C L IM A TE CHANGE IMPAC T S ON THE UNITED STAT E S

C L IM ATE CHANGE IMPAC TSON THE UNITED STAT ESThe Potential Consequences of Climate Variability and Change

Foundation

Humanity's influence on the global climate

will grow in the 21st century. Increasingly,

there will be significant climate-related

changes that will affect each one of us.

We must begin now to consider our

responses, as the actions taken today will

affect the quality of life for us and future

generations.

A Report of theNational Assessment Synthesis Team

US Global Change Research Program

PUBLISHED BY THE PRESS SYNDICATE OF THE UNIVERSITY OF CAMBRIDGEThe Pitt Building,Trumpington Street,Cambridge,United Kingdom

CAMBRIDGE UNIVERSITY PRESSThe Edinburgh Building,Cambridge,CB2 2RU, UK

40 West 20th Street,New York,NY 10011-4211,USA10 Stamford Road,Oakleigh,VIC 3166,Australia

Ruiz de Alarcón 13,28014 Madrid,SpainDock House,The Waterfront,Cape Town 8001,South Africa

http://www.cambridge.org

First published 2001

Printed in the United States of America

ISBN 0-521-00075-0 paperback

This report was produced by the National Assessment Synthesis Team,an advisorycommittee chartered under the Federal Advisory Committee Act to help the USGlobal Change Research Program fulfill its mandate under the Global ChangeResearch Act of 1990.The report was turned in to the Subcommittee on GlobalChange Research on October 31,2000. The National Science and TechnologyCouncil has forwarded this report to the President and Congress for their consider-ation as required by the Global Change Research Act.

Administrative support for the US Global Change Research Program is provided bythe University Corporation for Atmospheric Research,which is sponsored by theNational Science Foundation. Any opinions, findings and conclusions or recom-mendations expressed in this publication are those of the authors and do not nec-essarily reflect the views of the National Science Foundation or the UniversityCorporation for Atmospheric Research.

This report was prepared at the request of the US government and is therefore inthe public domain.However, some materials used in the report are copyrighted (asnoted in the figure captions) and permission was granted to the US government fortheir publication in this report. In all cases,credit must be given for copyrightedmaterials. For commercial uses that include copyrighted materials,permission forreproduction must be sought from the copyright holders.

The recommended citation of this report is as follows: National Assessment Synthesis Team

Climate Change Impacts on the United States:The Potential Consequences of Climate Variability and Change,

Report for the US Global Change Research Program,Cambridge University Press,Cambridge UK,620pp.,2001.

Comments on this report should be addressed to:Office of the US Global Change Research Program

400 Virginia Avenue SW, Suite 750,Washington DC 20024http://www.usgcrp.gov

The National Assessment of the Potential Consequences of Climate Variability and Change is alandmark in the major ongoing effort to understand what climate change means for the UnitedStates. Climate science is developing rapidly and scientists are increasingly able to project somechanges at the regional scale, identifying regional vulnerabilities, and assessing potential regionalimpacts. Science increasingly indicates that the Earth’s climate has changed in the past and con-tinues to change, and that even greater climate change is very likely in the 21st century. ThisAssessment has begun a national process of research, analysis, and dialogue about the comingchanges in climate, their impacts, and what Americans can do to adapt to an uncertain and continu-ously changing climate. This Assessment is built on a solid foundation of science conducted aspart of the United States Global Change Research Program (USGCRP).

This document is the Foundation report, which provides the scientific underpinnings for theAssessment. It has been prepared in cooperation with independent regional and sector assessmentteams under the leadership of the National Assessment Synthesis Team (NAST). The NAST is acommittee of experts drawn from governments, universities, industry, and non-governmental organ-izations. It has been responsible for preparing an Overview report aimed at general audiences andfor broad oversight of the Assessment along with the Federal agencies of the USGCRP. These twonational-level, peer-reviewed documents synthesize results from studies conducted by regional andsector teams, and from the broader scientific literature.

This Assessment was called for by a 1990 law, and has been conducted under the authority of theUSGCRP in response to a request from the President’s Science Advisor. The NAST developed theAssessment’s plan, which was then approved by the National Science and Technology Council, thecabinet level body of agencies responsible for scientific research, including global changeresearch, in the US government. We would like to acknowledge their contributions to this effort.The agencies and their representatives are listed in the appendix to this volume. Of particular notehave been Rosina Bierbaum and Peter Backlund of the Office of Science and Technology Policy,who provided consistent and helpful guidance throughout, and who organized our Oversight Board.In addition, Robert Corell (now a NAST Member), Aristides Patrinos, Paul Dresler, Richard Ball, JoelScheraga, and Tom Spence, along with many additional individuals, have played major roles onbehalf of the Subcommittee on Global Change Research, its National Assessment Working Group,and the ten cooperating agencies.

These assessment reports could not have been prepared without the extraordinary efforts of a largenumber of people. In addition to the members of the NAST, a number of individuals were entrainedinto development of the content and findings of the report, both as lead authors for the Overviewand as lead and contributing authors for the chapters in this Foundation report. We want to expressour sincere gratitude to these authors, the many names of whom are listed in the Overview reportand in the chapter headings of this book. Those playing particularly important roles in the prepara-tion of major sections of the Foundation report included Susan Bernard, Lynne Carter, DavidEasterling, Benjamin Felzer, John Field, Paul Grabhorn, Susan Jay Hassol, Schuyler Houser,Michael MacCracken, Michael McGeehin, Jonathan Patz, John Reilly, Joel Smith, Melissa Taylor,and Tom Wilbanks.

FOREWORD

The report itself is based in large part on workshops and assessment efforts of five sector teams and teamsin 20 regions across the US. Each of these groups has in turn involved many more experts from universi-ties, governments at various levels, public and private organizations, and others interested in or affected bythe changing global environment. All of these individuals have played an important role in developing andexpanding the dialogue on the potential impacts of climate change. We want to especially thank the variousregional and sector team leaders who are listed along with their team members in the chapters of thisreport.

In addition, we benefited from the comments of hundreds of reviewers, who helped encourage new insightsand new ways of thinking about and presenting the results of these studies. Of particular help were themembers of the Independent Review Board that was established by the President’s Committee of Adviserson Science and Technology. Co-chaired by Peter Raven and Mario Molina, the board included BurtonRichter, Linda Fisher, Kathryn Fuller, John Gibbons, Marcia McNutt, Sally Ride, William Schlesinger, JamesGustave Speth, and Robert White. They provided cogent and helpful comments throughout the many draftsof the assessment documents.

The complexity of coordinating the activities of far-flung authors, providing background data, managing theinputs and responses from hundreds of reviewers, designing the reports to be accurate, accessible, andappealing, and ensuring that the final products were printed under tight timetables was very challenging.Many people devoted their personal and professional attention to those tasks without asking for credit.Here we acknowledge their contributions and dedication to seeing this job through, and thank them, mostassuredly less than they deserve. Paul Grabhorn kept us focused on effectively communicating our mes-sage, helped us appreciate the importance of design, and he, Melody Warford, and their staff carried thisthrough with an inspired design implemented through layout, graphics, and production of the documents.Susan Joy Hassol, with the gracious cooperation of the Aspen Global Change Institute, played a major rolein making complex scientific issues more easily understood and helping our convoluted prose speak moreclearly.

The staff of the National Assessment Coordination Office (NACO) played an important role in facilitating theentire assessment process by supporting the activities not only of the NAST, but also by coordinating theefforts of the regions, sectors, and agencies. Under the leadership of Michael MacCracken, the coordinationand logistics associated with this very distributed effort came together. Melissa Taylor served as executivesecretary to the NAST through March 2000. Lynne Carter served as NACO liaison to the regions, JustinWettstein and LaShaunda Malone served as liaison to the sectors, and LaShaunda Malone also served asliaison with agencies and as coordinator for the various peer reviews. Thomas Wilbanks of Oak RidgeNational Laboratory (ORNL) served as chair of the Inter-regional Forum that helped to encourage and coor-dinate regional activities. In addition, Forrest Hoffman of ORNL handled the Web site through which muchof our information was distributed. The NACO staff were also assisted in their efforts by staff of the GlobalChange Research Information Office, including Robert Worrest, Annie Gerard, and Robert Bourdeau, whohave helped in the posting of the full report for public comment and access.

The assessment studies are based on extensive data sets of various types. Benjamin Felzer, with assis-tance of staff at the National Center for Atmospheric Research (NCAR), assembled and analyzed the datafrom climate models and prepared most of the climate graphics. David Easterling, Byron Gleason, and otherstaff at the National Climatic Data Center provided databases describing past changes in the climate. TimKittel at the National Center for Atmospheric Research was instrumental in carrying through the processing

of the climatic data to provide consistent sets for use across the US. We also very much appreciatethe willingness of colleagues at the various modeling centers to provide results of their simulations,including particularly David Viner at the University of East Anglia, Francis Zwiers and George Boerat the Canadian Centre for Climate Modelling and Analysis, and John Mitchell, Ruth Carnell, andJonathan Gregory at the Hadley Centre of the United Kingdom Meteorological Office. The availabili-ty of data for the assessment teams was made possible by Ben Felzer of NCAR and AnnetteSchloss and Denise Blaha of the University of New Hampshire.

Baseline distributions and simulations of changes in ecosystems were made available through theVegetation/Ecosystem Modeling and Analysis Project (VEMAP) and their many team members. TimKittel of NCAR graciously served as coordinator of our links to this effort. The social science datasets were provided by Nestor Terlickij of NPA Data Associates through an agreement with the OakRidge National Laboratory based on the efforts of David Vogt and Thomas Wilbanks. In addition,Robert Chen at the Consortium for International Earth Science Information Networks (CIESIN) pro-vided very helpful data sets on population and other social measures.

Many individuals have played important roles in carrying through the administrative aspects of thiseffort. We want to graciously acknowledge the contributions of Mary Ann Seifert of the MarineBiological Laboratory, Gracie Bermudez of the World Resources Institute, Rosalind Ledford of theNational Climatic Data Center, Nakia Dawkins and Robert Cherry of NACO, and Susan Henson,Karen York, and Matt Powell of the National Science Foundation, all of whom assisted in makingpossible our many meetings and exchanges of reports, among many other tasks. In addition, thestaff of the University Corporation for Atmospheric Research (UCAR) provided invaluable assis-tance with travel and contractual issues associated with the assessment process. Those playingparticularly helpful efforts have been Gene Martin, Kyle Terran, Tara Jay, Amy Smith, Chrystal Pene,James Menghi, and Brian Jackson.

Finally, as co-chairs of the National Assessment Synthesis Team, we would like to thank the othermembers of this team. We have had quite an adventure, working to develop and analyze informa-tion, working with fellow NAST members and leaders of assessment teams around the country, con-sidering and coming to agreement on findings, and writing and rewriting text in response to internaland external comments. Throughout there has been great comity, and we are very proud to havecome to full consensus on all of the findings. We want to thank all of you especially for devotingyour time and effort to this important effort; we know it has involved much more than any of youfirst thought, but we believe the product is also a very significant contribution to the Nation’sfuture.

Jerry MelilloAnthony JanetosThomas Karl

National Assessment Synthesis Team Members

Jerry M.Melillo,Co-chairEcosystems CenterMarine Biological Laboratory

Anthony C. Janetos,Co-chairWorld Resources Institute

Thomas R.Karl,Co-chairNOAA National Climatic Data Center

Eric J. BarronPennsylvania State University

Virginia BurkettUSGS National Wetlands Research Center

Thomas F. CecichGlaxo Wellcome Inc.

Robert Corell (from January 2000)American Meteorological Society andHarvard University

Katharine JacobsArizona Department of Water Resources

Linda JoyceUSDA Forest Service

Barbara MillerWorld Bank

M.Granger MorganCarnegie Mellon University

Edward A. Parson (until January 2000)Harvard University

Richard G. RichelsEPRI

David S. SchimelNational Center for Atmospheric Research

Independent Review Board of thePresident’s Committee of Advisers onScience and Technology (PCAST)

Peter Raven,Co-chairMissouri Botanical Garden and PCAST

Mario Molina,Co-chairMIT and PCAST

Burton RichterStanford University

Linda FisherMonsanto

Kathryn FullerWorld Wildlife Fund

John GibbonsNational Academy of Engineering

Marcia McNuttMonterey Bay Aquarium Research Institute

Sally RideUniversity of California San Diego and PCAST

William SchlesingerDuke University

James Gustave SpethYale University

Robert WhiteUniversity Corporation for AtmosphericResearch,and Washington Advisory Group

TABLE OF CONTENTS

About the Assessment Process . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Chapter 1: Scenarios for Climate Variability and Change . . . . . . . . . . . . . . . .

Chapter 2: Vegetation and Biogeochemical Scenarios . . . . . . . . . . . . . . . . . .

Chapter 3: The Socioeconomic Context for Climate Impact Assessment . . . .

Chapter 4: Northeast . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Chapter 5: Southeast . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Chapter 6: Midwest . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Chapter 7: Great Plains . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Chapter 8: West . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Chapter 9: Pacific Northwest . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Chapter 10: Alaska . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Chapter 11: Islands . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Chapter 12: Native Peoples and Homelands . . . . . . . . . . . . . . . . . . . . . . . . . .

Chapter 13: Agriculture . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Chapter 14: Water . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Chapter 15: Human Health . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Chapter 16: Coastal Areas and Marine Resources . . . . . . . . . . . . . . . . . . . . . .

Chapter 17: Forests . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Chapter 18: Conclusions & Research Pathways . . . . . . . . . . . . . . . . . . . . . . .

Appendix: USGCRP Leadership . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Color Figure Appendix . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

2

6

13

73

93

109

137

167

191

219

247

283

315

351

379

405

437

461

489

525

538

543

2

What is the purpose of thisAssessment?

The Assessment’s purpose is to synthe-size, evaluate, and report on what wepresently know about the potential con-sequences of climate variability andchange for the US in the 21st century. Ithas sought to identify key climatic vul-nerabilities of particular regions and sec-tors, in the context of other changes inthe nation’s environment, resources, andeconomy. It has also sought to identifypotential measures to adapt to climatevariability and change. Finally, becausepresent knowledge is limited, theAssessment has sought to identify thehighest priority uncertainties aboutwhich we must know more to understandclimate impacts, vulnerabilities, and ourability to adapt.

How did the process involveboth stakeholders and scien-tists in this Assessment?

This first National Assessment involvedboth stakeholders and scientific experts.Stakeholders included, for example, pub-lic and private decision-makers, resourceand environmental managers, and thegeneral public. The stakeholders fromdifferent regions and sectors began theAssessment by articulating their con-cerns in a series of workshops about cli-mate change impacts in the context ofthe other major issues they face. In theworkshops and subsequent consulta-tions, stakeholders identified priorityregional and sector concerns, mobilizedspecialized expertise, identified potentialadaptation options, and provided usefulinformation for decision-makers. TheAssessment also involved many scientif -ic experts using advanced methods,models, and results. Further, it has stim-ulated new scientific research in manyareas and identified priority needs forfurther research.

What is the breadth of thisAssessment?

Although global change embraces manyinterrelated issues, this first NationalAssessment has examined only climatechange and variability, with a primaryfocus on specific regions and sectors.In some cases, regional and sectoranalyses intersect and complement eachother. For example, the Forest sectorand the Pacific Northwest have both pro-vided insights into climate impacts onNorthwest forests.

The regions cover the nation. Impactsoutside the US are considered onlybriefly, with particular emphasis onpotential linkages to the US. Sectorteams examined Water, Agriculture,Human Health, Forests, and CoastalAreas and Marine Resources. This firstAssessment could not attempt to becomprehensive: the choice of these fivesectors reflected an expectation thatthey were likely to be both important andparticularly informative, and that relevantdata and analytic tools were available –not a conclusion that they are the onlyimportant domains of climate impact.Among the sectors considered, therewas a continuum in the amount of infor-mation available to support theAssessment, with some sectors being atfar earlier stages of development. Futureassessments should consider otherpotentially important issues, such asEnergy, Transportation, Urban Areas,and Wildlife.

Each regional and sector team is pub-lishing a separate report of its ownanalyses, some of which are still contin-uing. The Overview and Foundationreports consequently represent a snap-shot of our understanding at the presenttime.

After identifying potentialimpacts of climate change,what kinds of societalresponses does this reportexplore?

Responses to climate change can be oftwo broad types. One type involvesadaptation measures to reduce theharms and risks and maximize the bene-fits and opportunities of climate change,whatever its cause. The other typeinvolves mitigation measures to reducehuman contributions to climate change.After identifying potential impacts, thisAssessment sought to identify potentialadaptation measures for each region andsector studied. While this was an impor-tant first step, it was not possible at thisstage to evaluate the practicality, effec-tiveness, or costs of the potential adap-tation measures. Both mitigation andadaptation measures are necessary ele -ments of a coherent and integratedresponse to climate change. Mitigationmeasures were not included in thisAssessment but are being assessed inother bodies such as the United NationsIntergovernmental Panel on ClimateChange (IPCC).

ABOUT THE ASSESSMENT PROCESS

Does the fact that this reportexcludes mitigation meanthat nothing can be done toreduce climate change?

No. An integrated climate policy willcombine mitigation and adaptationmeasures as appropriate. If future worldemissions of greenhouse gases arelower than currently projected, for what-ever reason, including intentional mitiga-tion, then the rate of climate change, theassociated impacts, and the cost and dif-ficulty of adapting will all be reduced. Ifemissions are higher than expected, thenthe rate of change, the impacts, and thedifficulty of adapting will be increased.But no matter how aggressively emis-sions are reduced, the world will stillexperience at least a century of climatechange. This will happen because theelevated concentrations of greenhousegases already in the atmosphere willremain for many decades, and becausethe climate system responds to changesin human inputs only very slowly.Consequently, even if the world takesmitigation measures, we must still adaptto a changing climate. Similarly, even ifwe take adaptation measures, futureemissions will have to be curbed to sta-bilize climate. Neither type of responsecan completely supplant the other.

How are computer modelsused in this Assessment?

State-of-the-science climate models havebeen used to generate climate changescenarios. Computer models of ecologi-cal systems, hydrological systems, andvarious socioeconomic systems havealso been used in the Assessment tostudy responses of these systems to thescenarios generated by climate models.

What additional tools,besides models, were usedto evaluate potential climatechange impacts?

In addition to models, the Assessmenthas used two other ways to think aboutpotential future climate. First, theAssessment has used historical climaterecords to evaluate sensitivities ofregions and sectors to climate variabilityand extremes that have occurred in the20th century. Looking at real historicalclimate events, their impacts, and howpeople have adapted, gives valuableinsights into potential future impacts thatcomplement those provided by modelprojections. In addition, the Assessmenthas used sensitivity analyses, which askhow, and how much, the climate wouldhave to change to bring about majorimpacts on particular regions or sectors.For example, how much would tempera-ture have to increase in the South beforeagricultural crops such as soybeanswould be negatively affected? Whatwould be the result for forest productivi-ty of continued increases in temperatureand leveling off of the CO2 fertilizationeffect?

Has this report been peerreviewed?

This Overview and the underlyingFoundation document have been exten-sively reviewed. More than 300 scientificand technical experts have provideddetailed comments on part or all of thereport in two separate technicalreviews. The report was reviewed ateach stage for technical accuracy bythe agencies of the US Global ChangeResearch Program. The public also pro-vided hundreds of helpful suggestionsfor clarification and modification duringa 60-day public comment period. Apanel of distinguished experts convenedby the President's Committee ofAdvisors on Science and Technologyhas provided broad oversight and moni-tored the authors response to allreviews.

3

4

What are scenarios and why are they used?

Scenarios are plausible alternative futures – each an example of what might happen under particularassumptions. Scenarios are not specific predictions or forecasts. Rather, scenarios provide a startingpoint for examining questions about an uncertain future and can help us visualize alternative futures inconcrete and human terms. The military and industry frequently use these powerful tools for future plan-ning in high-stakes situations. Using scenarios helps to identify vulnerabilities and plan for contingencies.

Why are climate scenarios used in this Assessment and how were theydeveloped?

Because we cannot predict many aspects of our nation's future climate, we have used scenarios to helpexplore US vulnerability to climate change. Results from state-of-the-science climate models and datafrom historical observations have been used to generate a variety of such scenarios. Projections ofchanges in climate from the Hadley Centre in the United Kingdom and the Canadian Centre for ClimateModeling and Analysis served as the primary resources for this Assessment. Results were also drawnfrom models developed at the National Center for Atmospheric Research, NOAA's Geophysical FluidDynamics Laboratory, and NASA's Goddard Institute for Space Studies.

For some aspects of climate, virtually all models, as well as other lines of evidence, agree on the types ofchanges to be expected. For example, all climate models suggest that the climate is going to get warmer,the heat index is going to rise, and precipitation is more likely to come in heavy and extreme events. Thisconsistency lends confidence to these results.

For some other aspects of climate, however, the model results differ. For example, some models, includ-ing the Canadian model, project more extensive and frequent drought in the US, while others, including theHadley model, do not. The Canadian model suggests a drier Southeast in the 21st century while the Hadleymodel suggests a wetter one. In such cases, the scenarios provide two plausible but different alternatives.Such differences can help identify areas in which the models need improvement.

Many of the maps in this document are derived from the two primary climate model scenarios. In mostcases, there are three maps: one shows average conditions based on actual observations from 1961-1990;the other two are generated by the Hadley and Canadian model scenarios and reflect the models’ projec-tions of change from those average conditions.

What assumptions about emissions are in these two climate scenarios?

Because future trends in fossil fuel use and other human activities are uncertain, the IntergovernmentalPanel on Climate Change (IPCC) has developed a set of scenarios for how the 21st century may evolve.These scenarios consider a wide range of possibilities for changes in population, economic growth, tech-nological development, improvements in energy efficiency, and the like. The two primary climate scenariosused in this Assessment are based on one mid-range emissions scenario for the future that assumes nomajor changes in policies to limit greenhouse gas emissions. Some other important assumptions in thisscenario are that by the year 2100:

• world population will nearly double to about 11 billion people;• the global economy will continue to grow at about the average rate it has been growing,

reaching more than ten times its present size;• increased use of fossil fuels will triple CO2 emissions and raise sulfur dioxide emissions,

resulting in an atmospheric CO2 concentration of just over 700 parts per million; and• total energy produced each year from non-fossil sources such as wind, solar, biomass, hydroelectric,

and nuclear will increase to more than ten times its current amount, providing more than 40% of the world’s energy, rather than the current 10%.

ABOUT SCENARIOS AND UNCERTAINTY

Many of the mapsin this documentare derived fromthe two primary cli-mate model sce-narios. In mostcases, there arethree maps: oneshows averageconditions basedon actual observa-tions from 1961-1990; the other twoare generated bythe Hadley andCanadian modelscenarios andreflect the models’projections ofchange from pres-ent day conditions.

5

How is the likelihood of various impacts expressed?

To integrate a wide variety of information and differentiate more likely from less likely outcomes, the NASTdeveloped a common language to express the team's considered judgement about the likelihood of results.The NAST developed their collective judgements through discussion and consideration of the supportinginformation. Historical data, model projections, published scientific literature, and other available informa-tion all provided input to these deliberations, except where specifically stated that the result comes from aparticular model scenario. In developing these judgements, there were often several lines of supportingevidence (e.g., drawn from observed trends, analytic studies, model simulations). Many of these judge-ments were based on broad scientific consensus as stated by well-recognized authorities including theIPCC and the National Research Council. In many cases, groups outside the NAST reviewed the use ofterms to provide input from a broader set of experts in a particular field.

Language Used to Express Considered Judgement

The Assessment’s Emissions Scenario Falls in the Middle of the other IPCC Emissions Scenarios

0% 50% 100%

“LITTLE CHANCE”OR

“VERY UNLIKELY”

“UNLIKELY”OR

“SOME CHANCE”“POSSIBLE”

“LIKELY”OR

“PROBABLE”

“VERY LIKELY”OR

“VERY PROBABLE”

Common Language

Likelihood

The graph shows a comparisonof the projections of total car-bon dioxide emissions (in bil -lions of metric tons of carbon,GtC) and the human-inducedwarming influence due to allthe greenhouse gases and sul-fate aerosols for the emissionsscenarios prepared by the IPCCin 1992 and 2000. As is appar-ent from the graph, both theemissions scenario and thehuman-induced warming influ-ence assumed in thisAssessment lie near the mid-range of the set of IPCC sce-narios. Further detail can befound in the Climate chapter .See color figure section.

Both the emissionsscenario and the

human-inducedwarming influence

assumed in thisAssessment lie

near the mid-rangeof the set of IPCC

scenarios.

8

7

6

5

4

3

2

1

0

600 1000

Cumulative Carbon Emissions during 21st Century (GtC)

IPCC 2000 Emissions Scenarios for Year 2100

IPCC 1992 Emissions Scenarios for Year 2100

Hadley Centre Model Emissions Scenario for Year 2100

Canadian Centre Model Emissions Scenario for Year 2100

1600 2000 2600

SUMMARYCLIMATE CHANGE AND OUR NATION

ong-term observations confirm that our climate is now changing at a rapid rate.Over the 20th century, the average annual US temperature has risen by almost

1°F (0.6°C) and precipitation has increased nationally by 5 to 10%,mostly due toincreases in heavy downpours. These trends are most apparent over the past fewdecades. The science indicates that the warming in the 21st century will be signifi-cantly greater than in the 20th century. Scenarios examined in this Assessment,which assume no major interventions to reduce continued growth of world green-house gas emissions,indicate that temperatures in the US will rise by about 5-9°F(3-5°C) on average in the next 100 years,which is more than the projected globalincrease. This rise is very likely to be associated with more extreme precipitationand faster evaporation of water, leading to greater frequency of both very wet andvery dry conditions.

This Assessment reveals a number of national-level impacts of climate variability andchange including impacts to natural ecosystems and water resources. Naturalecosystems appear to be the most vulnerable to the harmful effects of climatechange,as there is often little that can be done to help them adapt to the projectedspeed and amount of change. Some ecosystems that are already constrained by cli-mate,such as alpine meadows in the Rocky Mountains,are likely to face extremestress,and disappear entirely in some places. It is likely that other more widespreadecosystems will also be vulnerable to climate change. One of the climate scenariosused in this Assessment suggests the potential for the forests of the Southeast tobreak up into a mosaic of forests,savannas,and grasslands. Climate scenarios sug-gest likely changes in the species composition of the Northeast forests,includingthe loss of sugar maples. Major alterations to natural ecosystems due to climatechange could possibly have negative consequences for our economy, whichdepends in part on the sustained bounty of our nation’s lands, waters,and nativeplant and animal communities.

A unique contribution of this first US Assessment is that it combines national-scaleanalysis with an examination of the potential impacts of climate change on differentregions of the US. For example,sea-level rise will very likely cause further loss ofcoastal wetlands (ecosystems that provide vital nurseries and habitats for many fishspecies) and put coastal communities at greater risk of storm surges,especially inthe Southeast. Reduction in snowpack will very likely alter the timing and amountof water supplies,potentially exacerbating water shortages and conflicts,particular-ly throughout the western US. The melting of glaciers in the high-elevation Westand in Alaska represents the loss or diminishment of unique national treasures ofthe American landscape. Large increases in the heat index (which combines tem-perature and humidity) and increases in the frequency of heat waves are very likely.These changes will,at minimum,increase discomfort,particularly in cities. It is veryprobable that continued thawing of permafrost and melting of sea ice in Alaska willfurther damage forests,buildings, roads,and coastlines,and harm subsistence liveli-hoods. In various parts of the nation,cold-weather recreation such as skiing willvery likely be reduced,and air conditioning usage will very likely increase.

L

The findings in this reportare based on a synthesisof historical data, modelprojections, published sci-entific research, and otheravailable information,except where specificallynoted.

6

Highly managed ecosystems appear more robust, and some potential bene-fits have been identified. Crop and forest productivity is likely to increase insome areas for the next few decades due to increased carbon dioxide in theatmosphere and an extended growing season. It is possible that some USfood exports could increase,depending on impacts in other food-growingregions around the world. It is also possible that a rise in crop productionin fertile areas could cause prices to fall,benefiting consumers. Other bene-fits that are possible include extended seasons for construction and warmweather recreation, reduced heating requirements,and reduced cold-weath-er mortality.

Climate variability and change will interact with other environmental stress-es and socioeconomic changes. Air and water pollution,habitat fragmenta-tion, wetland loss,coastal erosion,and reductions in fisheries are likely to becompounded by climate-related stresses. An aging populace nationally, andrapidly growing populations in cities,coastal areas,and across the South andWest,are social factors that interact with and alter sensitivity to climate vari-ability and change.

There are also very likely to be unanticipated impacts of climate change dur-ing the 21st century. Such "surprises" may stem from unforeseen changes inthe physical climate system,such as major alterations in ocean circulation,cloud distribution,or storms;and unpredicted biological consequences ofthese physical climate changes,such as massive dislocations of speciesor pest outbreaks. In addition,unexpected social or economicchanges,including major shifts in wealth,technology, or politi-cal priorities,could affect our ability to respond to climatechange.

Greenhouse gas emissions lower than thoseassumed in this Assessment would result inreduced impacts. The signatory nations of theFramework Convention on Climate Changeare negotiating the path they will ultimatelytake. Even with such reductions,however,the planet and the nation are certain toexperience more than a century of cli-mate change,due to the long lifetimes ofgreenhouse gases already in the atmos-phere and the momentum of the climatesystem. Adapting to a changed climate isconsequently a necessary component ofour response strategy.

The warming in the 21st centu-ry will be significantly greaterthan in the 20 th century.

Natural ecosystems, which areour life support system inmany important ways, appearto be the most vulnerable tothe harmful effects of climatechange...

Major alterations to naturalecosystems due to climatechange could possibly havenegative consequences for oureconomy, which depends inpart on the sustained bounty ofour nation’s lands, waters, andnative plant and animal com-munities.

7

8

SUMMARYCLIMATE CHANGE AND OUR NATION

Adaptation measures can,in many cases, reduce the magnitude of harmful impactsor take advantage of beneficial impacts. For example,in agriculture,many farmerswill probably be able to alter cropping and management practices. Roads,bridges,buildings,and other long-lived infrastructure can be designed taking projected cli-mate change into account. Adaptations,however, can involve trade-offs,and doinvolve costs. For example,the benefits of building sea walls to prevent sea-levelrise from disrupting human coastal communities will need to be weighed againstthe economic and ecological costs of seawall construction. The ecological costscould be high as seawalls prevent the inland shifting of coastal wetlands inresponse to sea-level rise, resulting in the loss of vital fish and bird habitat and otherwetland functions,such as protecting shorelines from damage due to storm surges.Protecting against any increased risk of water-borne and insect-borne diseases willrequire diligent maintenance of our public health system. Many adaptations,notably those that seek to reduce other environmental stresses such as pollutionand habitat fragmentation,will have beneficial effects beyond those related to cli-mate change.

Vulnerability in the US is linked to the fates of other nations,and we cannot evalu-ate national consequences due to climate variability and change without also con-sidering the consequences of changes elswhere in the world. The US is linked toother nations in many ways,and both our vulnerabilities and our potential respons-es will likely depend in part on impacts and responses in other nations. For exam-ple,conflicts or mass migrations resulting from resource limits,health,and environ-mental stresses in more vulnerable nations could possibly pose challenges for globalsecurity and US policy. Effects of climate variability and change on US agriculturewill depend critically on changes in agricultural productivity elsewhere,which canshift international patterns of food supply and demand. Climate-induced changes inwater resources available for power generation,transportation,cities,and agricul-ture are likely to raise potentially delicate diplomatic issues with both Canada andMexico.

This Assessment has identified many remaining uncertainties that limit our ability tounderstand fully the spectrum of potential consequences of climate change for ournation. To address these uncertainties,additional research is needed to improve ourunderstanding of ecological and social processes that are sensitive to climate, waysof applying climate scenarios and reconstructions of past climates to the study ofimpacts,and assessment strategies and methods. Results from these research effortswill inform future assessments that will continue the process of building our under-standing of humanity's impacts on climate,and climate's impacts on us.

The magnitude of climatechange impacts depends ontime period and geographicscale. Short-term impacts dif -fer from long-term impacts,and regional and local levelimpacts are much more pro-nounced than those at thenational level.

For the nation as a whole,direct economic impacts arelikely to be modest, while insome places, economic loss-es or gains are likely to belarge. For example, whilecrop yields are likely toincrease at the national scaleover the next few decades,large increases or decreasesin yields of specific crops inparticular places are likely.

Through time, climate changewill possibly affect the sameresource in opposite ways.For example, forest productiv-ity is likely to increase in theshort term, while over thelonger term, changes inprocesses such as fire,insects, drought, and diseasewill possibly decrease forestproductivity.

9

KEY FINDINGS

1. Increased warming Assuming continued growth in world greenhouse gas emissions, the primary climate models used in this Assessment projectthat temperatures in the US will rise 5-9ºF (3-5ºC) on average in the next 100 years. A wider range of outcomes is possible.

2. Differing regional impacts Climate change will vary widely across the US. Temperature increases will vary somewhat from one region to the next. Heavyand extreme precipitation events are likely to become more frequent, yet some regions will get drier. The potential impacts ofclimate change will also vary widely across the nation.

3. Vulnerable ecosystems Many ecosystems are highly vulnerable to the projected rate and magnitude of climate change. A few, such as alpine meadowsin the Rocky Mountains and some barrier islands, are likely to disappear entirely in some areas. Others, such as forests of theSoutheast, are likely to experience major species shifts or break up into a mosaic of grasslands, woodlands, and forests. Thegoods and services lost through the disappearance or fragmentation of certain ecosystems are likely to be costly or impossibleto replace.

4. Widespread water concerns Water is an issue in every region, but the nature of the vulnerabilities varies. Drought is an important concern in every region.Floods and water quality are concerns in many regions. Snowpack changes are especially important in the West, PacificNorthwest, and Alaska.

5. Secure food supplyAt the national level, the agriculture sector is likely to be able to adapt to climate change. Overall, US crop productivity is verylikely to increase over the next few decades, but the gains will not be uniform across the nation. Falling prices and competitivepressures are very likely to stress some farmers, while benefiting consumers.

6. Near-term increase in forest growthForest productivity is likely to increase over the next several decades in some areas as trees respond to higher carbon dioxidelevels. Over the longer term, changes in larger-scale processes such as fire, insects, droughts, and disease will possiblydecrease forest productivity. In addition, climate change is likely to cause long-term shifts in forest species, such as sugarmaples moving north out of the US.

7. Increased damage in coastal and permafrost areasClimate change and the resulting rise in sea level are likely to exacerbate threats to buildings, roads, powerlines, and otherinfrastructure in climatically sensitive places. For example, infrastructure damage is related to permafrost melting in Alaska, andto sea-level rise and storm surge in low-lying coastal areas.

8. Adaptation determines health outcomes A range of negative health impacts is possible from climate change, but adaptation is likely to help protect much of the US pop-ulation. Maintaining our nation's public health and community infrastructure, from water treatment systems to emergency shel-ters, will be important for minimizing the impacts of water-borne diseases, heat stress, air pollution, extreme weather events,and diseases transmitted by insects, ticks, and rodents.

9. Other stresses magnified by climate changeClimate change will very likely magnify the cumulative impacts of other stresses, such as air and water pollution and habitatdestruction due to human development patterns. For some systems, such as coral reefs, the combined effects of climatechange and other stresses are very likely to exceed a critical threshold, bringing large, possibly irreversible impacts.

10. Uncertainties remain and surprises are expected Significant uncertainties remain in the science underlying regional climate changes and their impacts. Further research wouldimprove understanding and our ability to project societal and ecosystem impacts and to provide the public with additional usefulinformation about options for adaptation. However, it is likely that some aspects and impacts of climate change will be totallyunanticipated as complex systems respond to ongoing climate change in unforeseeable ways.

PERMAFROST AREAS

It is very probable that ris-ing temperatures will causefurther permafrost thawing,damagingroads,buildings,andforests inAlaska.

SPECIES DIVERSITY

While it is possible that somespecies will adapt to changes inclimate by shifting their ranges,human and geographic barri-ers,and the presenceof invasive non-native specieswill limit thedegree of adap-tation that canoccur. Losses inlocal biodiversityare likely to accelerate towardsthe end of the 21st century.

FORESTRY

Timber inventories are likely toincrease over the 21st century.Hardwood productivity is like-ly to increasemore thansoftwoodproductivityin someregions,including theSoutheast.

10

IMPACTS OF CLIMATE CHANGEIt is very likely that the US will get substantially warmer. Temperatures areprojected to rise more rapidly in the next one hundred years than in the last10,000 years. It is also very likely that there will be more precipitationoverall, with more of it coming in heavy downpours. In spite of this, someareas are likely to get drier as increased evaporation due to higher temper-atures outpaces increased precipitation. Droughts and flash floods are like-ly to become more frequent and intense.

10

FRESHWATER ECOSYSTEMS

Increases in water temperature andchanges in seasonal pat-terns of runoff willvery likely disturbfish habitat andaffect recre-ational uses oflakes,streams,and wetlands.

ISLANDS

Sea-level rise and stormsurges will very likelythreaten public health andsafety and pos-sibly reducethe availabil-ity of freshwater.

CORAL REEFS

Increased CO2 and oceantemperatures,especially com-bined with other stresses,will possiblyexacerbatecoral reefbleachingand die-off.

WATER SUPPLY

Reduced summer runoff,increased winter runoff, and

increased demands are likely tocompound current stresses on

water supplies and flood manage-ment,especially in the western US.

COASTAL COMMUNITIES ANDINFRASTRUCTURE

Coastal inundation from storm surgescombined with rising sea level willvery likely increase threats to waterand sewer systems,transportation andcommunication systems,homes,andother buildings.

EXTREME EVENTS

It is very likely that morerain will come in heavydownpours,increasingthe risk of flash floods.

FOREST ECOSYSTEMS

Forest growth is likely to increase in many regions,at least over the next several decades. Over thenext century, tree and animal species’ ranges willprobably shift in response to the changing cli-mate. Some forests are likely to become more sus-ceptible to fire and pests.

AGRICULTURE

The Nation's food sup-ply is likely toremain secure.Theprices paid byconsumers andthe profit mar-gins for food pro-ducers are likely tocontinue to drop.

HUMAN POPULATIONS

Heat waves are very likely toincrease in frequency, resultingin more heat-related stresses.Milder winters are likely toreduce cold-related stresses insome areas.

RARE ECOSYSTEMS

Alpine meadows,mangroves,andtropical mountain forests in somelocations are likely to disappearbecause the newlocal climatewill not sup-port them orthere are barri-ers to theirmovement.

AdaptationThere are substantial opportunities to minimize the negativeimpacts and maximize the benefits of climate change throughadaptation. Examples include cultivating varieties of crops,trees, and livestock that are better suited to hotter conditions.This report includes an initial identification of potential adap-tation strategies, but an analysis of their effectiveness, practi-cality, and costs was not considered in this Assessment.

11

COASTAL ECOSYSTEMS

Sea-level rise is very likelyto cause the loss of somebarrier beaches,islands,marshes,and coastal forests,throughout the 21st century.

CHAPTER 1

SCENARIOS FOR CLIMATE VARIABILITYAND CHANGEMichael MacCracken1,2, Eric Barron3, David Easterling1,4, Benjamin Felzer1,5, andThomas Karl4

Contents of this Chapter

Summary:Scenarios for Climate Variability and Change

Introduction

Climate and the Greenhouse Ef fect

Human Activities and Changes in Atmospheric Composition

Human Activities and Climate Change

Approaches for Assessing the Impacts of Climate Change

Use of Historical Records

Use of Climate Model Simulations

Use of Vulnerability Analyses

Trends in Climate over the US during the 20th Century

Climate Model Simulations Used in the National Assessment

Climate Model Simulations of the 20th Century for the US

Scenarios for Changes in Atmospheric Composition and Radiative Forcing for the 21st

Century

Climate Model Scenarios of Changes in Temperature,Precipitation,Soil Moisture,and Sea

Level for the 21st Century

Climate Model Scenarios of Changes in Climatic Patterns,Variability, Storms,and Extremes

for the 21st Century

The Climatic Effects of Stabilizing the Carbon Dioxide Concentration

Crucial Unknowns and Research Needs

Summary

Literature Cited

Acknowledgments

1Coordinating author for the National Assessment Synthesis Team; 2Lawrence Livermore National Laboratory, onassignment to the National Assessment Coordination Office of the US Global Change Research Program;3Pennsylvania State University; 4National Climatic Data Center; 5National Center for Atmospheric Research 13

14

Potential Consequences of Climate Variability and Change

CHAPTER SUMMARY Climate of the Past Century

• Since the mid-1800s,the global average tempera-ture has warmed by about 1˚F (about 0.6˚C).The Northern Hemisphere average temperatureduring the 1990s is almost 1.5˚F (about 0.9˚C)warmer than during the few centuries prior tothe Industrial Revolution. While some of thiswarming may be due to an intensification ofsolar radiation and a small portion due to urbanwarming,a variety of analyses indicate that thecurrent warming is too large to be explained bynatural fluctuations alone. The observed magni-tude,pattern,and timing of the global warmingindicate that the rising concentrations of CO2

and other greenhouse gases caused by humanactivities are contributing significantly to therecent warming.

• During the 20th century, the average temperatureover the US increased by about 1˚F (0.6˚C),withsome regions warming as much as 4˚F (about2.4˚C) and some other regions showing slightcooling. In general,nighttime minimum tempera-tures rose more than daytime maximums,andwintertime temperatures rose more than thoseof summertime. Total annual precipitation alsoincreased,with most of the increase occurring inheavy precipitation events.

• Reconstructions of the climate of the past thou-sand years using ice cores,tree rings, vegetationtypes,and other proxy measures suggest that thewarming of the 20th century is unprecedentedcompared to natural variations prior to this cen-tury that were presumably caused by solar, vol-canic,and other natural influences. In addition,the current warming is much more extensiveand intense than the regional scale warming thatpeaked about 1000 years ago in Europe duringwhat is referred to as the Medieval Warm Period.The recent warming is also far more than can becharacterized as a recovery from the cool condi-tions centered in Europe and the North Atlanticregion a few hundred years ago that are oftenreferred to as the Little Ice Age. Looking backover the few thousand years for which we areable to provide some reconstruction of the tem-perature record,the current global warmthappears unprecedented.

• An ice-core record from Antarctica covering thepast 420,000 years indicates that temperatures inthat region have been up to about 10˚F (6˚C)colder than present values for about 90% of the

Climate Context

Climate1 provides the context for the environmentand for many human activities — changes in the cli-mate will thus have consequences for the environ-ment and for human activities. While solar radiationis the primary energy source for maintaining theEarth’s temperature,the atmospheric concentrationsof water vapor, carbon dioxide (CO2),methane(CH4),and other gases determine the intensity ofthe natural greenhouse effect that currently keepsthe Earth’s surface temperature at about 58˚F(14˚C). Without this natural greenhouse effect,theEarth’s surface temperature would be about 0˚F(about –18˚C),a temperature that would make theEarth uninhabitable for life as we know it. Over thelast 150 years,combustion of coal,oil,and naturalgas (collectively called fossil fuels),deforestation,plowing of soils,and various industrial activitieshave led,among other changes,to increases in theatmospheric concentrations of critical greenhousegases. In particular, the CO2 concentration hasincreased by about 30% and the CH4 concentrationby about 150%. The warming influence of thesechanges,amplified by associated increases in theatmospheric water vapor concentration,have inten-sified the natural greenhouse effect and initiatedchanges in the climate.

1 Throughout the National Assessment reports,the term “climate”isintended to include both climate variability and climate change.“Climate change” refers to long-term or persistent trends (over decadesor more) or shifts in climate,while “climate variability” refers to short-term (generally decadal or less) climate fluctuations.

420,000-year period. During these cold periods,massive glaciers covered much of the land areaof the Northern Hemisphere (e.g.,covering east-ern North America with roughly a mile of ice tosouth of the Great Lakes), even though globaltemperatures were only several degrees colder.Evidence suggests that these variations havebeen driven primarily by changes in the seasonaland latitudinal distribution of solar radiationcaused by cyclic variations in the Earth’s orbitaround the Sun,but amplified by a number offactors. These additional factors include changesin glacial height and extent,in ocean circulation,and in the atmospheric CO2 and CH4 concentra-tions that were apparently driven by the initialtemperature change.

• The geological record indicates that the globalclimate has varied markedly over the past billionor more years. It appears that these natural varia-tions resulted from changes in identifiable factorsthat still determine climatic conditions today.These factors include the amount of solar radia-tion and shape of the Earth’s orbit around theSun,the gas and particle composition of theatmosphere (which determines the efficiency ofthe absorption and reflection of incoming solarenergy),the geographical pattern of land andocean,the heights of mountains,the directionand intensity of ocean currents,the chaoticnature of the interactions among the atmos-phere,land,and oceans,and more. The geologi-cal record clearly indicates that changes in thesefactors can cause significant changes in climate.

Climate of the ComingCentury

• Projections of the expanding uses of coal,oil,andnatural gas as sources of energy indicate thathuman activities will cause the atmospheric CO2

concentration to rise to between 2 and 3 timesits preindustrial level by the end of the 21st cen-tury unless very significant control measures areinitiated. The concentrations of CH4 and someother greenhouse gases are also projected torise,whereas controls on chlorofluorocarbonemissions are expected to allow their concentra-tions to fall.

• The ongoing effects of past increases in the con-centrations of greenhouse gases and the changesprojected for the 21st century are very likely tocause the world to warm substantially in compar-ison to natural fluctuations that have been expe-rienced over the past 1000 years. Model-basedprojections for a mid-range emissions scenarioare that the global average temperature is likelyto rise by about 2 to 6˚F (about 1.2 to 3.5˚C),with a central estimate of almost 4˚F (2.4˚C), bythe end of the 21st century. The range of theseestimates depends about equally on ranges in theestimates of climate sensitivity and of growth infossil fuel emissions.

• For the mid-range emissions scenario,the pro-jected warming is likely to be greater in mid andhigh latitudes than for the globe as a whole,andwarming is likely to be greater over continentsthan over oceans. For this mid-range emissionsscenario,the models used for this Assessmentproject that the average warming over the USwould be in the range of about 5 to 9˚F (about2.8 to 5˚C). However, given the wide range ofpossible emissions scenarios and uncertainties inthe sensitivity of the climate to emissions scenar-ios,it is possible that the actual increase in UStemperatures could be higher or lower than indi-cated by this range.

• A warming of 5 to 9˚F (2.8 to 5˚C) would beapproximately equivalent to the annual averagetemperature difference between the northernand central tier of states,or the central andsouthern tier of states. Wintertime warming isprojected to be greater than summertime warm-ing and nighttime warming greater than daytimewarming.

• Even though less warming is projected in sum-mertime than in wintertime,the summertime

15

Chapter 1 / Scenarios for Climate Variability and Change

16


heat index,which combines the effects of heatand humidity into an effective temperature,isprojected to rise anywhere from 5 to 15˚F (oreven more for some scenarios) over much of theeastern half of the country, especially across thesoutheastern part of the country. If the project-ed rise in the heat index were to occur, summer-time conditions for New York City could becomelike those now experienced in Atlanta,those inAtlanta like those now experienced in Houston,and those in Houston like those in Panama.

• The amount of rainfall over the globe is also verylikely to rise because global warming willincrease evaporation;however, the pattern ofchanges is likely to vary depending on latitudeand geography as storm tracks are altered. Modelprojections of possible changes in annual precip-itation across the US are generally mixed. Resultsfrom the two models used in the NationalAssessment tend to agree that there is likely tobe an increase in precipitation in the southwest-ern US as Pacific Ocean temperatures increase,but do not provide a clear indication of the trendin the southeastern US.

• It is likely that the observed trends toward anintensification of precipitation events will contin-ue. Thunderstorms and other intensive rainevents are likely to produce larger rainfall totals.While it is not yet clear how the numbers andtracks of hurricanes will change,projections arethat peak windspeed and rainfall intensity arelikely to rise significantly.

• Although overall precipitation is likely toincrease across the US,the higher temperatureswill increase evaporation. Even with a modest

increase in precipitation,the increase in the rateof evaporation is expected to cause reductions insummertime soil moisture,particularly in thecentral and southern US.

• Sea level,which has risen about 4 to 8 inches(10-20 cm) over the past century, is projected toincrease by 5 to 37 inches (13-95 cm) over thecoming century, with a central estimate of about20 inches (50 cm). The range is so broadbecause of uncertainties concerning what mighthappen to the Antarctic and Greenland ice caps.To determine the amount of sea-level rise in par-ticular regions,the global rise in sea level mustbe adjusted by the local rise or sinking of coastallands.

• Limitations in scientific understanding mean thatthe potential exists for surprises or unexpectedevents to occur, for thresholds to be crossed,andfor nonlinearities to develop. Such surpriseshave the potential of either amplifying projectedchanges or, in rarer cases,moderating the poten-tial changes in climate. Examples might includeamplified rates of sea-level rise if deterioration ofthe Greenland or Antarctic ice caps is accelerat-ed;limited warming or perhaps even cooling insome regions if ocean currents and deep oceanoverturning is suppressed;disappearance ofArctic sea ice over a few decades;sufficientwarming of methane trapped in frozen soils toallow its release and subsequent amplification ofthe warming rate,etc. While such possibilitiescould cause large impacts,estimating the likeli-hood of their occurrence is presently highlyproblematic,making risk assessments quite diffi-cult.

17


ter draw upon the basis of scientific understandingdescribed in these and related reports and therecent scientific literature,providing a limited set ofcitations that can be expanded upon by reference tothese assessments.

Although these scientific studies indicate that thefuture will be different from the past,determininghow different it will be and the significance of thesedifferences presents a tremendous scientific chal-lenge. The future will be affected by how the cli -mate varies due to natural and human influences,how the environment may respond to climatechange and to other factors,and how society mayevolve due to a myriad of influences,including cli-mate variability and change. Quite clearly,definitivepredictions cannot be made,being too dependenton factors ranging from uncertainties introduced byour growing,but limited,understanding of the cli-mate system to the complexities introduced by thepace of technological development and social evolu-tion.

Given the seriousness and strength of the projec-tions of climate change arising from the scientificcommunity and from careful assessments,prudentrisk management led Congress in 1990 to call forassessments of the potential impacts of climatechange. During the 1990s,scientific assessmentshave focused on the global-scale consequences ofhuman activities,leading to the conclusion that “thebalance of evidence suggests a discernible humaninfluence on the global climate”(IPCC,1996a).IPCC assessments of the consequences of climatechange have also indicated that potentially impor-tant consequences could arise (IPCC,1996b,1996c).It was these global-scale findings that indicated boththe need for and the possibility of being able to con-duct an assessment of the potential consequencesof climate variability and change for the UnitedStates.

As a basis for this Assessment,and in the context ofthe uncertainties inherent in looking forward 100years,Assessment teams are pursing a three-prongedapproach to considering how much the climate maychange. The three approaches involve use of:(1)historical data to examine the continuation oftrends or recurrence of past climatic extremes;(2)

INTRODUCTIONThis National Assessment is charged with evaluatingand summarizing the potential consequences of cli-mate variability and change for the United Statesover the next 100 years (Dresler et al.,1998).Studies of the interactions of climate with both theenvironment and with societal activities show clear-ly that there are important interconnections. Thevery hot and dry conditions of the 1930s,coupledwith poor land management practices,not only cre-ated Dust Bowl conditions on the Great Plains,butalso led large numbers of people to migrate fromthe central US to settle in the Southwest andCalifornia. Drought conditions in 1988 and floodconditions in 1993 had devastating effects on manyregions in the upper Mississippi River basin.Climate variations along the West Coast have led toyears of drought (with subsequent fires) and offlood (with subsequent mudslides). It is these manyinteractions that have led to the focus on what willhappen in the future as climate variations continueand as human activities believed to be capable ofaltering the climate continue.

The hypothesis that human activities could be influ-encing the global climate was first postulated morethan a century ago (Arrhenius,1896) and hasbecome much better developed during the 20thcentury (e.g.,beginning with papers by Callendar,1938;Manabe and Wetherald,1975;Hansen et al.,1981 and continuing to include thousands of addi-tional scientific papers). Assessments of the scientif -ic literature to evaluate the basis for postulating thathuman activities are affecting the global climatehave been undertaken by many groups,includingthe Intergovernmental Panel on Climate Change(IPCC,1990,1992,1996a),eminent advisory groups(PSAC,1965;NRC,1979,1983; NAS,1992), govern-ment agencies (e.g.,USDOE,1985a,1985b),profes-sional societies (most recently, the AmericanGeophysical Union,see Ledley et al.,1999),andprominent scientific researchers (e.g.,Mahlman,1997). All of these analyses have come to similarconclusions,indicating that human activities arechanging atmospheric composition in ways that arevery likely to cause significant global warming dur-ing the 21st century. Results presented in this chap-

SCENARIOS FOR CLIMATE VARIABILITY ANDCHANGE

mate may seem hardly noticeable,the record of theEarth’s environmental history indicates that seem-ingly small changes in climate (e.g., changes in thelong-term average temperature of a few degrees)can have quite noticeable consequences for societyand the environment.

Many factors determine the Earth’s weather and cli-mate,including the intensity of solar radiation,con-centrations of atmospheric gases and particles,inter-actions with the oceans,and the changing characterof the land surface. The predominant source ofwarming is energy received from the Sun in theform of solar radiation. Energy from the Sun entersthe top of the atmosphere with an average intensityof about 342 watts per square meter. About 25% ofthis energy is immediately reflected back to spaceby clouds,aerosols (micron-sized particles anddroplets,including sulfate aerosols),and other gasesin the atmosphere;an additional 5% is reflectedback to space by the surface,making the overallreflectivity (or albedo) of the Earth about 30%. Ofthe other 70% of incoming solar radiation, about20% is absorbed in the atmosphere and the rest isabsorbed at the surface. Thus,70% of incomingsolar energy is the driving force for weather and cli-mate (Kiehl and Trenberth,1997).

Studies of the Earth’s climatic history extendingback hundreds of millions of years indicate thatthere have been global-scale climate changes associ-ated with changes in the factors that affect theEarth’s energy balance. Factors that have exertedimportant influences include changes in:solar irradi-ance,the Earth’s orbit about the Sun,the composi-tion of the atmosphere,the distribution of land andocean,the extent and type of vegetation,and thethickness and extent of snow and glaciers. Recordsof global glacial extent derived from ocean sedimentcores (e.g.,see Imbrie et al.,1992,1993) and of tem-perature and atmospheric composition derived fromdeep ice cores drilled in Greenland and Antarctica(e.g., Petit et al.,1999) provide strong indications ofthe interactions and associations of these variousinfluences. The Antarctic record (Figure 1), forexample,indicates that the atmospheric CO2 con-centration can be changed by up to 100 parts permillion by volume (ppmv)2 as a result of the climatechanges that occur due to the glacial-interglacialcycling over the past 420,000 years (Petit et al.,1999). While explanations of the relationshipsamong orbital forcing,atmospheric concentrationsof GHGs,and glacial extent are not yet fully quanti-fied,it is clear that the Earth’s climate has been dif-

comprehensive,state-of-the-science,model simula-tions to provide plausible scenarios for how thefuture climate may change;and (3) sensitivity analy-ses that can be used to explore the resilience ofsocietal and ecological systems to climatic fluctua-tions and change. This chapter provides backgroundand information concerning past and projectedchanges in climate needed to carry through theNational Assessment goal of analyzing potential con-sequences for society and the environment.

It should be emphasized that this chapter does notattempt a full scientific review of the adequacy oraccuracy of climate observations or climate simula-tions of the past or future. For such a review, thisAssessment relies on the very comprehensive,inter-national assessments being undertaken by the IPCC(e.g.,IPCC 1996a and the report now in preparationfor release in 2001). Rather, this chapter providesinformation needed to understand and explain theanalyses of the regional to national scale impactstudies that are described in this NationalAssessment report and the supporting regional andsector reports. In presenting the needed back-ground information,this chapter summarizes thestrengths and weaknesses of the various approachesthat need to be considered in interpreting theresults of the impact analyses. This considerationincludes balancing the many limitations that pre-clude making accurate specific predictions with theneed for providing the best available information forconducting a risk-based analysis of the potentialconsequences of climate change.

CLIMATE AND THEGREENHOUSE EFFECTThe ensemble of weather events at any locationdefines the climate in that place. The climate isdescribed by such measures as the averages of tem-perature,precipitation,and soil moisture as well asthe magnitude and frequency of their variations,thelikelihood of floods and droughts,the temperatureof the oceans,and the paths and intensities of thewinds and ocean currents. In contrast to climate’sfocus on average conditions over seasons to cen-turies and longer, weather describes what is happen-ing at a particular place and time (e.g.,when andwhere a thunderstorm occurs). Although the weath-er is constantly changing,the time- and space-aver-aged conditions making up the climate can also varyfrom season to season or decade to decade and canchange significantly over the course of decades orcenturies and beyond. While a slowly warming cli -

18


2 Parts per million by volume (ppmv) is equivalent to the number ofmolecules of CO2 to the number of molecules of air, which is made upmostly of nitrogen and oxygen.

ferent when atmospheric composition has been dif-ferent. Analyses indicate that these natural changesin atmospheric composition are being driven mainlyby the initial changes in climate due to the orbitalchanges,and are then acting as feedbacks thatamplify or moderate the initial changes in the cli-mate. Given the evidence that changes in atmos-pheric composition have been a factor in determin-ing climatic conditions over the Earth’s history,human-induced changes in atmospheric composi-tion (particularly greenhouse gas concentration)would also be expected to have an important influ-ence on the climate. Scientific understanding of thechanges in climates of the geological past would besignificantly compromised if the Earth’s climatewere not now responding to changes in atmospher-ic composition.

Changes in the Earth’s orbit around the Sun occurquite slowly, with periods ranging from about20,000 to 400,000 years (Berger, 1978;Berger andLoutre,1991). While these long periods mean thatchanges will be slow, their influences are steady andthe changes,along with other factors,seem to causetrends in temperature evident in records of a fewcenturies or more in length (Berger, 1999). On thetime scale of many centuries to millennia,observa-tions from Antarctic ice cores (Petit et al.,1999;Imbrie et al.,1989) suggests that these orbitalchanges cause changes in climate that lead tochanges in the amount of carbon dioxide in theatmosphere. These changes in the CO2 concentra-tion, working in parallel with the dynamics of icesheets and their underlying geological substrate,then seem likely to have reinforced the glacier-inducing and melting influences of the changes insolar radiation caused by the orbital variations(Pisias and Shackleton,1984;Shackleton et al.,1992;Petit et al.,1999;Clark et al.,1999). Following theend of the last glacial period about 10,000 yearsago,orbital changes appear to have contributed to a

Northern Hemisphere warming that peaked about6,000 years ago when the Earth was closer to theSun during the Northern Hemisphere summer.Subsequent to this peak,a slow and sometimesintermittent cooling of the Northern Hemispherestarted that seems to have continued until over-whelmed by the warming effects of the recentincreases in the CO2 concentration due to humanactivities (Thomson,1995).

The amount of solar radiation reaching a given loca-tion on the Earth can also be changed by changes insolar output (irradiance). Satellite observations ofsolar irradiance over the past 20 years indicate thatthe amount of energy put out by the Sun varies byabout 0.1% over the 11-year sunspot cycle,withmore energy coming out at sunspot maximum andless at solar minimum (Willson,1997). Analyses ofrecords of atmospheric conditions indicate thatstratospheric temperatures do vary somewhat with


400,000 Years of Antarctic CO2 and Temperature Change

Figure 1: Changes in the global average concentration of carbondioxide (light) and the local surface air temperature (dark) havebeen reconstructed for the past 420,000 years using informationderived from an ice core drilled at the Vostok station in Antarctica(Petit et al., 1999). The local temperature record is derived frommeasurements of oxygen-18 isotope concentrations in the waterfrozen as snow. The record shows a series of long-term variationsin the lower tropospheric (above the inversion layer) temperaturethat are similar to changes in solar radiation caused by changes inthe Earth’s orbit around the Sun. For most of past 420,000 years,temperatures in Antarctica (and by implication the globe) have beenlower than recent values. Independent geological evidence indi-cates that glacial ice amounts peaked on Northern Hemisphere con-tinents during these cold periods, most recently about 20,000 yearsago. The very brief warm periods coincide with interglacial periodsover the world’s continents, with the Eemian interglacial of about120,000 years ago being the last warm period until the presentinterglacial started about 10,000 years ago. In the absence ofhuman influences on the climate, models of the advance and retreatof glaciers that include representations of changes in the Earth’sorbit, natural variations in atmospheric composition, effects of cli-mate change on land cover, sinking and rising of land areas due tothe presence or absence of glaciers, and other factors suggest thatthe Earth would not return to glacial conditions for many thousandsof years (Berger et al., 1999). These studies also suggest that glob-al-scale glaciation would be unlikely if the CO2 concentration isabove about 400 ppmv.

CO2

Temperature

20


the sunspot cycle,but most scientists believe thatthese variations are too small to have caused adetectable impact on global average surface temper-atures,especially with the thermal buffering provid-ed by the global ocean. However, over the longer-term, reconstructions of changes in solar ir radiancesuggest that there may have been an increase of0.24% to 0.3% in solar output over the past severalcenturies (Lean et al.,1995;Hoyt and Schatten,1993). Calculations indicate that this increase insolar energy may have created a global warming ofas much as 0.4˚F (about 0.2˚C) from the 17th toearly 20th century, and perhaps contributed to asmall cooling influence since solar irradiancepeaked near the middle of this century (Lean andRind,1998). Over hundreds of millions of years,astronomical studies indicate that the amount ofsolar energy emitted by the Sun has been slowlyincreasing,but that these changes are too slow to beinducing noticeable climate change during humanexistence.

For global average temperatures to be relatively sta-ble over time,there must be a balance of incomingsolar energy and outgoing energy radiated away asheat (or infrared) energy. Observations from satel-lites confirm that the amount of outgoing energy isin close balance with the amount of absorbed solarenergy. However, the observations of the amount ofenergy being emitted are consistent with a celestialbody (like the Moon) that has an average tempera-ture close to 0˚F (about –18˚C). Were 0˚F really thesurface temperature,the Earth’s surface would becovered with snow and ice and it would be too coldfor life as we know it. Observations indicate,how-ever, that the Earth’s atmosphere acts to warm thesurface in a manner similar in effect (but different indetail) to the glass panels of a greenhouse. TheEarth’s natural “greenhouse”effect occurs becauseonly a small fraction of the infrared radiation emit-ted by the surface and lower atmosphere is able tomove directly out to space. Most of this heat radia-tion is absorbed by gases in the atmosphere andthen,along with other contributions of energy tothe atmosphere (e.g.,from absorption of solar ener-gy or heat released by the condensation of precipi-tation) is re-emitted,either out to space or backtoward the surface. Because the downward emittedenergy is available to further warm the surface,thisblanketing effect raises the average surface tempera-ture of the Earth to about 58˚F (about 14˚C) (Joneset al.,1999).

The gases that absorb and reemit infrared radiationare called greenhouse gases (GHGs). The set ofGHGs includes water vapor (the most important

greenhouse gas),carbon dioxide (the most impor-tant greenhouse gas whose concentration is beingdirectly influenced by human activities),methane,nitrous oxide, chlorofluorocarbons,stratosphericand tropospheric ozone,and others. Most of theGHGs occur naturally in the atmosphere,contribut-ing to the natural greenhouse effect that acts tokeep the Earth at a higher temperature than it other-wise would be were these gases not present.Observations and laboratory experiments indicatethat as the amount of these GHGs is increased,moreof the infrared radiation emitted upward from thesurface and lower atmosphere is absorbed beforebeing lost out to space. This process intensifies thenatural greenhouse effect,trapping more energynear the surface and causing the temperatures ofthe surface and atmosphere to rise (e.g.,see Goodyand Yung,1989).

Small particles or droplets (known collectively asaerosols) and changes in cloudiness and land reflec-tivity can affect how much energy is absorbed bythe Earth,creating a warming influence if the over-all reflectivity decreases,or a cooling influence ifoverall reflectivity increases. For example,aerosolscan result from major volcanic eruptions or burningof sulfur-laden coals or vegetation (e.g.,both naturaland human-induced fires). Cooling can result whenlight colored aerosols (such as sulfate aerosols orvolcanically injected aerosols) increase the amountof solar energy reflected back to space and therebydecrease the amount of energy absorbed by theatmosphere and surface. In addition to their directeffect,it is possible that sulfate aerosols exert anindirect cooling influence by increasing the reflec-tivity, extent,and character of clouds. By contrast,carbonaceous aerosols,such as organic compoundsand soot that are injected by fires and inefficientcombustion can increase solar absorption by theatmosphere,thereby creating a warming influenceby adding to the amount of energy that can be recy -cled by the greenhouse effect. Changes in the vege-tation cover can themselves affect the energy bal-ance, changing surface reflectivity, evapotranspira-tion rates,wind drag,and the amount by whichsnow cover can increase surface reflectivity in win-ter (Pitman et al.,1999). Unfortunately, the under-standing of these direct and indirect influences isquite limited,although they are not thought to bedominant (IPCC,1996a).

While the large-scale,long-term climate of the Earthas a whole is determined by the balance of incom-ing solar radiation and outgoing infrared radiation(moderated by the movement of energy within theEarth system),the climate at a particular place

21


in the 10,000 years prior to the start of human con-tributions suggests that the fluxes tended to be inbalance,with the amounts of carbon (or CO2) in anyparticular reservoir not changing significantly overtime.

Over the past few hundred years, evidence clearlyindicates that human activities have started tochange the balance. The lower curve in Figure 2provides the best available reconstruction of carbonemissions to the atmosphere (as CO2) since about1750 (Marland et al.,1999). Deforestation and thespread of intensive agriculture initiated a growth inemissions of CO2 in the mid-18th century that hasmoved about 130 GtC from the biosphere into theatmosphere (updated from Houghton,1995) sincethat time. Starting in the 19 th century and accelerat-ing in the 20th century, combustion of coal,oil,andnatural gas has led to emissions totaling more than270 GtC (extended from data presented in Andres etal.,2000). These fuels are collectively referred to asfossil fuels because they were formed many millionsof years ago from the fossil remains of plants andanimals. The effect of combustion of fossil fuels isto add carbon to the atmosphere that has been iso-lated in geological formations for many millions ofyears. Combustion of fossil fuels is currently addingmore than 6 GtC per year to the atmosphere.

As indicated in the middle curve of Figure 2,theatmospheric concentration of CO2 has beenresponding to these additions. The concentrationsshown here are derived from air bubbles trapped inice cores (Neftel et al.,1994) and since 1957 fromdirect measurements taken at the Mauna LoaObservatory in Hawaii (Keeling and Whorf, 1999;Conway et al.,1994). These observations,and oth-ers from around the world,provide convincing evi-dence that there has been an increase in the atmos-pheric CO2 concentration from historical levels ofabout 270-280 ppmv in the early 19th century toover 365 ppmv at present. Many types of studiesconfirm that it has been the rise in CO2 emissionsfrom land clearing and fossil fuel use that havecaused the rise in the atmospheric CO2 concentra-tion over the last 200 years (e.g.,Wigley andSchimel,2000).

Although the natural fluxes of carbon beingexchanged each year between the atmosphere andthe oceans and between the atmosphere and vegeta-tion are at least 10 times larger than the 6 GtC/yrfrom fossil fuel emissions,only about half of the fos-sil fuel carbon can be taken up by the vegetationand oceans. The other half of the atmosphericincrease, for reasons that relate to the slow over-

depends on interactions of the atmosphere,land sur-face (including its latitude,altitude,type,and vegeta-tive cover),and oceans. The atmosphere and oceanstransport energy from place to place,store it in theupper ocean,transform the form of energy fromheat to water vapor through evaporation and backthrough condensation,and create the climate expe-rienced at particular places. Some of the interac-tions are very rapid,as in the creation and move-ment of storms that have important local influences.Others,however, are quite slow, as in the severalyear cycle of El Niño (warm) and La Niña (cold)events in the tropical eastern and central PacificOcean that influence the weather around much ofthe world. Changes in land cover also causechanges in the amount of energy absorbed or emit-ted. Such changes can occur as a result of deforesta-tion, changes in snow cover, growth or decay of gla-ciers,or other factors. Thus, changes in the process-es that determine how energy is absorbed,movedaround,and stored cause the climate to fluctuate oreven change over long periods.

HUMAN ACTIVITIES ANDCHANGES IN ATMOSPHERICCOMPOSITIONObservations from the Vostok ice core record andother ice core records (e.g., Petit et al.,1999;Neftelet al.,1994) indicate that,until the last couple ofcenturies,the atmospheric CO2 concentration hadvaried between about 265 and 280 ppmv over thepast 10,000 years (Indermuehle et al.,1999). Eventhough the average atmospheric concentration var-ied over this time by only a few ppmv, exchanges ofcarbon were occurring among the atmosphere,oceans,and vegetation (each referred to as being areservoir for carbon,in that carbon comes in andgoes out over time). For example,carbon was beingtaken up by vegetation into living plants and beingreturned to the atmosphere as soil carbon decayed.Carbon dioxide was also being released into theatmosphere as cold,upwelling ocean waterswarmed in low latitudes,and CO2 was being takenup in the cold waters sinking in high latitudes.Estimates of the annual fluxes (transfers) of carbonbetween the atmosphere and ocean (and back),andthe atmosphere and vegetation (and back),suggestthat transfers of 60 to 90 billion metric tons of car-bon (abbreviated as GtC, for gigatonnes of carbon)per year have been taking place for each pathwayfor thousands of years (Schimel et al.,1995). Therelatively stable atmospheric concentration of CO2

22


turning rate of the oceans and limits on how muchvegetation can accumulate,is destined to remain inthe atmosphere for at least 100 years, even if globalemissions are substantially reduced. Just as addingwater to a multi-pool fountain raises its level eventhough the amount of water being pumped throughis many times larger than the amount of water beingadded,adding carbon (as CO2) from geological stor-age to the amount being exchanged among theatmosphere,ocean,and vegetation reservoirs causesa rise in the atmospheric concentration (as well asin ocean and vegetation levels).

HUMAN ACTIVITIES ANDCLIMATE CHANGEBased on scientific understanding of the greenhouseeffect,increasing the atmospheric composition ofgreenhouse gases should cause the global tempera-ture to rise. The top curve of Figure 2 presents areconstruction of the annual-average near surface airtemperature for the last 1000 years for the NorthernHemisphere (Mann et al.,1999);Crowley (2000)finds similar results. Because instrumental data aresparse or non-existent before the mid-19th century,these estimates of temperature are based on suchproxy indicators as widths of tree-rings,types ofvegetation,amounts of snowfall as recorded in icecores,etc. While these measures are not as preciseas thermometers,such indicators have proven to bereasonably accurate for reconstructing the fluctua-tions in Northern Hemisphere average temperature,providing a good indication of the variations thathave occurred prior to the start of instrumental datain the mid 19 th century. Although not as precise intheir time resolution, records of subsurface groundtemperatures also confirm that long-term warmingis occurring (Huang et al.,2000).

These proxy data suggest that for most of the past1000 years,the Northern Hemisphere average tem-perature had been slowly cooling at about–0.03˚C/century (Thomson,1995;Mann et al.,1999). Then,starting in the late 19th century, thetemperature started to rise,and has risen especiallysharply during the latter part of the 20th century.This 20th century warming appears to be unprece-dented compared to natural variations prior to thiscentury that were presumably caused by solar, vol-canic,and other natural influences. In addition,thecurrent warming is much more extensive andintense than the regional scale warming that peakedabout 1000 years ago in Europe during what isreferred to as the Medieval Warm Period (Mann et

a. Carbon Emissions

c. Temperature Change

1000 Years of Global CO2 and Temperature Change

1

0.8

0.6

0.4

0.2

0

-0.2

-0.4

-0.6

9

8

7

6

5

4

3

2

1

0

Figure 2: Records of CO2 emissions, CO2 concentrations, andNorthern Hemisphere average surface temperature for the past 1000years: (a) Reconstruction of past emissions of CO2 as a result ofland clearing and fossil fuel combustion since about 1750 (in billionsof metric tons of carbon per year) [data from CDIAC, 2000; Andres etal., 2000; Marland et al., 1999; Houghton, 1995; Houghton andHackler, 1995]; (b) Record of the CO2 concentration for the last 1000years, derived from measurements of CO 2 concentration in air bub-bles in the layered ice cores drilled in Antarctica, a location that hasbeen found to be representative of the global average concentration[data from Etheridge et al., 1998; Keeling and Whorf, 1999]; (c)Reconstruction of annual-average Northern Hemisphere surface airtemperatures based on paleoclimatic records (Mann et al., 1999). Forthe Mann et al. data, the zero change baseline is based on the aver-age conditions over the period 1902-80. The error bars for the esti-mate of the annual-average anomaly increase somewhat going backin time, with one standard deviation being about 0.25˚F (0.15˚C).Although this record comes mostly from the Northern Hemisphere, itis likely to be a good approximation to the global anomaly based oncomparisons of recent patterns of temperature fluctuations.See Color Plate Appendix.

b. CO2 Concentration380

360

340

320

300

280

260

Proxy Records

Instrumental Records

al.,1999;Crowley, 2000). The recent warming is alsofar more than can be characterized as a recoveryfrom the cool conditions centered in Europe and theNorth Atlantic region a few hundred years ago thatare often referred to as the Little Ice Age (Crowleyand North,1991;Mann et al.,1999;Crowley, 2000).Overall,looking back over the few thousand yearsfor which we can reconstruct estimates of large-scaletemperatures,the current warmth of global condi-tions appears unprecedented.

Figure 3 presents the instrumental records of tem-perature change for the globe and for the US. Theglobal results indicate that the annual average tem-perature has risen about 1.0˚F (about 0.6˚C) sincethe mid-19th century, with sharp rises early and latein the 20th century and a pause in the warming nearthe middle of the century. Sixteen of the 17warmest years this century have occurred since1980,and,counting the projected temperature for1999,the seven warmest years in the instrumentalrecord have all occurred in the 1990s. The globalaverage temperature in 1998 set a new record by awide margin, exceeding that of the previous recordyear, 1997, by about 0.3˚F (Karl et al.,2000). Higherlatitudes have warmed more than regions nearer theequator and nighttime temperatures have warmedmore than daytime. To the extent that available dataare globally representative,the 1990s are thewarmest decade in the last 1000 years (the periodfor which we have adequate data,see Mann et al.,1999). A recent report by the National ResearchCouncil (NRC,2000) confirms that,although satellite-measured temperatures of the lower atmospheresince that record began in 1979 are rising moreslowly than surface temperatures,the two measuresof the global climate have been rising at similar ratesover the four-decade long record of balloon measure-ments (Angell,2000). The NRC report also confirmsthat there is good reason to accept the evidence thatthe increase in the surface temperatures is real andhas become relatively rapid compared to the rates ofwarming earlier in the 20th century.

Of course,distributions of temperature changearound the world are more varied,with some regionswarming at a rate substantially greater than the glob-al average and others even experiencing a modestcooling. Observations derived from the UnitedStates Historical Climatology Network (USHCN) for1200 of the highest quality observing stations in theUS indicate that surface temperatures have increasedover the past century at near to the global averagerate. As is the case around the world,the largestobserved warming across the US has occurred inwinter. Note that it is generally not appropriate to

23


compare the spatial patterns of the satellite observa-tions with the spatial patterns of the surface temper-ature record because, for example,the atmosphereredistributes temperature anomalies,near surfaceinversions disconnect surface and atmospheric tem-perature changes,and forcings such as by volcaniceruptions and ozone changes have different effectson the surface and atmospheric temperature trends.However, other measures of climate change acrossthe US indicate that changes are indeed occurring.

Figure 3: (a) Global annual-average surface temperature andtemperature change for combined land and ocean regions forthe period 1900-1999 based on the method of Quayle et al.(1999); (b) US annual-average surface temperature and tempera-ture change for the period 1900-1999 using the USHCN data set(Easterling et al., 1996). See Color Plate Appendix.

Global 20th Century Temperature

U.S. 20th Century Temperature

Temperature Change (ºF)

Temperature (ºF)

24


An increasing number of studies indicates that thetime histories of greenhouse gas emissions,concen-trations,and surface temperature are closely relatedrather than just random correlations (IPCC 1996a;Tett et al.,1999). Each type of factor that could con-tribute to the observed warming of the climatewould have a distinctive character or “fingerprint”that can be searched for in the observations. Forexample,an increase in solar radiation would beexpected to warm both the lower and upper atmos -phere, yet the lower atmosphere has warmed whilethe upper atmosphere has cooled. Although there issome evidence that some of the warming in the firsthalf of the 20 th century may have been due to anincrease in the intensity of solar radiation,majorwarmings like that of the 20th century have not beenevident in the records of the past thousand years(and likely much longer),suggesting that an increasein solar radiation is unlikely to be the primary causeof the recent warming (IPCC,1996a).

It is also becoming more clear that the change is notdue to a diminution of the influence of major vol-canic eruptions,especially because the relativelyrecent El Chichón (in 1983) and Pinatubo (in 1991)eruptions injected very large amounts of aerosol intothe stratosphere and yet,although there was somecooling,global average temperatures remained wellabove temperatures following the Krakatoa eruptionin 1883 and major eruptions during the first decadeof the 20th century (see Figure 3). Third,were thewarming due mainly to a change in the coupling ofthe atmosphere and oceans, we would expect to seevariations of this size and rate in the past. However,such variations do not appear to have occurred,except perhaps as the world was emerging from thelast glacial period when large ice sheets were melt-ing. Lastly, the possibility of urban heating contami-nating the temperature record has been examined innumerous studies and in each case only about 0.1˚Cor less of the observed 0.6˚C warming over the 20th

century can be linked to urban contamination oftemperature records (Karl et al.,1988, Jones et al.1990,Easterling et al.,1997). Based on the inadequa-cy of natural factors to explain the recent change,the IPCC (1996a) concluded that

“the probability is very low that these correspon-dences [i.e.,the observed time history of the geo-graphical,seasonal and vertical patterns of atmos-pheric temperature change] could occur bychance as a result of natural internal variabilityonly. The vertical patterns of change [i.e.,withstratospheric cooling and tropospheric and sur-face warming] are also inconsistent with thoseexpected for solar and volcanic forcing.”

More recent studies are confirming these findings

(e.g.,see Hegerl et al.,1997;Barnett et al.,1999;Knutson et al.,1999).

Climatic changes due to factors being influenced byhuman activities also have characteristic finger-prints. Because greenhouse gases are essentiallytransparent to solar radiation, yet absorb infraredradiation,increasing the concentrations of green-house gases creates a warming influence at the sur-face and a cooling influence in the stratosphere(which is consistent with what has been occurring).Increases in sulfate aerosols that have occurred overthe 20th century as a result of sulfur dioxide emis-sions resulting from coal combustion would beexpected to have led to a surface cooling thatwould be greater in the Northern Hemisphere thanin the Southern Hemisphere and most dominant inthe mid-20th century. Depletion of stratosphericozone as a result of the emissions of chlorofluoro-carbons would be expected to have led to surfacewarming and cooling of the lower stratosphere.

Accounting for the effects of the increases in green-house gas and aerosol concentrations and thechanges in stratospheric ozone,the time and spacepatterns of temperature changes are consistent witha strong warming during the 20th century caused bythe changes in greenhouse gas concentrations and acooling influence due to aerosols that grew instrength in the middle of the 20th century. Based onthese diverse results,the IPCC (1996a) concludedthat “the balance of evidence suggests a discerniblehuman influence on global climate.” Since thatassessment,an increasing number of studies are pro-viding more quantitative information indicating thatthe 20th century warming is unlikely to be due tosolely to changes in solar radiation and is likely tobe a result of the increasing concentrations ofgreenhouse gases and aerosols,especially during thelatter half of the 20th century (Tett et al.,1999;Stottet al.2000).

APPROACHES FORASSESSING THE IMPACTSOF CLIMATE CHANGEBecause it would be very disruptive to rapidly termi-nate use of fossil fuels around the world3, it is clearthat the atmospheric CO2 concentration will contin-ue to increase for many decades into the future. Inaddition,the concentrations of other greenhousegases are increasing,and limitation of emissions ofthese gases would require implementing significant

emission control measures. Theoretical analyses,measurements in laboratory and field experiments,and knowledge of the processes determining thetemperatures of the Earth,Mars,and Venus all indi-cate that increasing the concentrations of GHGs inthe atmosphere will increase the natural greenhouseeffect,causing the world to warm. Given theweight of evidence provided by assessments of thepotential for climate change,prudent risk manage-ment demands an assessment of the potentialimpacts of climate changes that will occur over the21st century.

Early attempts to investigate the potential conse-quences of climate change often simply assumedthat climate would change by an arbitrary amount(e.g.,temperatures increase by 5˚F, or precipitationgoes up or down by 20%). For other studies (e.g.,USEPA,1989), results from model simulations with adoubled CO2 concentration were all that were avail-able. Such studies,however, could only be used toinvestigate the potential sensitivity of existing sys-tems to a different climate, rather than to explorehow such changes might evolve over time and in sodoing how these changes might spur natural andsocietal adaptations that could moderate the poten-tial consequences.

In this Assessment,our goal is to examine the conse-quences of time-dependent climatic change. Doingthis requires a two-step process. First,estimates ofhow the climate may change in the future must bedeveloped. The development of scenarios of climatechange that can be used in this effort is the primarysubject of this chapter. Second,estimates must bedeveloped of how the climate will affect the envi-ronment and society, and of how society mightrespond. These topics are the subject of subsequentchapters,but are coupled to this chapter in that thepotential impacts often depend on having certaintypes of information available about how the weath-er or climate will change. To assist in these analyses,this chapter summarizes our understanding about anumber of the particular climatic influences thatmay occur.

Three approaches have been used to develop theinformation base needed to evaluate the potentialconsequences of climate change on the US:

• Carefully checked historical data are being usedto examine the potential consequences of thecontinuation of past climatic trends and weatherand climate extremes in order to evaluate theconsequences of recurrences of the types of cli-mate fluctuations and variations that occurred inthe past (e.g.,the Dust Bowl period);

• Results from general circulation model simula-tions extending out to the year 2100 are beingused to generate plausible quantitative estimatesof the combined influences on climate of pro-jected changes in greenhouse gas and aerosolconcentrations;and

• Sensitivity analyses are being encouraged to facil-itate exploration of the limits of vulnerability(both strengths and weaknesses) for particularregions,sectors,societal activities,and ecosys-tems.

This strategy has several advantages in that it servesmultiple purposes and addresses several needs.These include:(a) providing a historical basis forassessing the significance of potential changes inthe climate;(b) providing a range of plausible futureclimatic conditions as a means of recognizing thelimitations and degree of uncertainty in the modelformulations or assumptions;(c) incorporating therange and character of natural variability for consid -eration, given its importance for human and naturalsystems;(d) providing opportunities to comparemodel simulations with observations in order toevaluate model capability;and (e) ensuring opportu-nities to include sensitivity analyses to explore theimplications of thresholds or limits in human andecosystem adaptability. While not all groups havebeen able to pursue all approaches fully, having avariety of approaches has helped broaden theapproach and served many of these purposes.

Thus,in this multi-pronged approach, climate mod-els provide the Assessment process with physicallyconsistent projections that are sufficiently plausibleand quantitative to investigate the potential impactsof climate change on water, health,ecosystems, foodproduction,and coastal areas,among other types ofconsequences. Use of the model projections is guid-ed by knowledge of the climate of the last centuryand sensitivity analyses,and experience with theweather and climate during the historical recordprovides a benchmark,a personal and national con-text for assessing the future.

25


3 Estimates are that it would take a reduction in cumulative emissionsof somewhat over 50% during the 21st century to stop any further risein CO2 concentration,with virtually no emissions allowed thereafter.Even stabilizing the atmospheric concentration at twice its preindustri-al level would require limiting average emissions over the 21st centuryto about 130% of current levels,as opposed to the projected tripling ofCO2 emissions by the year 2100 as projected by the mid-range emis-sions scenario (IPCC,1996a).

important,contamination of temperature and pre-cipitation observations. A similar high-quality dataset has been developed for Alaska,but the spatialrepresentativeness of these data sets is not as highdue to the sparsity of stations.

In addition to the monthly average station data,datasets of daily maximum and minimum temperatureand precipitation have been used to examine vari-ability and trends in climatic parameters. The DailyHistorical Climatology Network data set containsobservations for 187 high-quality stations in the con-tiguous US for the period 1910-1997 and observa-tions for 1000 stations in the contiguous US for theperiod 1948-1997. An additional data set(“Probabilities of Temperature Extremes in theU.S.A.”CD-ROM, available from NCDC) has beendeveloped that includes observations of daily maxi-mum and minimum temperatures for 300 stations inthe contiguous US,Alaska,and Hawaii for the period1948-1996. The software on this CD-ROM uses a sta-tistical model described in Karl and Knight (1997)to provide probability estimates of how dailyextreme temperatures and heat waves may changeunder various warming scenarios. This CD-ROMalso contains software to allow the user to examineprobabilities of extreme daily temperatures underthe observed climate and how they might changewith climate change.

An important additional data set for sensitivity stud-ies in examining ecosystem impacts has been pro-vided by the Vegetation-Ecosystem Modeling andAnalysis Project (VEMAP Members,1995;Kittel etal.,1995,1997). The VEMAP data set extends from1895-1993. This record was created by using statisti-cal models that could link data from long-term sta-tions to help fill in records at stations spanning onlypart of the period 1895-1993. The statistical meth-ods allowed information for missing periods to beinferred and provided a spatially and temporally uni-form data set for driving ecosystem and agriculturalmodels, for example. The VEMAP record is based onUSHCN stations plus USDA-Natural ResourcesConservation Service Sno-Tel stations for high eleva-tion precipitation. Altogether, the data set draws oninformation from about 8000 stations. The process-ing algorithm for deriving a high spatial resolutiondata set accounts for elevation and slope changes.The primary data set provides gridded monthly aver-age data for minimum and maximum temperature,precipitation,humidity (both relative and absolute)and solar radiation at a 0.5˚ x 0.5˚ latitude-longitudespacing (about 27 miles or 43 kilometers in longi-tude and 35 miles or 55 km in latitude). Becausesome ecosystem and agricultural models require

1. Use of Historical Records4

Records of how the climate has actually changedover the past century and over earlier times providean important context for evaluating the potentialconsequences of future changes in climate.Climatologists have used two types of data to identi-fy changes and variations in climate. The first con-sists of actual obser vations made over the 20th cen-tury of temperature,precipitation,and other weath-er-related variables that have been routinely meas-ured at thousands of locations across much of theglobe,including the US. These data have been sup-plemented over the past few decades by space-based measurements. Because the observing meth-ods,instruments,and station locations have changedover time, climatologists have used various methodsto assess and correct for the non-climate related fac-tors that can affect these data.The second type is“paleoclimate”data:physical,biological,and chemi-cal indicators recorded in rocks,ice,trees,and sedi-ments that can be used to infer past climate condi-tions. Examples include the width or density of treerings,ice cores containing air that has been trappedinside the ice for thousands and even hundreds ofthousands of years,sediment at the bottom of lakesand the ocean,and others. These data are calibratedagainst modern-day climate measurements,and indi-cate that,on a global scale, climate fluctuations havebeen at most several tenths of a degree F (fewtenths of a degree C) over the past several thousandyears,indicating a quite stable climate compared toconditions occurring over the past million years.

For the US,carefully documented data records existfor the 20th century that provide climate informa-tion for most of the inhabited areas of the country.Data from the United States Historical ClimatologyNetwork (USHCN),which has been developed froma carefully selected and processed set of observa-tions from the US Cooperative Observing Network,have been thoroughly quality controlled (Easterlinget al.,1996). The data set was developed and ismaintained by NOAA’s National Climatic DataCenter (NCDC). It contains monthly averaged maxi-mum,minimum,and mean temperature and totalprecipitation data for 1200 of the highest qualityobserving stations in the continental United Statesfor the period 1895 to 1997. These data have beencarefully screened for recording errors and,basedon well-defined procedures,adjusted for long-termvariability or trends that might be introduced bychanges in instrumentation,station location,urbanwarming,or other factors that can cause small,but

26


4The data sets described in this section are available athttp://www.nacc.usgcrp.gov/scenarios/

estimates of daily projections of these variables, astatistically based “weather-generator”technique hasbeen used to provide estimates of daily temperatureand precipitation for each grid location.

2. Use of Climate ModelSimulations5

As a second approach,physically consistent projec-tions of future climatic conditions derived from cli-mate models provide an important tool for investi-gating the potential consequences of climatechange. Climate models have been developed andare used because the Earth’s atmosphere/ocean/land/ice system is far too complex to reproduce in alaboratory and simple extrapolations of pastchanges in climate cannot account for the rapidchanges in human influences on the climate. Thesemathematical representations of the Earth atmos-phere/ocean/land/ice system rely on the well-estab-lished laws for conservation of mass,momentum,and energy, and on empirical relationships derivedfrom observations of how particular processeswork,to specify transfers of these conserved quanti-ties among latitude/longitude/altitude grid boxesthat cover the Earth like tiles. The typical size of thegrid boxes that cover the Earth in current atmos-pheric models is about the size of a modest sized USstate and these boxes average several thousand feet(about a kilometer) thick;ocean models tend tohave finer grid sizes to represent the smaller oceaneddies.

Developing models that can be used to project pos-sible future climatic conditions requires incorporat-ing the most important physical principles andprocesses that determine climatic conditions. Themost comprehensive models of Earth’s climate sys -tem to date are called General Circulation Models orGCMs6 (e.g.,see Nihoul,1985;Washington andParkinson,1986;Mote and O’Neil,2000). Thedomains for these models include the global atmos-phere (up to mid-stratospheric altitudes),the oceans(from surface to the bottom),the land surface(although with limited detail in mountainousregions),and sea ice and snow cover (withGreenland and Antarctic ice caps assumed to bepresent). The processes represented include solar

and infrared radiation,transfer and transformation ofenergy, evaporation and precipitation,winds andocean currents,snow cover and sea ice,and muchmore. While full detail cannot always be included,present models are constructed so as to representkey processes with suf ficient detail that the large-scale climate and its sensitivity to potential changesby human activities can be self-consistently calculat-ed. Tests are performed to determine the ability ofthe models to simulate the evolution of tempera-ture, rainfall,snow cover, winds,soil moisture,seaice,ocean circulation,and other key variables overthe entire globe through the seasons and over peri-ods of decades to centuries (e.g.,Gates et al.,1999;Meehl et al.,2000a).

The advantages of using model simulations are thatthey are quantitative and are based on the funda-mental laws of physics and chemistry, often affectedand moderated by biological interactions. However,while attempts are made to ensure climate modelsare adequately comprehensive,such models areobviously simplified versions of the real Earth that,in their current versions,cannot capture its fullcomplexity, especially at regional and smaller scales.The level of confidence that can be placed in suchmodels can be evaluated by testing their ability tosimulate past and present climate conditions.Among the tests that have been used to evaluate theskill of climate models have been comparisons ofmodel simulations of the weather (to the limit thatit is predictable),the cycle of the seasons, climaticvariations over the past 20 years when globally com-plete data sets are available, climatic changes overthe past 150 years during which the world haswarmed,and climatic conditions for periods in thegeological past when the climate was quite differentthan at present. Studies on comparisons of modelsimulations of paleoclimatic variations also suggestthat models can simulate some of the types ofchanges that have been reconstructed from the geo-logical records (e.g.,COHMAP, 1988; Kutzbach et al.,1993; Joussaume et al.,1999). Beyond studies ofparticular periods,only quite simplified models havebeen able to be tested on their simulations of theonset,duration,and termination of the glacial peri-ods of the past million years,and these results sug-gest that the GCMs likely do not adequately includeall of the feedback processes that may be importantin determining the long-term climate (Berger, 1999;Berger et al.,1999).

The capabilities of the most developed of thesemodels have been carefully reviewed by the IPCCand as part of other national and international scien-tific efforts to evaluate their ability to represent

27


5Results from the models described in this section are available athttp://www.nacc.usgcrp.gov/scenarios/6 Some studies refer to these models of the global climate system asGlobal Climate Models (also condensed to GCMs). However, technical-ly, it is only the atmospheric and oceanic components of such modelsthat are actually considered to be General Circulation Models in thatthey calculate how air and oceans move. Because the atmospheric andoceanic parts of global climate models are so dominant and so widelydiscussed, we have chosen to refer to the overall climate system mod-els as General Circulation Models.

the future. This choice is meant to emphasize thatwe must recognize that climate model simulationsdo not provide precise forecasts,but rather are bestused to develop insights about plausible climatechanges resulting from specific assumptions such asabout how energy technologies and emissions willevolve. Relying on this approach, even with therecognized uncertainties,can be useful,just as it isin other cases where individuals and organizationsmake use of information, even if it is associatedwith some level of uncertainty. For example,manypeople plan their days around weather forecastswith uncertainty conveyed both in words and num-bers,e.g.,a 30% chance of rain,or snow likely witha probability of 70%,etc. Others invest financialresources based on economic trends or decide topurchase a new home based on interest rate analy-ses. Understood in this light,the model-based sce-narios can help to provide useful insights about theconsequences of climate variability on the US,butthe model results should be considered as plausibleprojections rather than specific predictions.

3. Use of Vulnerability Analyses

The third approach to exploring potential impactsof future climate change is to ask what degree ofchange would cause significant impacts in areas ofcritical human concern,and then to seek to deter-mine the likelihood that such changes might occur(based on the historical record,model simulations,etc.). This approach is a form of “sensitivity analy-sis”conducted to determine under what conditionsand to what degree a system might be sensitive tochange. Such analyses are not predictions that suchchanges will occur; rather, they examine what theimplications would be if the specified changes didoccur.

For example,questions that might be exploredcould include:What would happen to weather con-ditions over the US if El Niño conditions occurredmore frequently or more intensely? What would bethe implications if there were simultaneousdroughts in the US and in other grain-growingregions? What if the 1980s California drought last-ed ten years instead of six? What if the deepwatercirculation of the North Atlantic Ocean were dis-rupted and colder conditions prevailed from NewEngland across to Europe? Alternatively, such ques-tions could be phrased:How large would climatechange have to be in order to cause a particularimpact? How dry would conditions need to be forfire frequency and extent to increase significantly inthe southeastern US? How high do ocean tempera-tures have to become for coral reefs to be seriously

most aspects of the present and historical climates(e.g.,see discussions in IPCC,1996a;Gates et al.,1999;Meehl et al.,2000a). These evaluations indi-cate that climate models represent many, but not all,of the important large-scale aspects of the global cli -mate quite well. The evaluations also show, howev-er, that there are important limitations of their simu-lations of regional conditions,particularly in anddownwind of mountainous regions,because impor-tant local influences are not well represented in themodels. Model capabilities for representing naturalclimate variations over periods of years (e.g.,the ElNiño/La Niña fluctuations) to several decades (e.g.,over the Pacific and Atlantic oceans) are only begin-ning to show success. Basically, the model evalua-tions indicate that the models can be used to pro-vide important and useful information about poten-tial long-term climate changes over periods of up toa few centuries on hemispheric scales and acrossthe US,but care must be taken in interpretingregionally specific and short-term aspects of themodel simulations. Rather than repeat the full analy-sis of model results being undertaken as part of theongoing IPCC assessments,this chapter focuses onthe performance of the two selected models overthe US,while the Assessment as a whole focuses ondetermining how the selected climate change sce-narios may impact human and natural systems.

Because these models are based on quantitative,physically based relationships governing,to theextent of current understanding,the global distribu-tions of air pressure,heat,moisture,and momentum,climate models can be used to investigate how achange in greenhouse gas concentrations,or a vol-canic eruption,may modify the Earth’s climate.Using models in this way enables the generation ofinformation that can potentially be used in assess-ment of impacts across the regions and sectors ofthe country. Because of continual ef forts atimprovement over the last several decades,thesemodels provide a state-of-the-science glimpse intothe climate of the 21st century and represent agrowing capability to learn how climate change mayimpact the nation. However, real uncertaintyremains in the ability of models to simulate manyaspects of the future climate such that the modelresults must be viewed as providing a view of futureclimate that is physically consistent and plausible,but incomplete.

To convey the importance of the limitations,assumptions,and uncertainties in the model results,the IPCC has adopted the terms “projection”and“scenario” rather than “prediction”or “forecast”torefer to the results of climate model simulations of

28


threatened? How low does river flow in theMississippi-Missouri basin have to become for exten-sive areas of hypoxia (lack of oxygen) to occur inthe Gulf of Mexico?

While there are always values for which one couldget disastrous consequences,this approach is mostuseful when it focuses on basing the questions ontypes of climatic fluctuations, changes, or conditionsthat might have occurred before the instrumentalrecord began. For example,a recent study byWoodhouse and Overpeck (1998) suggests that the1930s drought in the Great Plains,while severe, wasmuch shorter than earlier droughts that haveoccurred in the past several hundred years. Theyalso found that droughts of similar magnitude to the1930s drought are expected to occur about once ortwice a century. Thus,a return of the 1930sdrought,perhaps even lengthened,seems a plausiblescenario for the future (Stahle et al.,2000 reportsimilar findings). Similarly, various proxy recordsindicate that droughts in California have lastedmuch longer than the 1980s drought. The fact thatsuch conditions have occurred suggests that theycould occur again,and that it would be prudent tothink about the impacts such climate fluctuationsmight have, given the way society has developed.

Because generating scenarios for sensitivity analysesnecessarily focuses on considering particular condi-tions in particular places inducing particular typesof impacts,the details of this approach are notdeveloped in this chapter. Instead,the region andsector chapters pose the questions and contain theinformation underpinning these analyses and theirapplication.This chapter is instead devoted to build-ing the base of national-scale information that thesestudies have used.

TRENDS IN CLIMATE OVERTHE US DURING THE 20th

CENTURYThe climate of the United States contains an incredi-ble variety of climatic types. It ranges from the highlatitude Arctic climate found in northern Alaska,totropical climates in Hawaii,the Pacific Islands andCaribbean,with just about every climate regime inbetween. Because of this wide array of climate,andthe large area involved,the interannual variations(year-to-year variability) of climate in different partsof the country are affected differently by a variety ofexternal forcing factors. Perhaps the most well-known of these factors is the El Niño-Southern

Oscillation (ENSO) which has an irregular period ofabout 2-7 years. ENSO has reasonably well-knowneffects in different parts of the country. In the ElNiño phase,which involves unusually high sea sur-face temperatures (SSTs) in the eastern and centralequatorial Pacific from the coast of Peru westwardto near the international date line,effects includemore winter-time precipitation in the southwesternand southeastern US,and above average tempera-tures in the Midwest that,with a strong El Niño,canextend into the northern Great Plains. The La Niñaphase,which involves unusually low SSTs off thewest coast of South America,often leads to higherwinter-time temperatures in the southern half of theUS,with more hurricanes in the Atlantic and moretornadoes in the Ohio and Tennessee valleys (Boveet al.,1998;Bove,personal communication).Furthermore,in the summertime,La Niña conditionsmay contribute to the occurrence of drought in theeastern half of the country (Trenberth andBranstator, 1992).

Other factors that af fect the interannual variabilityof the US climate include the Pacific DecadalOscillation (PDO),and the North Atlantic Oscillation(NAO). The PDO is a phenomenon similar to ENSO,but is manifest in the SSTs of the North PacificOcean (Mantua et al.,1997). The PDO has an irregu-lar period that is on the order of decades,and likeENSO, has two distinct phases,a warm phase and acool phase. In the warm phase,SSTs are higherthan normal in the equatorial Pacific,and lower thannormal in the northern Pacific,leading to a deepen-ing of the Aleutian Low, higher winter temperaturesin the Pacific Northwest,and relatively high SSTsalong the Pacific coast. This condition also leads todry winters in the Pacific Northwest,and wetterconditions both north and south of there.Essentially, the opposite conditions occur in thecool phase. The NAO is a phenomenon that displaysa seesaw in temperatures and atmospheric pressurebetween Greenland and northern Europe. However,the NAO also includes ef fects in the US such thatwhen Greenland is warmer than normal,the easternUS is usually colder, particularly in winter, and viceversa (Van Loon and Rogers,1978).

As context for evaluating the importance of climatechange during the 21 st century, it is useful to reviewhow the climate over the US has changed over the20th century. Whereas Figure 3b showed the resultsfor the US as a whole, Figure 4a displays the spatialpattern of the trend in annual average temperatureacross the US for the past 100 years calculated usingthe USHCN data set. Over most areas of the US,except for the Southeast,there has been warming of

29


strong. The Southeast is one of thehandful of places in the world indi-cating some cooling,due perhaps tothe increased presence of sulfateaerosols, changes in atmosphericcirculation regimes,and/or changesin cloud cover (Karl et al.,1996).Locally in some areas,interannualvariability is high enough andtrends small enough that sometrends are not statistically signifi-cant. However, wherever theabsence of statistically significantresults occurs,significant trends arefound nearby, reinforcing the overallobserved pattern of warming forthe US.

Not only are average temperatureschanging,but the variability of the global climatealso seems to be changing. For example, Parker etal.(1994) compared spatially averaged variances ofannual temperature anomalies between the twoperiods 1954-1973 and 1974-1993. An increase intemperature variability of between 4 and 11% wasfound for the latter period. In some areas,such asNorth America,the increase was even larger.However, Karl et al.(1995a) analyzed changes invariability over the 20 th century on a variety of timescales,from 1-day to 1-year for most of the NorthernHemisphere. They found evidence of a decrease invariability on shorter time scales (e.g.,1-day),but nobroad scale patterns for longer scale variability.Thus,it appears that, for example,temperature vari-ability on longer time scales (e.g., year-to-year vari-ability) is increasing,but variability on shorter timescales (e.g.,day-to-day or month-to-month variabili-ty) is decreasing.

Recent analysis of changes in the number of dayswhere the minimum temperature drops belowfreezing indicates that the frequency of such condi-tions is changing across the US. Over the 20th cen-tury, averaged over the country, there has been adecline of about two days per year (i.e.,-2 days/100years). The spatial pattern of the change mirrors thechanges in average annual temperature,showingcooling in the southeastern US and warming every-where else. Thus,the Southeast has experienced anincrease in the number of days below freezing whilethe western portion of the country has experiencedstrong decreases,with moderate declines or nochange elsewhere. Seasonally, this change is mostapparent in winter and spring,with little change inthe autumn. Examination of changes in the dates ofthe first autumn frost and the last spring frost shows

more than 1˚F, which is consistent with theobserved warming of the world as a whole (Karl etal.,1996). In some regions,particularly in theNortheast,the Southwest,and the upper Midwest,the warming has been greater, in some places suchas the northern Great Plains, reaching as much as3˚F. Warming in interior Alaska has also been quite30


Trends in Annual Average of Selected Climate Variables

Observed 20th Century - TEMPERATURE

Observed 20th Century - PRECIPITATION

Observed 20th Century - PDSI

Figure 4: Trends in the annual average of selected climatic vari-ables over the US during the 20th century as derived from observa -tions compiled in the USHCN data set (Easterling et al., 1996). (a)Temperature (˚F/century); (b) Precipitation (percent change/centu-ry); (c) Palmer Drought Severity Index (percent change/century).See Color Plate Appendix.

>15ºF

+10ºF

+5ºF

0ºF

-5ºF

>10

8

6

4

2

0

-2

-4

-6

-8

-10

Figure 5: US trends (1910-1996) in mean precipitation (in percentchange per century) for various categories of daily precipitationintensity. Values are plotted for each 5%, such that 5 representsfrom the lowest to 5 th percentile and 95 represents the 95 th to high-est values of precipitation intensity. The lowest to 5th percentileare the lightest daily precipitation amounts and the 95th to highestare the heaviest daily amounts (Karl and Knight, 1998).

Observed US Trends in Daily Precipitation Intensity

>100%

75%

50%

25%

0

-25%

-50%

-75%

-100%

a similar pattern,with little change in autumn,but achange to earlier dates of the last spring frost. Thisshift has resulted in a lengthening of the frost-freeseason over the country with a trend of 1.1 days perdecade (Easterling,2000).

Observations indicate that total annual precipitationis increasing for both the globe and over the US.Although global precipitation has only increased byabout 1%,the increase north of 30˚N has been sig-nificantly larger, estimated to be 7-12% (IPCC,1996a). For the conterminous US,the increase inprecipitation during the 20 th century is estimated tobe 5-10% (Karl and Knight,1998),which is broadlyconsistent with the global-scale changes in mid-lati-tudes. Although there is more spatial variationacross the US for precipitation trends than for tem-perature trends,and although the high year-to-yearvariability means that small changes are not as likelyto be statistically significant,there is an overallincreasing trend that is highly significant,both statis-tically and practically (see Figure 4b). Across theUS,most regions have experienced increased pre-cipitation with the exception of localized decreasesin the upper Great Plains,the Rocky Mountains,andparts of Alaska. Recent analyses suggest that muchof this increase in precipitation is due to increasesin heavier precipitation events (see Figure 5) and anincrease in the number of rain-days (Karl andKnight,1998). Not only is this trend evident in daily(24-hour) precipitation events,but the frequency ofheavy multi-day (7-day) precipitation events is alsoincreasing (Kunkel et al.,1999). Trends in additionaltypes of variability and extreme events are onlystarting to become available (Smith,1999;Easterlinget al.,2000a).

Soil moisture is a function of how much precipita-tion falls and when,as well as how much evapora-tion and runoff occur. Figure 4c shows the trendsin soil moisture across the US during the 20th centu-ry, calculated using a Palmer Drought Severity Indexmodel (Palmer, 1965). Overall,there has been rela-tively little change, except for some areas of theRocky Mountains and northern Great Plains thathave become somewhat drier, and for theMississippi River Valley, which,the way this index iscalculated,tends to show a recovery from thedrought years of the 1930s.

CLIMATE MODELSIMULATIONS USED IN THENATIONAL ASSESSMENTOver the past decade,models have been developedthat can quite reasonably simulate the climatic con-ditions of the 20th century and that can be used tosimulate the climatic effects of changes in atmos-pheric composition in the 21st century. These mod-els offer simulations of time-dependent scenarios,based on quantitative relationships, grounded inobservational evidence and theoretical understand-ing. Around the world,there are more than twodozen groups that are developing models to simu-late the climate (Gates et al.,1999;Meehl et al.,2000a). However, the various models are in variousstages of development and validation,and theirtreatments of greenhouse gases,aerosols,and othernatural and human-induced forcings continue toevolve. The various models that have been used tosimulate the climate of the 21st century have beenused in various types of simulations,including equi-librium and time-dependent simulations (IPCC,1996a). The most important characteristics of themodels that were considered as possible choices foruse in the National Assessment are summarized inTable 1.

For the purposes of the National Assessment,toensure use of up-to-date results,and to promote ahelpful degree of consistency across the broad num-ber of research teams participating in this activity,the National Assessment Synthesis Team (NAST)developed a set of guidelines to aid in narrowingthe set of simulations to be considered for use bythe regional and sector teams. To build the basis forits set of guidelines,the NAST developed a set ofobjectives for the characteristics of model simula-tions that would be most desirable. The criteria formaking the selections,which included aspects con-cerning the structure of the model,the character ofthe simulations,and the availability of the neededresults,included that the models must,to the great-est extent possible:

• be coupled atmosphere-ocean general circulationmodels that include comprehensive representa-tions of the atmosphere,oceans,and land sur-face,and the key feedbacks affecting the simula-tion of climate and climate change;

• simulate the evolution of the climate throughtime from at least as early as the start of thedetailed historical record in 1900 to at least as faras into the future as the year 2100 based on a 31


mum and maximum temperature and to be ableto represent the development of summertimeconvective rainfall;

• be capable,to the extent possible,of represent-ing significant aspects of climate variations suchas the El Niño-Southern Oscillation cycle;

• have completed their simulations in time to beprocessed for use in impact models and to beused in analyses by groups participating in theNational Assessment;

well-documented scenario for changes in atmos-pheric composition that takes into account time-dependent changes in greenhouse gas andaerosol concentrations (equilibrium simulationsassuming a CO2 doubling were excluded) 7;

• provide the highest practicable spatial and tem-poral resolution (roughly 200 miles [about 300km] in longitude and 175 to 300 miles [about275 to 425 km] in latitude over the central US);

• include the diurnal cycle of solar radiation inorder to provide estimates of changes in mini-

32


Model Component or

Feature

Atmospheric reso-lution in horizontal(latitude-longitude)

and vertical

Treatment of landsurface, evaporationand evapotranspira-

tion

Includes diurnalcycle

Oceanic resolutionin horizontal (lati -

tude-longitude) andvertical

Treatment of sea ice

Treatment of atmos-phere-ocean

coupling

Treatment of multi-ple greenhouse

gases

Treatment of sul-fate chemistry

Equilibrium tem-perature responseof system model to

CO2 doubling

Year when resultsfrom 1900 to 2100

simulation weremade available

Characteristics of ClimateModels Recommended for Use

in the National Assessment

CanadianClimateCentre

(CGCM1)

3.75˚ by 3.75˚ (spectral T32)

10 layers

Modifiedbucket for soil

moisture

Yes

1.8˚ by 1.8˚29 layers

(based on GFDLMOM 1.1)

Thermo-dynamic only

Flux-adjusted

No,CO2 usedas surrogate

Albedo changeonly

3.5˚C

6.3˚F

1998

Hadley Centre,United

Kingdom(HadCM2)

2.5˚ by 3.75˚(grid)

19 layers

Soil layers,plant canopy,and leaf stom-atal resistance

included

Yes

2.5˚ by 3.75˚

20 layers

Dynamic and thermo-dynamic

Flux-adjusted


Albedo changeonly

2.6˚C (4.1˚Cfor AGCM withsimple ocean)

4.7˚F (7.4˚F)

1998

Characteristics of Climate Models for which Some Results

Were Available for the National Assessment

Max PlanckInstitute,Germany

(ECHAM4/OPYC3)

2.8˚ by 2.8˚ (spectral T42)

19 layers


included

Yes

2.8˚ by 2.8˚

9 layers


Flux-adjusted


Albedo changeonly

2.6˚C

4.7˚F

1998 (but onlythrough 2049)

GeophysicalFluid

DynamicsLaboratory

(GFDL)

3.75˚ by 2.25˚ (spectral R30)

14 layers

Simplifiedbucket for soil

moisture

No

1.875˚ by2.25˚

18 layers(GFDL MOM 1.1)


Flux-adjusted


Albedo changeonly

3.4˚C

6.1˚F

1999

NationalCenter for

AtmosphericResearch

(NCAR CSM)


18 layers


included

Yes

2.4˚ by 1.2˚(variable)

45 layers

Dynamic and thermo

-dynamic

Not flux-adjusted

Yes

Yes,withreduced sulfur

emissions

2.0˚C

3.6˚F

1999

ParallelClimate Model

(PCM)


18 layers


included

Yes

0.66˚ by 0.66˚(variable)

32 layers


Not flux-adjusted

Yes

Sulfate loadingspecified from

NCAR CSM

2.0˚C

3.6˚F

1999

Hadley Centre,United

Kingdom(HadCM3)

2.5˚ by 3.75˚(grid)

19 layers

Soil layers,plant canopy,

stomatal resist-ance,and CO2

processesincluded

Yes

1.25˚ by 1.25˚

20 layers

Dynamic and thermo-dynami

Not flux-adjusted

Yes

Yes

3.3˚C

5.9˚F

2000

Table 1: Characteristics of Global Models

7 Note that although vegetation is an important feature of the land surface that can affect the climate,human-induced changes in future vegetationcover and changes in vegetation due to changes in climate are not yet being treated in these climate models.

• be models that are well-documented and whosegroups are participating in the development ofthe Third Assessment Report of the Intergovern-mental Panel on Climate Change (IPCC) in orderto ensure comparability between the US ef fortsand those of the international community;

• provide a capability for interfacing their resultswith higher-resolution regional modeling studies(e.g.,mesoscale modeling studies using resolu-tions finer by a factor of 5 to 10);and

• allow for a comprehensive array of their resultsto be provided openly over the World Wide Web.

Including at least the 20th century in the simulationadds the value of comparisons between the modelresults and the historical record and can be used tohelp initialize the deep ocean to the correct valuesfor the present-day period. Having results frommodels with specific features,such as simulation of

the daily cycle of temperature,which is essential foruse in cutting edge ecosystem models, was impor-tant for a number of applications that Assessmentteams were planning. Despite uncertainties sur-rounding available emissions scenarios,using resultswith consistent assumptions about increases ingreenhouse gases and sulfate aerosols helps toensure that the assessment ef forts of the variousregional and sector teams can be combined into aconsistent national synthesis and could then beinterfaced with international assessments.

These restrictions led to a decision to considermainly model simulations that used emissions sce-narios that were close to the IPCC’s “IS92a”scenario(see IPCC,1992) (see box,“What Does the IS92aScenario Assume?”) so that there could be readycomparison with international studies and analyses.As shown in Figure 6,the net radiative forcing for

33


What Does the “IS92a” Scenario Assume?

To prepare a projection of future changes in climate,a scenario of future concentrations of greenhouse gasesmust be developed. This is often done by starting with a scenario for changes in emissions of greenhousegases. The future emissions scenario most used for analysis throughout the 1990s,including to drive modelsimulations of climate change,has been the IS92a scenario. This scenario is near the middle of the range ofsix peer-reviewed scenarios of possible alternative futures published by the IPCC in 1992 (IPCC,1992). Basedon calculations done with models of greenhouse gas and aerosol concentrations,the IS92a scenario results ina climate forcing that is similar to that used in the two models chosen for primary use by the NationalAssessment. The recently published set of IPCC 2000 emissions scenarios finds that the net radiative forcingof this emissions scenario (i.e., greenhouse gas induced warming minus aerosol induced cooling influence) isstill well within the range of what the IPCC has recently concluded are plausible scenarios for how energytechnologies,energy use,economic development,and population growth of the 21st century may evolve(IPCC,2000).

The IS92a scenario makes a number of assumptions based on current and projected trends and expectations.Like each of the IPCC’s 1992 scenarios,it assumes that the nations of the world will implement no majorchanges in their policies that would limit the growth of activities that are contributing to climate change. Thescenario also assumes that global population will approximately double over the 21 st century and that contin-ued economic growth at rates typical of the recent past will raise total economic output by a factor of about10; growth by this amount would mean that global average per capita economic activity would go up by afactor of 5. Because of increasing efficiencies and new technologies,the scenario assumes world energygrowth will,however, only need to increase by about a factor of 4. To meet this increase in energy demand,energy derived from fossil fuels (coal,oil,and natural gas) is projected to more than double (increased use ofcoal,however, would increase CO2 emissions by a factor of about 3). To provide the rest of the energy, theIS92a scenario assumes that energy derived from non-fossil fuel energy sources (e.g.,solar, wind,biomass,hydroelectric,and nuclear) will increase by a factor of about 15. The scenario assumes that this growth innon-fossil energy sources will occur without any implementation of climate-specific policies because thecosts of these energy sources will decline relative to fossil fuels. If this scenario comes to pass,it would meanthat the fraction of energy coming from non-fossil sources would rise from just over 10% of all energy now toover 40% by 2100. Scenarios forecasting less rapid availability of non-fossil technologies would lead to greaterCO2 emissions to meet the same growth in population and economic activity;scenarios leading to reducedCO2 emissions would require some combination of more rapid increases in efficiency improvements, fasterdevelopment of non-fossil technologies,a slower rate of economic development,and reduced populationgrowth.

narios prepared by the IPCC (2000). Thenew scenarios suggest that the upper limitof possible increases in radiative forcing by2100 is greater than for the IS92 emissionsscenarios due to the recent recognition thatsignificantly intensified use of fossil fuelscould lead to substantial increases in emis-sions of methane,carbon monoxide,nitro-gen oxides,and volatile organic compoundsthat would significantly increase the con-centration of tropospheric ozone,a stronggreenhouse gas. What is clear from this dia-gram,and is discussed more fully later in thetext for the particular models used,is thatthe IS92a emissions scenario is a quite plau-sible choice for consideration if the resultsfrom only one emissions scenario are avail-able. However, it must be emphasized thatthe climate model results that are availableare simply one representation of what couldhappen,and are not predictions or forecasts

of what might actually happen. This restrictioncould start to be relaxed in future assessments byconsidering results from a wider range of climatemodels and a wider range of emissions scenarios.

In the selection of the particular set of model resultsto be used for the Assessment,a number of addition-al constraints were also considered. For example,time and computer resource constraints generallyprevented the completion of a new set of modelsimulations with these models specifically designedfor this Assessment. Given the limited duration ofthe Assessment,and the desire to process the GCMresults through the VEMAP processing package inorder to better account for changes in mountainousregions,it was essential that scenarios be completedearly in the assessment process (i.e.,mid to late1998) in order to enable timely availability ofprocessed model results. In addition,the limitationsin capabilities and resources have meant that the setof cases and situations that all teams would beasked to use needed to be kept to a minimum. Forthese reasons,it was necessary to limit the selectionto a minimum,but representative,set of model simu-lations.

Given these guidelines and considerations,theresults from particular simulations of two modelswere selected to be the primary sources of simula-tion-based projections for this first NationalAssessment. The specific simulations selected werethose runs that are closest to the IS92a emissionsscenario from the GCMs developed by the CanadianCentre for Climate Modelling and Analysis (hence-forth referred to as the “Canadian model scenario”)

the 21st century for the IS92a emissions scenario isnear the mid-range of radiative forcing scenariosconstructed based on the new set of emissions sce-

34


8

7

6

5

4

3

2

1

0

600 1000

Cumulative Carbon Emissions during 21st Century (GtC)

Radiative Forcing Versus 21st Century Carbon Emissions forVarious Emissions Scenarios

1600 2000 2600

IPCC 2000 Emissions Scenarios for Year 2100IPCC 1992 Emissions Scenarios for Year 2100Hadley Centre Model Emissions Scenario for Year 2100Canadian Centre Model Emissions Scenario for Year 2100

Figure 6: Comparison of the projections of total carbon emissionsand overall human-induced radiative forcing for the six emissionsscenarios prepared by the IPCC in 1992 (IS92 scenarios; IPCC,1992) and the 35 emissions scenarios prepared by the IPCC in 2000for which radiative forcing could be estimated (SRES scenarios;IPCC, 2000). These scenarios are based, although in different ways,on projected changes in emissions resulting from changes in popu-lation, economic development, energy use, efficiency of energy use,the mix of energy technologies, etc. The horizontal axis gives thetotal emissions of fossil fuel-derived carbon dioxide projected forthe 21st century (in billions of tonnes of carbon, GtC). For refer-ence, if the current level of global carbon emissions is maintainedfrom 2000 to 2100, cumulative emissions over the 21st centurywould be roughly 650 GtC. Assuming no climate-related controlson emissions are introduced, this value is near the lowest valueprojected by any of the scenarios for the 21st century. The verticalaxis gives the projected change in net radiative forcing at a pres-sure level approximating the tropopause (in watts per square meter)for all human-induced changes in greenhouse gases and aerosols(both direct and indirect contributions) over the 21st century usingrelationships employed in the IPCC Second Assessment Report(IPCC, 1996a; Smith et al., 2000), including the uptake of CO2 by theoceans and land. Radiative forcing is important because it is thedriving force for global warming; for reference, the projectedchange in radiative forcing up to the year 1992 is about 1.6 wattsper square meter (IPCC, 1996a). The figure also shows the netradiative forcing and the approximate emissions of carbon used inthe Hadley and Canadian scenarios. For these scenarios, whichincrease the equivalent CO2 concentration by 1% per year, the car-bon emissions are estimated by calculating the emissions neededto match the net radiative forcing after subtracting the radiativeeffects of other greenhouse gases and aerosols based on the aver-age of IS92a and IS92f scenarios, and is an amount between theIS92a and IS92f scenarios. Based on these calculations, theCanadian and Hadley scenarios lie near the mid-range of the pro-posed scenarios in terms of both carbon emissions and net radia-tive forcing. See Color Plate Appendix.

and the Hadley Centre for Climate Prediction andResearch of the Meteorological Office of theUnited Kingdom (“Hadley model scenario,” specifi-cally the simulation using the HadCM2 GCM).Although careful consideration was given,the tim-ing and types of simulations available from USmodeling centers did not meet as many of theimportant criteria as the models selected (seeNRC,1998),although results from US modelinggroups were able to be used by some regionalteams and for some types of investigations.

Using the results from more than one major mod-eling center helps to capture a sense of the rangeof conditions that may be plausible in the future,even though the range of possible futures is likelyto be broader due to the wide range of possibleemissions scenarios as well as uncertainties arisingfrom model limitations. Both of the models select-ed are coupled ocean-atmosphere models that arewell documented and have been peer-reviewed bythe scientific community (Boer et al.,1984,2000b;Johns et al.,1997). Both models include the day-night cycle,which enables them to provide esti-mates of changes in minimum and maximum tem-perature.Both models reasonably represent thebroad scale features of the global climate,includ-ing the major high and low pressure centers andthe major precipitation belts that generate theweather. Even though each simulation can takeseveral hundred hours on the fastest supercomput-ers that are available (Karl and Trenberth,1999),both models have available ensembles of simula-tions (Mitchell et al.,1995;Mitchell and Johns,1997;Boer et al.,2000a).

Although the fundamental physical principles driv-ing these models are similar, there are differencesin how, and even whether, the models incorporatesome important processes. Therefore,there aresome differences in the results of these models.One important factor in causing these differencesis the uncertainty remaining in how best to repre-sent such processes as changes in cloud cover inresponse to global climate change (e.g.,seeMitchell et al.,1987). Because of such uncertain-ties,it is considered important to use models rep-resenting a range of possible values in impact stud-ies. In addition,it needs to be noted that none ofthe model projections consider the potential influ-ences of changes in natural forcings, even thoughit is likely that fluctuations will continue to occuras a result of variations in solar forcing and occa-sional volcanic eruptions (Hyde and Crowley,2000).

Figure 7 and Table 2 provide a comparison of theprojected changes in annual average surface temper-ature for the globe and for the US based on resultsfrom the Canadian and Hadley models. Results are

35


Figure 7: Comparison of the annual average changes in (a) glob-al average surface air temperature (˚F), and (b) US average sur-face air temperature (˚F) from the Canadian model scenario andHadley model scenario simulations used in the NationalAssessment and from the simulations of other modeling groups,including a very recent result from the Hadley Centre model ver-sion 3, Germany’s Max Planck Institute/German ClimateComputing Center (DKRZ), NOAA’s Geophysical Fluid DynamicsLaboratory, and from the Parallel Climate and the ClimateSystem models from the National Center for AtmosphericResearch (which used a slightly lower greenhouse gas emissionscenario and a significantly lower sulfate emissions scenariothan the other models). Decadal means have been plotted tosuppress the natural year-to-year variability. The baseline peri-od is 1961-1990. The anomalies are with respect to the year2000, calculating the values from a 2nd order polynomial fit overadjacent decades. See Color Plate Appendix.

year

U.S. Mean Temperature Anomalies (b)

year

Global Mean Temperature Anomalies (a)

a

b

emissions scenario than IS92a for greenhouse gasesand aerosols (ACACIA-BAU, see Dai et al.,2001) andthat are carried out with a model with a climate sen-sitivity in the lower part of the range of 2.7 to 8.1˚F(1.5 to 4.5˚C). It is important to note that thesemodel results also indicate that substantial warmingoccurs even assuming that emissions are reducedsignificantly below the IS92a scenario.

Although the emissions scenarios are the same forthe Canadian and Hadley simulations,the Canadianmodel scenario projects that the world will warmmore rapidly than does the Hadley model scenario.This greater warming in the Canadian model sce-

also provided for a set of simulations done withother models,some of which became available afterprocessing of results for use in impacts studies hadbeen completed. The Canadian and Hadley simula-tions each use an emissions scenario for changes ingreenhouse gas and aerosol concentrations over the21st century that is designed to represent the IS92a(or no policy intervention) case of the IPCC (1992).New simulations are being carried out by theworld’s modeling groups for the new range of cli-mate scenarios developed by the IPCC (2000). Asan example of the results of this type of simulation,the figure also includes the newer simulations withthe NCAR CSM and PCM models that use a lower

36


Table 2: Model-Simulated Changes in 20th and 21st CenturySurface Temperatures for the US

Model simulated changes in annual-average surface temperature for the 20th and 21st centuries based on linear fits to thedecadal average values derived from the model simulations with comparison to estimates of observed changes for the 20th

century and the range of warming projected by the IPCC (1996a) for the various emission scenarios and climate model sensi-tivities.

Source of Estimate Simulated Change in Global Simulated Change in Average Surface Temperature Average Surface Temperature for

Conterminous US

20th Century 21st Century 20th Century 21st Century

Hadley - Version 2 1.0˚F 4.7˚C 0.8˚F 4.7˚F0.55˚C 2.6˚C 0.4˚C 2.6˚C

Canadian Centre 1.2˚F 7.5˚F 1.9˚F 9.0˚F0.7˚C 4.2˚C 1.05˚C 5.0˚C

Max Planck 1.0˚F 3.4˚F* 1.6˚F 4.1˚F*Institute (MPI) 0.55˚C 1.9˚C 0.9˚C 2.3˚C

Geophysical Fluid Dynamics 1.4˚F 5.7˚F* 1.65˚F 7.8˚F*Laboratory (GFDL) 0.8˚C 3.2˚C 0.9˚C 4.3˚C

Hadley - Version 3 1.1˚F 5.6˚F 1.4˚F 8.85˚F0.6˚C 3.1˚C 0.8˚C 4.9˚C

Parallel Climate 0.9˚F 3.7˚F 0.7˚F 4.1˚F Model 0.5˚C 2.0˚C 0.4˚C 2.3˚C

Climate System 0.9˚F 2.8˚F 0.7˚F 3.3˚FModel 0.5˚C 1.5˚C 0.4˚C 1.8˚C

Observed (Quayle,1999 and 0.7-1.4˚F 0.5-1.4˚FKarl et al.,1995b) 0.4-0.8˚C 0.3-0.8˚C

IPCC (1996a) for 1990 to 2100 1.6-6.3˚F^(uncontrolled sulfur emissions) 0.9-3.5˚C^

IPCC (1996a) for 1990-2100 1.4-8.1˚F^(level sulfur emissions) 0.8-4.5˚C^

*Estimates for less than the full 21st century have been linearly extrapolated to develop an estimate for change over the full century (MPI from 2049to 2100;GFDL from 2090 to 2100).^For estimates for just the 21 st century, about 0.2-0.3˚F (0.1-0.2˚C) must be subtracted,depending on scenario considered.

nario occurs in part because the Hadley model sce-nario projects a wetter climate at both the nationaland global scales,and in part because the Canadianmodel scenario projects a more rapid melting ofArctic sea ice than the Hadley model scenario.Results from other models,with the exception ofthe latest results from the National Center forAtmospheric Research (Dai et al.,2001),are general-ly within or slightly below the lower bound of thisrange. The larger reduction in the NCAR modelresults from the slower rise in greenhouse gas con-centrations that is assumed and due to a projectedincrease of low cloud cover that is not evident insimulations by other models,although these effectsare somewhat offset by reduced loadings of sulfateaerosols. Compared to the range suggested for theyear 2100 in the IPCC results (IPCC,1996a),theHadley model scenario projects warming for the21st century that is slightly above the central IPCCestimate of about 4˚F (2.4˚C) after adjusting for thechange in baseline years. The Canadian model sce-nario projects global average warming that is slight-ly above the high-end of the IPCC suggested range ifsulfur emissions are not controlled,but within therange if they are assumed to be controlled. Thegreater warming for the Canadian model(Hengeveld,2000),as for the Hadley-3 model,is like-ly a result of their higher climate sensitivity. Whileneither the Hadley nor Canadian model scenariosprojects a rate of warming coincident with the lowend of the IPCC range,this lower bound is also gen-erally not consistent with estimates of climate sensi-tivity derived from comparison of model simulationswith the paleoclimatic record or with the extent ofwarming that has occurred over the last two hun-dred years.

All of the models,with the exception of the Hadleyversion 2 GCM (HadCM2),project greater warmingover the US than for the globe as a whole. The vari-ation of results among model results is also greaterfor the US than for the globe. It is especially inter-esting that the projected warming due to thesechanges in greenhouse gas concentrations is ver yrapid after the mid-1970s,when much of the recentwarming began. As an indication of how thesequential improvement of models by the variousgroups may change the results,it is instructive tocompare the results from the HadCM2 that wereused in this Assessment,with the results from theHadley version 3 GCM (HadCM3) that were notavailable in time for full use in this Assessment. Themore recent Hadley model results suggest signifi-cantly more warming over the US than the Hadleymodel selected for this Assessment. Recognizingthat all model results are plausible projections

rather than specific quantitative predictions,the pri-mary models used for this Assessment project thatthe average warming over the US will be in therange of about 5 to 9˚F (about 2.8 to 5˚C).However, given the wide range of possible emis-sions scenarios and uncertainties in model simula-tions,it is possible that the actual increase in UStemperatures could be higher or lower than indicat-ed by this range. Such a warming is approximatelyequivalent to the annual average temperature differ-ence between the northern and the central tier ofstates,or the central and the southern tier of states.

Figure 8 provides similar information for projectedchanges in precipitation. For the globe,the two pri-mary Assessment models represent a range of plausi-ble conditions that are typical of results from otherclimate models that have used the same emissionsscenario,although the simulation of NOAA’sGeophysical Fluid Dynamics Laboratory (GFDL)does suggest an even greater increase in global pre-cipitation than either of these primary models.Over the US,the spread among model results isgreater than over the globe due to the patchiernature of precipitation and changes in precipitation.The Hadley model scenario projects a very largeincrease in precipitation (which is one reason itstemperature increase is lower than for other mod-els) whereas the Canadian model scenario resultsshow an increase mainly in the second half of the21st century. The greater variability of the precipita-tion results,compared to the temperature results,reflects the larger natural variability of precipitation.By using the selected results from the Canadian andHadley models, we are not only capturing results fordiffering model sensitivities,but also,to a largeextent, for much of the wet/dry and hot/warmrange of future climate conditions generated by thewider set of climate models. As such,these casesseem quite representative of the types of conditionsthat could occur.

While the available information provides quite plau-sible estimates for the future,there are importantlimitations that need to be recognized:

• Each model simulation provides a snapshot ofthe temporal and spatial variations of the climateas the global climate is evolving through time inresponse to changes in greenhouse gases andaerosols. Because of inherent variability in themodel that results from small differences in theinitial model conditions,only by employing anensemble of simulations would we be able toassess the statistical significance of the modelresults for any decade over this interval. When

37


emissions scenario being used for the model runswe are using is described in a subsequent sec-tion). For the future,the actual emissions ofgreenhouse gases and aerosols are likely to be dif-ferent than the baseline used. For example,it isquite possible that emissions of both greenhousegases and aerosols may be lower (as a result ofsocietal development,control measures,etc.) orhigher if oil shale and coal become the fuels ofchoice throughout the world. Changing the emis-sions scenario would give different results,although the rate of climate change over the nextfew decades is not likely to differ significantlyfrom the model results because of the momentumcreated by climate and global energy systems.

• Use of only two model simulations provides a lim-ited opportunity to investigate the consequencesof climate variability and change. To help over-come this limitation, regions and sectors havebeen asked,as explained earlier, to look both atthe historical record and to consider cases thatreflect educated guesses based on the nature andimportance of specific regional and sector sensi-tivities. One tool developed for use in the sensi-tivity analyses is the “Probabilities of TemperatureExtremes”CD-ROM that has been developed byNCDC. Other approaches focus on drawing infor-mation from the regional paleoclimatic record.

Recognizing the limitations in the minimum strategyapproach that could be proposed for the entire setof Assessment teams,some groups have had theresources available to carry through additionalimpact studies using results from the models devel-oped at the National Center for AtmosphericResearch (NCAR),NOAA’s Geophysical FluidDynamics Laboratory (GFDL), NASA’s GoddardInstitute for Space Studies (GISS),and the MaxPlanck Institüt/Deutsches Klimarechencentrum(MPI/DKRZ, referred to simply as MPI) in Germany.To support this extended effort,access to the widerset of climate information is provided through theNational Assessment web site.

While GCMs have shown significant improvementover recent decades,and the models used in theAssessment are considered among the world’s best,there are a number of shortcomings that arise inapplying the models to study potential regional-scaleconsequences of climate change. For thisAssessment,several types of effort have been used tostart to address these problems. Of most importancefor the analyses done as part of the NationalAssessment,the results of the GCMs have beenpassed through the VEMAP processing algorithms sothat information could be provided at a scale that

an ensemble of simulations is analyzed,the long-term trends in variables have been found to begenerally consistent across multiple simulations,but quite variable for particular years,decades,and locations.

• The particular simulations we have selectedreflect only one particular emissions scenariorather than a range of emission scenarios (the

38


Figure 8: Comparison of the annual average changes in (a) globalaverage precipitation (inches per month), and (b) average precipita-tion over the US from the Canadian model scenario and Hadleymodel scenario simulations used in the National Assessment andfrom the simulations of other groups (same as for Figure 7). Thebaseline period is assumed to be 1961-1990. Although decadalmeans have been applied to suppress year-to-year fluctuations, thegreater variability of precipitation than temperature still reveals sig-nificant variations due to natural factors; the magnitude, althoughnot the timing, of the remaining fluctuations may be consideredplausible. The anomalies are with respect to the year 2000, calculat-ing the values from a 2 nd order polynomial fit. See Color PlateAppendix.

year

Global Precipitation Anomalies

US Precipitation Anomalies

39


Normalization of Results from Climate Models

While the Canadian and Hadley models both provide reasonable simulations of the large-scale features of the20th century climate over the US,there are differences in the absolute values of temperature and precipitationthat could affect many types of impact studies if adjustments were not made. For the 20

thcentury, this

process is accomplished by simply driving the impact models with the observed climatic conditions ratherthan the model-generated conditions. For studying the 21

stcentury, this procedure is not possible as observa-

tions of the future are not available. Instead,the assumption is made that the differences between models andobservations for the 20

thcentury are systematic – that is,that the differences between models and observa-

tions are a result of limitations in the model formulation and will be present in simulations for both the 20th

and 21st centuries. If this assumption is valid,then the changes in climate due to human activities can bedetermined by taking the difference between a model simulation with increasing concentrations of green-house gases and one simulation without such changes and adding the difference to the observations for the20

thcentury to yield plausible estimates for the changing climatic conditions of the 21

stcentury. Although

this assumption is certainly not completely valid,it is likely to be sufficiently valid that the uncertainties intro-duced in making this assumption will, for many types of situations,be of less importance than uncertaintiesresulting from other factors (e.g.,the differences between models,uncertainties in climate sensitivity tochanges in greenhouse gases,uncertainties in impact models,etc.).

To carry out this normalization of the model results using the differencing approach,and to provide improvedspatial resolution of key climate variables,the VEMAP methodology applied initially to the observed stationdata was used to process the Canadian and Hadley model scenarios of climatic changes during the 21

stcentu-

ry. This procedure was done by interpolating the monthly average changes calculated by the models to theVEMAP grid and then basing the scenario for the 21

stcentury on the model calculated increment to the 20

th

century climate baseline. In the case of temperature,the adjustment was carried out by adding the modelestimate for the monthly average change in temperature from the model’s 1961-90 baseline to the local valueof the observed monthly baseline temperature for the same period. For precipitation,the adjustment wasmade based on the multiplying by the ratio (percentage) change calculated by the model. In this way, projec-tions for the 21

stcentury were made for changes in mean maximum surface air temperature,mean minimum

surface air temperature,and total precipitation on a monthly basis. A weather generator was used to derivedaily values for these variables,and incoming solar radiation and humidity were then derived from these vari-ables. These data sets are available at http://www.nacc.usgcrp.gov/scenarios and values for particular regionsor time periods can be extracted by going to http://eos-webster.sr.unh.edu.

While this application of the VEMAP technique of using the changes calculated by the models to simulate thechanges to the historical record provides a practical way of accounting for the systematic offsets betweenmodeled and locally observed conditions,this technique is not without its limitations. For example,care muststill be taken when analyzing the effects of special situations where thresholds effects might occur (e.g.,thepresence or absence of snow cover in mountainous regions) resulting in projected changes that may be toostrong or too weak. Also,assuming that the temperature changes will be the same in valleys and on moun-taintops fails to deal with the effects of inversions and the special weather conditions of mountain regions.Using the ratioing approach to estimate precipitation change also assumes,at least to some extent,that weath-er systems will be of the same type,being different only in overall intensity or number, while not recognizingthat changes in storm direction into mountainous regions could have a large effect. Darwin (1997) arguesthat at least some of these limitations can be reduced,especially in desert regions, by using absolute amountsof precipitation to make the adjustment;however, in mountainous regions,this approach seems to fail to dealwith the strong gradients in precipitation with altitude. Alcamo et al.(1998) have compared the risk forworldwide natural vegetation using the two approaches,and find a lower risk using the ratioing approachthat is used in the Assessment than the difference-adjustment technique,suggesting that the conclusionsdrawn in this Assessment may be somewhat conservative,although this uncertainty is likely less than theuncertainty resulting from the differences in the model projections.

What is most clear is that, for future assessments,meso-scale models need to be used to more rigorously andaccurately simulate regional patterns of changes in precipitation (and such efforts are already underway in acouple of the regions).

tions very accurately, several complications must beaccounted for in making the comparisons. First,themodel simulations have not been designed to,andcannot be designed to, exactly reproduce the cli-mate of the 20 th century. One reason that reproduc-tion of the 20 th century climate is not possible isthat observations are poor or entirely lacking ofchanges in some of the factors that could lead topart of the naturally induced fluctuations in the cli-mate. These factors include changes in solar radia-tion,9 injection of volcanic aerosols into the strato-sphere,and the state of the global ocean and icesheets at the start of the century. Over the longterm,omitting such natural forcing factors shouldtend to average out to a near zero net effect onglobal average temperatures. For this reason,theeffect of these omissions is often assumed to besmall over periods of many decades compared tothe steady and long-term growth of the greenhouseeffect. Second,because of the chaotic nature of theclimate, we cannot expect to match the year-by-yearor decade-by-decade fluctuations in temperaturethat have been observed during the 20th century.Third,these particular model simulations do not yetinclude consideration of all of the effects of human-induced changes that are likely to have influencedthe climate,including changes in stratospheric andtropospheric ozone and changes in land cover (andassociated changes relating to biomass burning, dustgeneration,etc.). Finally, while it is desirable formodel simulations not to have significant biases inrepresenting the present climate,having a modelthat more accurately reproduces the present climatedoes not necessarily mean that projections ofchanges in climate developed using such a modelwould provide more accurate projections of climatechange than models that do not give as accuratesimulations.This can be the case for at least two rea-sons. First,what matters most for simulation ofchanges in future climate is proper treatment of thefeedbacks that contribute to amplifying or limitingthe changes,and accurate representation of the 20th

century does not guarantee this will be the case.Second,because projected changes are calculatedby taking differences between perturbed and unper-turbed cases,the effects of at least some of the sys-tematic biases present in a model simulation of the

was comparable to the information in the data setsfor the historical period and in a way that account-ed for at least some of the shortcomings and biasesin the models. In particular, the model scenarioresults used in the impact assessments were adjust-ed to remove the systematic differences with obser-vations that are present in the GCM calculations inparticular regions due to mountainous terrain andother problems. The VEMAP normalization processis described in the box “Normalization of Resultsfrom Climate Models.”

In addition,some regional teams have applied othertypes of “down-scaling”techniques to the GCMresults in order to derive estimates of changesoccurring at a finer spatial resolution. One suchtechnique has been to use the GCM results asboundary conditions for mesoscale models thatcover some particular region (e.g.,the West Coastwith its Sierra Nevada and Cascade Mountains).These models are able to represent importantprocesses and mountain ranges on finer scales thando GCMs. However, these simulations are very com-puter intensive and it has not yet been possible toapply the techniques nationally or for the entire 21st

century. With the rapid advances in computingpower expected in the future,this approach shouldbecome more feasible for future assessments. Toovercome the computational limitations ofmesoscale models,other participants in theAssessment have developed and tested empiricallybased statistical techniques to estimate changes atfiner scales than do the GCMs,and these efforts arediscussed in the various regional assessmentreports. These techniques have the importantadvantage of being based on observed weather andclimate relationships,but have the shortcoming ofassuming that the relationships prevailing today willnot change in the future.

CLIMATE MODELSIMULATIONS OF THE 20th

CENTURY FOR THE USAn important measure of the adequacy of the appli-cability of these models for simulation of future cli-matic conditions is to compare their results for sim-ulation of the climate of the 20th century over theUS with observations8. In conducting these simula-tions,the models are driven by observations and,particularly for aerosols, reconstructions of thechanging composition of the atmosphere. Whileone might want the simulations to match observa-

40


8It should be noted that while we are interested in changes over theUS,these changes are in many cases determined by how well themodel represents changes in the global scale features of the climatethat in turn then af fect what is happening over the US and in particu-lar regions. Although the models selected do include flux adjustmentsto reduce drift in global average temperatures,these flux adjustmentshave only a limited influence on determining the patterns of continen-tal-scale climate simulated by these models. It should also be notedthat models not including flux adjustments give a generally similar pat-tern and range of model projected changes in climate.9The newest GCM simulations are beginning to in vestigate the effectsof past variations in solar radiation on climate, even though reconstruc-tions of past levels of solar output are uncertain.

present climate can be eliminated. While potentialnonlinearities and thresholds make it unlikely thatall biases can be removed in this manner, it is alsopossible that the projected changes calculated bysuch a model could turn out to be more accuratethan simulations with a model that provided a bet-ter match to the 20th century climate.

Recognizing these many limitations, evaluation ofthe simulations of the Canadian and Hadley modelsare presented here to give an indication of the gen-eral adequacy of the models for use in these studies.Analyses at the global scale by the two modelinggroups indicate that there is general agreement withthe observed long-term trend in temperature overthe 20th century, although there is significant varia-tion over decadal time scales (e.g., Johns et al.,1997;McFarlane et al.,1992;Flato et al.,2000). As shownby Stott et al.(2000),simulations with the Hadleymodel also show that, by accounting for changes ingreenhouse gas concentrations,sulfate aerosols,andsolar forcing,there is a close similarity between theobserved and the modeled climates,with bothmodel simulations warming about 1˚F during the20th century and showing a roughly similar temporalpattern even though not all influences were consid-ered.

Few of these comparisons have focused on the char-acter of the simulations at the continental andnational scale that are of interest in this Assessment,and so this section presents a selection of thesemodel results. At these scales,so many types ofcomparisons can be made,and there are so manyways to display and interpret the results,that the setof comparisons included here is augmented by addi-tional comparisons available on the Web site10 toprovide the interested reader the opportunity togain a more complete perspective. The set of fig-ures here have been chosen to illustrate that resultsfrom these models,while not predictions,are plausi-ble and suitable for use in investigating the potentialconsequences of climate variability and change forthe US.

Figure 9 compares the Canadian and Hadley modelscenarios to observations,presenting results forannual average temperature and for seasonal tem-perature range11 (summer average temperatureminus winter average temperature) for the period1961-1990;this period, by common convention,isconsidered the baseline climate period. For annualaverage temperature,the model results and observa-

tions have quite similar values and distributionsacross the US,with average temperatures exceeding80˚F (about 28˚C) along the southeastern edges ofthe US and near 40˚F (about 5˚C) across the north-central US. The maps of the seasonal range in tem-perature across the US (summer minus winter)show that the seasonal ranges of temperature forthe models extend from about 5˚F (about 3˚C) nearsouthern and southwestern coastal regions to over50˚F (about 28˚C) in the northern Great Plains,inreasonable concurrence with observations.

The comparisons also show that the models are abit warmer than obser vations along mountain ridges(e.g.,the Sierra Nevada and Cascade Mountains) anda bit colder than observed over mountain basins.Doherty and Mearns (1999) report that both modelsexhibit large year-round cold biases over mountain-ous regions of the West when compared to theLegates and Wilmott (1990a) climatology. However,that climatology likely has a warm bias (making themodels look cold) because most observing stationsare located in valleys in mountainous regions. TheVEMAP surface climatology used in the NationalAssessment comparisons improved on the Legatesand Wilmott climatologies by adding in informationfrom a large number of high altitude stations andotherwise accounting for the ef fects of mountains.Compared to this presumably more accurate repre-sentation of the observed conditions,the model dif-ferences with observations are smaller, but not elimi-nated.

Differences with observations remain particularlylarge over the southern Rocky Mountains and GreatBasin (see Web site for a map of actual differences).These differences are most likely due to the effectsof smoothing the mountain ridges and uplifting themountain valleys to match the relatively coarse reso-lution available in current climate models (figures ofthe differences in topographic height of models andobservations are also shown on the Web site). Bothprimary models also exhibit a warm bias overHudson Bay during winter that extends southwardinto the northern US. This bias may be partly due toinsufficient observational measurements overHudson Bay itself, so that the obser ved surface tem-perature is likely more representative of cold landareas than of water bodies covered by sea ice(Doherty and Mearns,1999). Other biases may wellreflect the limited spatial resolution and representa-tion of climatic processes in the models. For exam-ple,both models also have a warm bias during sum-mer in the central Great Plains and Midwest thatprobably reflects inadequate treatment of summerconvection and soil moisture processes. This bias

41


10See additional figures at www.cgd.ucar.edu/naco/found/figs.html.11Figures showing the model projections of temperature for the sum-mer and winter seasons and for differences between simulations andobservations are available on the Web site.

ocean areas (Doherty and Mearns,1999). These dif-ferences of several degrees can create problems inthe direct application of model results,but theagreement of the overall patterns and seasonalranges provides considerable confidence that theprojected changes in temperature due to humaninfluences are plausible for use in impacts studies.

extends further into the eastern US in the Canadianmodel scenario than in the Hadley model scenario.Over adjacent ocean areas,the Canadian model alsoindicates temperatures slightly above observationswhereas the Hadley model indicates temperaturesare slightly below observations,likely reflectingremaining problems with representation of coastal

42


Comparison of Annual Average Temperatures & Seasonal Range

VEMAP Observed Seasonal Temperature Range

Canadian Model Seasonal Temperature Range

Hadley Model Seasonal Temperature Range

+50ºF

45ºF

40ºF

35ºF

30ºF

25ºF

20ºF

15ºF

10ºF

5ºF

0ºF

+50ºF

45ºF

40ºF

35ºF

30ºF

25ºF

20ºF

15ºF

10ºF

5ºF

0ºF

+50ºF

45ºF

40ºF

35ºF

30ºF

25ºF

20ºF

15ºF

10ºF

5ºF

0ºF

Figure 9: Comparison of annual average temperatures and seasonalrange (summer/winter) (˚F) for the US from (a, d) observations, (b, e)the Canadian model scenario, and (c, f) the Hadley model scenario.Results are for the period 1961-90. The model-simulated temperatures,their spatial patterns, and their seasonal ranges are in quite goodagreement with observations generated by the VEMAP project (Kittel etal., 1995, 1997; VEMAP Members, 1995). Mean temperature is calculat-ed as the mean of the minimum and maximum temperatures, so thatthe model data are consistent with the VEMAP data. [Seasonal and dif-ference plots are also provided on the Web site containing the figures.]See Color Plate Appendix.See Color Plate Appendix

Observed 1961-1990 Average

Canadian Model 1961-1990 Average

Hadley Model 1961-1990 Average

100ºF

90ºF

80ºF

70ºF

60ºF

50ºF

40ºF

30ºF

20ºF

10ºF

0ºF

+100ºF

90ºF

80ºF

70ºF

60ºF

50ºF

40ºF

30ºF

20ºF

10ºF

0ºF

+100ºF

90ºF

80ºF

70ºF

60ºF

50ºF

40ºF

30ºF

20ºF

10ºF

0ºF

a

b

c

d

e

f

Figure 10 presents similar results for annual totalprecipitation and seasonal range (summer minuswinter, in inches/month). Precipitation amounts incomplex terrain are highly variable as a result of thelocal interaction of storms with mountains and localvariations in the surface warming that drives con-vective rain systems (Legates and DeLiberty, 1993;

Legates 1997). The relative coarseness of the modelresolution means,therefore,that agreement is notlikely to be as good,especially over the western US.Both models and observations (from VEMAP andLegates and Wilmott,1990b) show a similar rangefrom a minimum in the dry areas of the Southwestto much larger amounts over other parts of the

43


Comparison of Annual Total Precipitation & Seasonal Range

VEMAP Observed Seasonal Precipitation Range

Canadian Model Seasonal Precipitation Range

Hadley Model Seasonal Precipitation Range

>10

8

6

4

2

0

-2

-4

-6

-8

-10

>10

8

6

4

2

0

-2

-4

-6

-8

-10

>10

8

6

4

2

0

-2

-4

-6

-8

-10

Figure 10: Comparison of annual total precipitation and seasonalrange (summer minus winter) in inches per month for the US from(a,d) observations, (b,e) the Canadian model scenario, and (c,f) theHadley model scenario. Results are average inches/month for theperiod 1961-90. The model-simulated precipitation totals, their spatialpatterns, and their seasonal ranges are in reasonable agreement withobservations generated by the VEMAP project (Kittel et al., 1995,1997; VEMAP Members, 1995). [Difference plots are also provided onthe Web site containing the figures.] See Color Plate Appendix.

VEMAP Observed Annual Precipitation

Canadian Model Annual Precipitation

Hadley Model Annual Precipitation

>5”

4”

3”

2”

1”

0”

>5”

4”

3”

2”

1”

0”

>5”

4”

3”

2”

1”

0”

a

b

c

d

e

f

plausibly represented,with the indication that pre-cipitation in the West occurs much more in winterthan summer whereas over the rest of the countrythere tends to be a modestly larger amount in sum-mer. Overall,the model results show broad agree-ment with observations, except in the Canadianmodel over Florida. Similarities,however, are evi-dent in the simulation of high amounts of precipita-tion in the West in winter and low amounts in sum-mer giving a negative seasonal range,and asmoother seasonal cycle in the eastern US.However, there are important dif ferences,especiallyin the regions of mountainous terrain where obser-vations of precipitation are also problematic due tothe great spatial variability.

As for the temperature differences,differences withobservations arise because the models do not fullyrepresent the high reach of mountain ridgelines.Because of this discrepancy, the models do not cre-ate as much precipitation along the Pacific coastridgelines as is observed,allowing more precipita-tion further inland. For the rest of the country, com-parisons by Doherty and Mearns (1999) indicatethat the models have a wet bias over northeasternNorth America in spring and summer, and a dry biasin southern North America in both summer and win-ter. In our comparisons,the Canadian model (butnot the Hadley) shows a wet bias in the northeast-ern US,but a dry bias when integrated over thewhole country. The biases in coastal regions mayresult from the relatively coarse resolution of themodels,which does not allow adequate representa-tion of the relatively small-scale spatial patterns ofthe sea breeze and other coastal meteorology. Also,the tropical rainbelt created by the IntertropicalConvergence Zone (ITCZ) does not extend farenough northward in either model,creating drybiases in some of the equatorial regions of theNorthern Hemisphere. That there are differencesthat must be accounted for in the analyses becomesespecially clear when focusing on very particularregions (e.g.,Florida),and the Web site provides dif-ference maps for simulations of the total and season-al precipitation.

Because some impact studies require scenarios ofchanges in day-to-day variability in the weather, com-parison should also be made over this time scale,considering, for example,the adequacy of modelsimulations of the frequency, intensity, and amountsof precipitation. Unfortunately, such detailed com-parisons are only beginning to be carried out and socaution must be exercised in interpretations thandepend on these results. Nonetheless,as for temper-ature,if account is taken of systematic differences,

country. Although the very broad-scale patterns aresimilar, the role of mountain chains in concentratingprecipitation into particular locations is much moreevident in the observations than in the models withtheir very smoothed representation of mountainranges. The pattern of the seasonal range is also

44


Figure 11: Time histories of the changes in (a) annual average tem-perature (˚F), and (b) annual total precipitation (inches per year) forthe 20th century based on observations and on simulations from theCanadian and Hadley models, calculated as 10-year running meansfrom 1900 to 2000. Mean temperature is the actual mean tempera-ture from the models, rather than the mean of the minimum andmaximum temperatures. Anomalies are shown with respect to1961-1990. In these simulations, unlike in intercomparisons of theatmospheric models as in the AMIP project (Gates et al., 1999), theocean temperatures are freely calculated and the concentrations ofgreenhouse gases and aerosols are imposed; natural forcings,such as changes in solar radiation and volcanic eruptions that arelikely affecting the observed climate are not, however, being treatedin the models because observations of their precise radiative influ-ences are not available. See Color Plate Appendix.

Mean Temperature Change

Precipitation Change

a

b

Year

Year

the model results would seem to give a plausible setof baseline conditions to use in estimating changesto temperature and precipitation that could occur asthe climate changes.

In evaluating model performance,it is also impor-tant to look at how well the models simulate thetemporal variations of climatic conditions in theimmediate past. Figure 11 shows a comparison overthe US of the observed and modeled time historiesof changes in annual average temperature and annu-al total precipitation during the 20th century.Remembering that complete agreement of each cli-mate fluctuation should not be expected due to thenatural variability of the climate,these plots indicatethat the models generally have the right magnitudeand duration of natural climate anomalies and that,with the exception of the start of rapid warminglate in the century in the Canadian model scenario,the trends are plausibly similar. The Web site pro-vides diagrams that go beyond these comparisons toprovide estimates of the actual values of tempera-ture and precipitation,thereby illustrating the sys-tematic differences that are present between themodels and observations. These differences ariseboth because of limitations in the models (e.g.,inad-equate resolution,inadequate representation of vari-ous processes,etc.) and shortcomings in the moni-toring network (e.g., few stations at high latitudes,etc.). To the extent that these differences are sys-tematic,the model projections of changes can beused if care is taken in working near thresholdssuch as the freeze line. To the extent that the differ-ences are inherent in the treatment of climateprocesses and how they might respond with a dif-ferent climate,uncertainties are introduced into theclimate scenarios, again emphasizing that theseresult must be viewed as scenarios rather than pre-dictions.

While these analyses indicate that the model resultsare generally similar to observations,it is clear thatsystematic errors are present,especially in moun-tainous areas. To account for these dif ferences,his-torical analyses have generally been based on com-pilations of observational data,such as the USHCNor VEMAP data sets, rather than numerical modelresults,and appropriate adjustments need to bemade when applying model results for the future (asexplained in the box on page 28 on Normalizationof Results).

SCENARIOS FOR CHANGES INATMOSPHERIC COMPOSITIONAND RADIATIVE FORCING FORTHE 21st CENTURYProjecting changes in climate for the 21 st centuryrequires not only a tested climate model,but also ascenario for the development and evolution of thehuman activities that are expected to affect the cli-mate. In particular, projections of climate changerequire a projection of how atmospheric composi-tion will be changing in the 21st century as a resultof the ongoing use of fossil fuels and the release ofother greenhouse gases12. To provide the basis forsuch estimates,scenarios of societal and technologi-cal evolution during the 21 st century must be devel-oped;these in turn can be used to develop emis-sions scenarios. The accuracy of these scenarios isnecessarily limited by uncertainties in insights andassumptions about what will happen many decadesinto the future. Because of the resulting uncertain-ties,the concentration scenarios that are used,likethe climate scenarios,cannot be viewed as predic-tions of the future. Instead,they must be treated asplausible estimates of future conditions that areappropriate for use in exploring vulnerabilitiesthrough analysis and assessment.

A range of scenarios has been developed by a num-ber of groups to describe how atmospheric concen-trations of CO2, other GHGs,and aerosols maychange in the future. These scenarios are generallybased on projections of future changes in popula-tion,energy technology, economic development,environmental controls,and other factors. The 1992scenarios proposed by the Intergovernmental Panelon Climate Change (IPCC,1992) have become wide-ly used because of the international effort that wentinto their consideration13. The set of 1992 IPCCgreenhouse-gas emission scenarios was based uponsix plausible demographic and socioeconomic sce-narios that spanned a wide range of possibilities forpopulation growth,types of energy use,and rates ofeconomic growth. The range of projected emissionsfor the 21st century is quite broad (see IPCC,1992and Figure 6).

The central baseline (sometimes called “business-as-usual”) estimate from the set of IPCC 1992 scenar-ios is closely comparable to the radiative forcingscenario represented by a 1% per year compounded

45


12Note that for the purposes of these studies,the level of solar insola-tion and the occurrence of volcanic eruptions are assumed to remainas they were for the 20 th century. Even though changes are likely (e.g.,see Hyde and Crowley, 2000),the net effect of these changes are likelyto be small in comparison to the human-induced influences on radia-tive forcing.

13These scenarios are presently being updated as part of the effortleading up to the IPCC’s Third Assessment Report (IPCC,2000). Thenewer scenarios tend to span a similar range to the 1992 scenarios.

the IS92a scenario (see Figure 6). To the extent thatactual greenhouse gas emissions might be greater orless than this central scenario over the long term,the climatic changes at a given time would begreater or less. Alternatively, the climatic changesthat are projected with this scenario would be pro-jected to occur either earlier or further in thefuture,although the difference would likely be lessthan one or two decades. Although the potentialconsequences of a somewhat faster or slower rise ingreenhouse gas emissions has not yet been evaluat-ed,it seems likely that such changes in emissionsscenarios would have a relatively small influenceover the climate changes projected for the first halfof the 21st century.

Figure 12 shows the projected changes in CO2 andequivalent CO2 concentration for the IS92a scenario(projected changes in the concentrations of othergreenhouse gases are described in IPCC,1992) andthe 1% per year change in equivalent CO2 concen-

increase in the equivalent CO2 concentration thathas been used by most climate modeling groups togenerate their central estimates of potential climatechange for the 21st century. This scenario has beentaken as the baseline scenario for this study becauseof its wide use,because it represents neither maxi-mum nor minimum emissions projections (seeFigure 6),and because this Assessment did not havethe resources to either construct better alternativescenarios or ensure that such scenarios would beused by the climate modeling groups for calcula-tions that would be available in time for impact eval-uation as part of this Assessment. Although theIS92a emissions scenario tended to overestimategreenhouse gas emissions during the 1990s (Hansenet al.,1998),it is not clear that the recent tendencytoward lower emissions compared to IS92a will per-sist as the global economy recovers from its recentrecession. In particular, the new IPCC (2000) sce -narios suggest a wide range of possible future emis-sions scenarios,some higher and some lower than

46


Figure 12: Comparison of different projections for aerosol effectsand for (a) the CO 2 and equivalent CO2 concentrations, and (b) theassociated radiative forcings for the period 1850-2100. In the topfigure, the lavender line shows the IPCC’s IS92a scenario estimateof the CO2 concentration; values prior to 1990 are based on obser-vations. Based on this projection, the CO2 concentration wouldrise to about 705 ppmv in 2100 from a level of about 353 ppmv in1990. Because many of the climate models treat the effects of theset of human-affected greenhouse gases by use of an equivalentCO2 concentration, the green line shows the scenario for the equiv-alent CO2 concentration, which rises to about 1022 ppmv in 2100from a value of about 410 ppmv in 1990. For this curve, the equiva-lent CO2 concentration is calculated so as to incorporate the radia-tive effects of changes in the concentrations of all greenhousegases using the IPCC radiative forcing equivalents (the conversionfactor is 6.3 based on Appendix 2 in IPCC, 1997). The light blue lineshows the equivalent CO2 concentration that results from using theHadley radiative forcing equivalents to approximate the IS92a sce-nario; the conversion factor used is 5.05 (John Mitchell, personalcommunication). Using the Hadley conversion factor, the equiva-lent CO2 concentration for the IS92a scenario would rise to about1409 ppmv in 2100. The red line shows that the Hadley IS92a equiv-alent CO2 scenario is quite well fitted by use of a 1% per year com-pounded increase in the Hadley equivalent CO2 concentration. Inthis case, the CO2 equivalent concentration in 2100 reaches about1346 ppmv. The deep blue line shows the IPCC IS92a scenario forsulfur emissions, which shows a rise until about 2050, when emis-sions roughly level off. While there are some differences in theprojected concentrations of equivalent CO2 between the IPCC(1996a) and the Hadley model scenario, the bottom figure showsthat these differences are mostly overcome when comparing theradiative forcings that are projected by the IPCC and are actuallyused in the Hadley model scenario. The red and blue lines, respec-tively show the radiative forcings as projected by the IPCC (solidlines) and as included in the Hadley model (dotted lines). For bothforcings, the Hadley model projects slightly less influence than theprojections using the IPCC conversion factors. When these forc-ings are combined, as shown by the green lines, the net radiativeforcings projected by the IPCC and used in the Hadley model 1%per year scenario are very close. See Color Plate Appendix.

Forcing Scenarios

Radiative Forcing

Year

Year

tration. For the National Assessment,the IS92a timehistory of the CO2 concentration (which rises toabout 708 ppmv by 2100) has been used in thestudies of the consequences of a rising CO2 concen-tration for plants,coral,etc. However, because con-centrations of CH4, N2O,and CFCs are also changing,the model simulations need to be forced by the netradiative effect of these greenhouse gas changes.This consideration is implemented in the Hadleyand Canadian models by increasing the equivalentCO2 concentration by 1% per year (compounded)starting in 1990 to account for the combined radia-tive effects of all the greenhouse gases (Boer et al.,2000a). Thus,while the 1990 concentration of CO2

is about 354 ppmv, the model uses an equivalentCO2 concentration of about 420 ppmv to accountfor the effects of the other greenhouse gases. Forthe year 2100, rather than reaching a CO2 concentra-tion of about 708 ppmv, an equivalent CO2 concen-tration of about 1055 ppmv is reached (note thatthe 1% per year case is slightly higher than this con-centration,actually being closer to case IS92f). Interms of radiative forcing,the 1% simplification over-estimates the net forcing in 2100 of all greenhousegases by about 10% compared with IS92a (ofcourse,there are other scenarios that have higherforcing than IS92a). As shown in Figure 12,theIS92a scenario used by the models also significantlyincreases sulfate emissions (and therefore sulfateaerosol loadings) until about 2050,after which levelsare projected to remain roughly constant. The netchanges in forcing for both the Hadley and Canadianmodel scenarios are,as indicated in Figure 6,nearthe middle of the range for all emissions scenarios.

In that sulfate aerosols contribute significantly to airpollution and acid rain,the newer scenarios in IPCC(2000) suggest that sulfate aerosol levels (and sotheir cooling influence) will be lower than in IS92a,thereby raising the overall warming influence. Inaddition,the new scenarios suggest that significantincreases in the use of fossil fuels will lead toincreased emissions of methane,carbon monoxide,nitrogen oxides,and volatile organic compounds,which will lead to significant increases in bothregional and hemispheric levels of troposphericozone,a strong greenhouse gas.

What is quite clear from Figure 6 is that the radiativeforcing could be either higher or lower than thecase that has been the most frequent reference casefor the modeling groups and is being used in thisAssessment. The normal way to treat such a range ofpossible futures would be to treat a range of possi-ble future conditions rather than rely on only onecase. In this first National Assessment,constraints on

time and resources,however, have forced a limita-tion to considering the consequences of only oneconcentration scenario. While this approach is alimitation that should be relaxed in future assess-ment efforts,the constraints of this limitation arereduced by the recognition that much of the climatechange over the next few decades will be due toalready recorded changes in atmospheric composi-tion. In addition,with the momentum created bythe world’s present use of fossil fuel energy, devia-tions in the concentration scenario for the variousGHGs are likely to have only a limited influence onthe climate over the next few decades. To explorethis issue further, a later section of this chapter doessummarize the climatic consequences of using ascenario that moves toward stabilization of theatmospheric CO2 concentration at double its prein-dustrial value. However, even if such a stringentemissions limitation were imposed now, the effecton CO2 concentrations and climate would be rela-tively modest during the early 21st century beforeincreasing and becoming quite significant duringthe 22nd century.

CLIMATE MODELSCENARIOS FOR CHANGESIN TEMPERATURE,PRECIPITATION, SOILMOISTURE, AND SEA LEVELOVER THE US FOR THE 21st CENTURY

Temperature and Heat Index

All climate models project significant warming forthe 21st century. Results shown in Figure 7 clearlyindicate that the global warming projected for the21st century will be significantly greater than duringthe 20th century. This increase in the rate of warm-ing is due to both the continuing rise in the CO2

concentration projected for the 21st century and thecontinuing response of the climate system to theincreasing rate of rise in the CO2 concentration inthe second half of the 20th century. Figure 7 alsodemonstrates that the projections for warming overthe US are very likely to be greater than for theglobal average,both because warming is greaterover land areas than over ocean areas and becausethe US is located in mid-latitudes. This figure alsoshows that,although the rate of warming is not like-

47


is very likely to increase substantially over comingdecades.

Figure 13 shows the annual average geographic pat-terns of the projected warming across the US as cal-culated by the Canadian and Hadley models14. Thetrends are in degrees (˚F) of warming per centuryand represent the expected warming for the severaldecades around 210015. In the Canadian model sce-nario for the next 100 years,increases in annualaverage temperature of 10˚F (5.6˚C) are projectedacross the central US,with changes about half thislarge projected along the East and West Coasts. Theprojections indicate that the changes will be particu-

ly to be uniform over this period,the average ratewarming rate is very likely to increase during the21st century. This change in rate may occur in anuneven way, with some very warm years,and thensome not-so-warm or even cooler years. Althoughwe do not yet have the ability to forecast theseshort-term fluctuations precisely, the model scenar-ios clearly show that the long-term rate of warming

48


Trends in Annual Average Temperature -21st Century

Canadian Model - 21st Century

Hadley Model - 21st Century

Hadley CM3 Model - 21st Century

Figure 13: Projections across the US of the increase in annualaverage temperature (˚F) over the 21 st century from the (a)Canadian model scenario (VEMAP-processed), (b) Hadley modelscenario (VEMAP-processed), and (c) HadCM3 models. TheHadCM3 results are shown here to point out that different gener-ations of the same basic model can yield results that are as dif-ferent as results of different models. See Color Plate Appendix.

>15

10

5

0

-5

>15

10

5

0

-5

>15

10

5

0

-5

Figure 14: Time histories of (a) maximum and (b) minimum tempera-ture over the US (˚F). The values prior to the present are based onobservations from 1900-1998 (the HCN data set) and values for thefuture are based on the VEMAP version of the Canadian and Hadleymodel scenarios (i.e., in the VEMAP data sets, model projections ofclimate change are added to the observed 1961-90 baseline climate).See Color Plate Appendix.

Maximum Temperature in the US (annual average) Minimum Temperature in the US (annual average)

a b

Year Year

14 Maps of projected changes for the winter and summer seasons are available on the Web site.

larly large in winter, with minimum temperatures ris-ing more than maximum temperatures. Largeincreases in temperature are also projected overmuch of the South in summer. In the Hadley modelscenario,temperatures in the eastern US are project-ed to increase by 3 to 5˚F (2-3˚C) by 2100. Regionsacross the rest of the nation are projected to warmby up to about 7˚F (4˚C). Such changes would beequivalent to shifting the climate of the southern USto the central US and the climate of the central USto the northern US.

Model results from the HadCM3 model are alsoshown in Figure 13. This model is a more recentversion of the Hadley Centre model than was avail-able at the start of this Assessment. This modelshows greater warming in the eastern half of theUS. This reinforces the point that although all themodels agree that there will be a strong warmingtrend,projections of the spatial and temporal pat-tern of the warming differ among the models.

While the maps included in this chapter provideresults for the conterminous 48 states,model resultsare also available for Alaska and for the Pacific andCaribbean Islands on the Web site. Both modelsproject that Alaska will experience even moreintense warming than the conterminous US16. Incontrast,Hawaii and the Caribbean islands are likelyto experience somewhat less warming than the con-tinental US,because they are at lower latitudes andare surrounded by ocean,which warms more slowlythan land. These results are shown in maps appear-ing in the respective chapters of this report.Although the details of the projected climate fluctu-ations over time are less reliable than the projec-

tions of the overall trends,it is useful to examine theprojected time histories of the changes. Figure 14shows the time histories for the projected changesover the US in the annual averages of minimum andmaximum temperature for the two models. As issuggested by the maps,the time series show thatthe warming is projected to be greater in theCanadian model scenario than in the Hadley modelscenario. The larger increase in minimum than max-imum temperature indicates that nighttime tempera-tures are projected to increase more than daytimetemperatures. Factors causing this difference couldinclude the increase in downward infrared radia-tion,the increase in the dew point temperature,changes in cloud cover, changes in soil moisture,and changes in snow and ice cover, each of whichwould act to raise nighttime temperatures morethan daytime temperatures. In addition,an increasein sulfate concentrations or increases in cloud covermight act to limit daytime warming by reflectingmore solar radiation back to space. That both mod-els suggest that minimum temperatures will risemore rapidly than maximum temperature is consis-tent with what has been observed over the pastcentury (Easterling et al.,1997).

Although the two primary models used here projectthat the temperature increase will be greater in thewestern than in the eastern US,the intensification ofthe hydrologic cycle caused by the warming willalso cause an increase in the amount of moisture inthe air. This increase is particularly important forthe southeastern and eastern US,where humidity isrelatively high and upward trends in temperatureare quite large (Karl and Knight,1997). Figure 15shows the projected increases in the heat index

49


Canadian Model Hadley Model+25ºF

+20ºF

+15ºF

+10º

+5ºF

0º

July Heat Index Change - 21 st Century

Figure 15: Projections across the US of the increase in the July heat index (˚F) over the 21st century (˚F per century) from VEMAP ver-sions of the (a) Canadian model scenario and (b) Hadley model scenario. See Color Plate Appendix.

15Because the model simulations are most valid over long-periods of time,these results are based on a linear fit to the model projected changes forthe 21st century rather than being based on the differences between particular years or decades at the beginning and end of the century. Thischoice is intended to make clear that it is the overall century-long rate of change that is the result in which we can have the most confidence.Because of the long-term warming, great care should be taken in comparing this projected rate of change to observed changes over shorter periodsbecause it is widely recognized that there will be considerable natural variability through the century as a result of the effects of natural influencessuch as solar variations, volcanic eruptions,and the ocean-atmosphere interactions that create such fluctuations as ENSO events.16For results for Alaska,see http://www.cgd.ucar.edu/naco/alaska/tx.html.

both primary models project Pacific Ocean warmingand a southward movement of the storm-generatingAleutian Low which would together lead toincreased precipitation along the West Coast. Forthese conditions,a greater fraction of the increasedwintertime precipitation would be expected to fallas rain rather than snow, causing,on average, areduction in mountain snow pack. These changesare likely to increase wintertime and decrease sum-mertime river flows in the West. Even with an accu-rate projection of Pacific Ocean changes,the region-al pattern of this precipitation increase could onlybe roughly estimated due to the limited representa-tion of the region’s mountains (e.g.,see Mearns etal.,1999). As global scale models improve,mesoscale models will be able to be used to explorethis issue further.

Across the Northwest and over the central and east-ern parts of the US,the precipitation projectionsfrom the models are in less agreement. The differ-ences between model projections are likely a resultof a number of factors. For example,the two mod-els show different positions and intensities of thestorm tracks in the Southeast during winter in theirsimulations of recent decades. The Canadian modelscenario projects that there will be a decrease inannual precipitation across the southern half of thenation east of the Rocky Mountains. Decreases areprojected to be particularly large in easternColorado and western Nebraska in the west centralPlains,and in the southern states in an arc fromLouisiana to Virginia. These projected decreases inprecipitation are largest in the Great Plains during

across the US based on the model projections for thechanges in maximum temperature;similar resultshave been reported from the GFDL model(Delworth et al.,1999). The heat index is a measureof the rise of apparent temperature and is a goodmeasure of discomfort because it combines bothheat and humidity ef fects (Steadman,1979). Theseresults indicate that, even though the relative humid-ity may drop slightly (not shown),the rise in theheat index will be more than double the actual risein temperature across much of the South and East,making the projected warming in these parts of thecountry feel particularly significant. By the end ofthe 21st century, the heat index of the Northeast islikely to feel more like that of the Southeast today;the Southeast is likely to feel more like today’s southTexas coast;and the south Texas coast is likely to feelmore like the hottest parts of Central America today.

Precipitation

Figure 16 shows the projected pattern of changes inprecipitation across the contiguous US, expressed asa percentage change from the present amount17. Themost noticeable feature in both models is a project-ed increase in precipitation in California and thesouthwestern US. The projected increase is larger inthe Canadian than in the Hadley model scenario.This feature is a result of a warmer Pacific Oceancausing an increase mainly in wintertime precipita-tion. Although the projected changes over thePacific Ocean are not well-established,particularlywith regard to how El Niño conditions may change,

50


Trends in Annual Average Precipitation -21st Century

Canadian Model - 21st Century

Hadley Model - 21st Century

Hadley CM3 Model - 21st Century

Figure 16: Projections across the US of the changes inannual precipitation over the 21st century (percent changeper century) from the (a) Canadian model scenario (VEMAP-processed), (b) Hadley model scenario (VEMAP-processed),and (c) HadCM3 models. See Color Plate Appendix.

100%

80%

60%

40%

20%

0

-20%

-40%

-60%

-80%

-100%

100%

80%

60%

40%

20%

0

-20%

-40%

-60%

-80%

-100%

100%

80%

60%

40%

20%

0

-20%

-40%

-60%

-80%

-100%

17Changes in the absolute amount of precipitation are shown in fig-ures available on the web site.

summer and in the East during both winter and sum-mer. In the Hadley model scenario,virtually theentire US is projected to experience increases in pre-cipitation,with the exception of small areas alongthe Gulf Coast and in the Pacific Northwest.Precipitation is projected to increase in the easternhalf of the nation and in southern California andparts of Nevada and Arizona in summer, and in everyregion except for the Gulf States and northernWashington and Idaho during the winter. However,while the Hadley (HadCM2) scenario used in thisAssessment suggests greater precipitation in theSouthwest,the more recent HadCM3 model suggeststhat there will be less rainfall in the Southwest;theprojected pattern of change is similar to theCanadian model scenario in parts of the Southeast.Because of the dif ferences among these results,theprojected direction of the trend for changes in pre-cipitation in any given region needs to be viewed asuncertain,although continuation of the increasingprecipitation trend for the US as a whole seems plau-sible. Resolving these differences in precipitationprojections will occur only by increasing resolutionand implementing other improvements in the cli -mate models.

Figure 17 provides the time histories of the project-ed changes in precipitation for the US. Both modelsproject a long-term increase in total annual precipita-tion across the US. However, the time histories clear-ly indicate that the very large variability that current-ly exists is likely to continue,with the possibility ofperiods of both increased and even reduced precipi-tation within the overall upward trend.

Soil Moisture

Projections of changes in soil moisture depend onthe balance between precipitation, evaporation, run-

off, and soil drainage. By itself, an increase in precip-itation would tend to increase soil moisture.However, higher air temperatures increase the rateof evaporation and may remove moisture from thesoil faster than it can be supplied by precipitation.Under these conditions,some regions are likely tobecome drier even though rainfall increases. In fact,soil moisture has already decreased in portions ofthe Great Plains and Eastern Seaboard,including insome locations where precipitation has increasedbut air temperature has risen. Figure 18 shows theprojected changes in the summer soil moistureacross the US. In the Canadian model scenario,theSoutheast and the region extending through the

51


Figure 17: Time history of model projected changes in precipitationover the US (inches per year). The values prior to the present arebased on observations from 1900-1998 (the HCN data set) and val-ues for the future are based on the VEMAP version of the Canadianand Hadley model scenarios. See Color Plate Appendix.

Canadian Model Hadley Model

Summer Soil Moisture - 21st Century

Figure 18: Projections across the US of changes in summertime soil moisture over the 21st century (percent change per century) fromthe (a) Canadian model scenario, and (b) Hadley model scenario. Figure prepared by the National Climatic Data Center. See ColorPlate Appendix.

>100

75

50

25

0

-25

-50

-75

-100

Annual Precipitation in the US

Year

Sea Ice and Sea Level

The two primary model scenarios include projec-tions of a decline in sea ice cover and a rise in sealevel,both of which are of particular importance forassessing the potential consequences of climatechange along coastlines. The Canadian model sce-nario projects that sea ice in the Arctic Ocean isvery likely to melt completely each summer and besignificantly reduced in winter thickness and extentby the end of the 21st century,whereas the Hadleymodel scenario envisions a slower process of melt-ing. Observations indicate that the average depth ofsea ice in the Arctic has dropped by 40%,fromabout 10 feet (3.1 meters) to about 6 feet (1.8meters) over the past three decades (Rothrock etal.,1999);this suggests that the model projections ofmelting of sea ice in the future are likely to be quiteplausible. Sea ice is particularly important forcoastal regions because its presence suppresseswaves from wintertime storms that erode coastlines.In addition,some marine species depend on theprotection or convenience of sea ice to feed andreproduce,making the meltbacks ecologicallyimportant.

Because sea ice floats,its melting does not affect sealevel. However, the melting of glaciers on land andthe warming of ocean waters do cause sea level torise. Over the 20th century, observations indicatethat sea level has risen about 4 to 8 inches (10-20cm). In estimating the potential rise in sea level dur-ing the 21st century, the Canadian model scenarioincludes consideration of only the sea-level risecaused by the warming of ocean waters (the ther-mal expansion effect),whereas the Hadley modelscenario also includes consideration of the risecaused by the melting of mountain glaciers.Although melting of the polar ice sheets may con-tribute to sea level rise in the long-term,neither ofthe model estimates includes consideration of thechanges in sea level caused by the accumulation ormelting of snow on Greenland and Antarctica18. Theglobal climate models also do not include the local,but significant,component of sea-level changecaused by changes in the heights of coastlines asthey rise or fall due to regional or even local effects(e.g.,the pumping out of groundwater, earthquakes,isostatic adjustment from the last glacial period,

central US to just east of the Rocky Mountains areprojected to experience the largest decreases in soilmoisture. Increases in soil moisture are projectedfor the areas surrounding Iowa and from Utah toCalifornia. In the Hadley model scenario,summersoil moisture is projected to increase in the easternhalf of the US and is generally unchanged or slightlydecreased from the Rocky Mountains westward,except for Southern California.

Increased drought becomes a national problem inthe Canadian model scenario and is also found inthe GFDL model (Wetherald and Manabe,1999).Intense drought tendencies occur in the region eastof the Rocky Mountains and throughout the Mid-Atlantic-Southeastern states corridor. Increased ten-dencies toward drought are also projected in theHadley model scenario for the regions immediatelyeast of the Rocky Mountains. California and Arizona,as well as the region from eastern Nebraska to theVirginia coastal plain are projected to have reduceddrought tendency. The differences in soil moistureand drought tendencies are likely to be the mostcritical for agriculture, forests, water supply, and lakelevels.

52


Figure 19: Historic and projected changes in sea level (inchesabove baseline) based on the Canadian and Hadley model scenar-ios. The Canadian model projection includes only the effects ofthermal expansion of warming ocean waters (F. Zwiers, personalcommunication). The Hadley model simulation adds on the sealevel increment of melting of mountain glaciers (Gregory andOerlemans, 1998). Neither model includes consideration of possi-ble changes of sea level (upward or downward) due to melting oraccumulation of snow on Greenland and Antarctica. See Color PlateAppendix.

Sea Level Rise

Year

18 In the IPCC 1996 report,the accumulation and melting of Antarcticaand Greenland were assumed to balance to give no net contribution tosea level change over the 21st century;more recent studies are findingthat some parts of Greenland are melting while others seem to beaccreting,and that the global warming at the end of the last glacialperiod apparently has initiated deterioration of parts of the WestAntarctic Ice Sheet. Due to limitations in the obser vations of these icesheets,ho wever, projections included here do not include contribu-tions from changes in the size of these polar ice sheets, even thoughstrong warming seems likely to contribute to their melting and thenceto sea-level rise.

etc.). Given these caveats, Figure 19 presents themodel projections for the estimated rise in globalsea level from the two models used in the NationalAssessment. Over the next hundred years,a rise ofsea level of about one and a half feet (about 0.5meters) is considered likely based on these projec-tions. Maps of the regional pattern of sea-levelchange around the US are presented in the Coastalchapter, where the effects of local changes in thelevel of the coastline are also considered. A rise ofthis amount would be several times as much asoccurred during the 20th century. As indicated inthe Coastal chapter, even a relatively modest risecan cause extensive coastal erosion (e.g.,seeLeatherman et al.,2000).

CLIMATE MODELSCENARIOS OF CHANGES INCLIMATIC PATTERNS,VARIABILITY, STORMS, ANDEXTREMES FOR THE 21st

CENTURY

Changes in Climate Patterns,Variability, and Storms

Examination of the patterns of global-scale climatechange provides the broader context needed tounderstand the changes in storm tracks,precipita-tion belts,and other variations over the US. Suchanalyses can be undertaken because the simulationof large-scale natural variability by the climate mod-els is generally reasonable (e.g.,Stouffer et al.,2000). As for virtually all global models,both mod-els used in the National Assessment project that,incomparison to global average changes, warming athigh latitudes will be greater during winter andwarming of the land will be greater than of theocean (Figure 20). The dramatic wintertime warm-ing in high latitudes is very likely due to feedbacksinvolving the reduction in the reflectivity of the sur-face as sea ice melts and because weakening of thenear-surface inversion allows a relatively large tem-perature change to occur. Land warms more thanocean because of the oceans’ greater ability to limitand redistribute the trapped energy by evaporatingmoisture,mixing heat downward,transporting heataround by ocean currents,and the ocean’s largerheat capacity. In addition,the warming of land areasincreases as soil moisture is reduced,which reducesthe potential for evaporative cooling. Although not

shown in these figures,another robust feature ofglobal warming is greater warming at upper levelsof the tropical atmosphere. This warming occursbecause of the way that the vertical atmosphericstructure is determined through convection and theremoval of moisture with altitude. The upper atmos-phere warming affects how the atmospheric circula-tion changes,the generation and intensity of convec-tive rainfall (rainfall resulting from vertical motion in

53


Global Patterns of Surface TemperatureChanges - 21st Century

Figure 20: Global patterns of projected changes in surface tempera-ture (˚F) over the 21 st century [future (2090-2099) and modern (1961-1990)] for (a) December, January, February (DJF) from the Canadianmodel scenario, (b) DJF from the Hadley model scenario, (c) June,July, August (JJA) from the Canadian model scenario, and (d) JJAfrom the Hadley model scenario. See Color Plate Appendix.

Canadian Model Average Temperature (winter)

Hadley Model Average Temperature (winter)

>15

10

5

0

-5

>15

10

5

0

-5

Canadian Model AverageTemperature (summer)

Hadley Model Average Temperature (summer)

>15

10

5

0

-5

>15

10

5

0

-5

precipitation response. Other models (Meehl et al.,2000b) also show this type of response (Meehl andWashington,1996;Knutson and Manabe,1995,1998;Timmermann et al.,1999),although some modelsshow a La Niña-like (Noda et al.,1999),or an initialLa Niña-like,pattern that transitions into an El Niño-like pattern (Cai and Whetton,2000). This responseappears to be highly dependent upon how cloudfeedbacks are represented by the models,soremains quite uncertain (Meehl et al.,2000b).

The global precipitation anomalies (Figure 21) pro-jected by the models show increased precipitationcoinciding with the region of these warm anom-alies. The increased precipitation in the Southwestappears to be largely a result of the warmer SSTs inthe Pacific Ocean off the coast of North America.During winter, decreased precipitation along thenorthern branch of the Hadley Circulation (theatmospheric circulation with rising air near theequator and sinking air near 30˚ latitude, resultingin the trade winds and subtropical dry regions)extends over the eastern US in the Canadian modelscenario,but not in the Hadley model scenario.During summer, the Hadley model scenario shows alarge area of decreased precipitation in the easternPacific and Atlantic Oceans,whereas the Canadianmodel scenario projects decreased precipitationover land areas. Recent analyses of the 6-hourly datafrom the Hadley model indicate that the model isaccurately reproducing the Southwest monsoon dur-ing summer. In a simulation with increased green-house gases (although without sulfates),there areindications of a strengthening of the monsoon(Arritt et al.,2000),which correlates with the regionof increased summer precipitation in the Southwest.

Because the Northern Hemisphere’s atmosphericcirculation is more vigorous during winter, examin-ing the winter circulation pattern provides an indi-cation of the causes of these precipitation changes.The polar jet stream is known to be dependentupon both global and local temperature gradients.While the reduced pole-to-equator temperature gra-dient at the surface suggests a weaker or northward-shifted jet stream,the increased pole-to-equator tem-perature gradient in the upper troposphere suggeststhe reverse. The models calculate the relative influ-ence of each factor and provide a result that is aphysically and quantitatively consistent representa-tion of how temperatures,winds,and other atmos-pheric features might change in the future.

As shown in Figure 22,both models project thestrengthening and southward shift in the region ofmaximum upper atmospheric winds in the eastern

the atmosphere),and the development of tropicalstorms.

Model projected changes in regional temperaturesover the Pacific Ocean indicate greater warmingover the equatorial and northern East Pacific Ocean,and that these changes extend to the West Coast ofthe US in both models (Figure 20). This pattern ofwarming resembles an El Niño pattern of sea sur-face temperature (SST) anomalies,and so it wouldseem very likely to lead to an El Niño-like wind and

54


Global Precipitation Percent Differences - 21st Century

Figure 21: Global precipitation percent differences [(future - mod-ern)/modern) X 100] for (a) December, January, February (DJF) fromthe Canadian model scenario, (b) DJF from the Hadley model sce-nario, (c) June, July, August (JJA) from the Canadian model sce-nario, and (d) JJA from the Hadley model scenario. See Color PlateAppendix.

Canadian Model Precipitation Percent Dif ference (winter)

Hadley Model Precipitation Percent Dif ference (winter)

Canadian Model Precipitation Percent Difference (summer)

Hadley Model Precipitation Percent Dif ference (summer)

100%

80%

60%

40%

20%

0

-20%

-40%

-60%

-80%

-100%

100%

80%

60%

40%

20%

0

-20%

-40%

-60%

-80%

-100%

100%

80%

60%

40%

20%

0

-20%

-40%

-60%

-80%

-100%

100%

80%

60%

40%

20%

0

-20%

-40%

-60%

-80%

-100%

Pacific and across the West Coast (Sousounis,1999;Felzer, 1999). The changes in these winds,whichseem in the models to be a combination of thepolar and subtropical jets,are indicative of a deep-ened and southward-shifted Aleutian Low in bothmodels,especially for the Hadley model scenario.The Aleutian Low is a center for storms coming intoNorth America off the Pacific,so a deepening andsouthward-shift in the Aleutian Low would allowmore storms to penetrate further southwardtowards the California coast,helping to explain theprecipitation increases projected for that region.The projected weakening of the Pacific SubtropicalHigh (centered near Hawaii) would reduceupwelling of colder ocean waters,allowing SSTs torise and enabling more storms to penetrate into theSouthwest. Storm counts (Figure 23) confirm thatthe models are projecting more storms associatedwith the stronger Aleutian Low. Although theHadley model scenario shows a slight decrease instorms over the Southwest,there is more moisturein the atmosphere, resulting from the higher SSTs(Felzer and Heard,1999). As a result,the amount ofprecipitation is actually projected to increase. Othermodels,however, show reduced storm activity alongthe Pacific coast (Christoph et al.,1997),so thatthese results are apparently model dependent.

The region of storm formation off the East Coast ofthe US is locally dependent upon the land-sea tem-perature gradient. Warm Gulf Stream waters and acold land surface in winter provide ideal conditionsfor generating storms (e.g.,nor’easters). With warm-ing of the land surface,the land-sea contrast isreduced and the intensity of these storms could bereduced. The storms in the Hadley model scenariostart in the Mid-Atlantic region and track north andeast over the Atlantic Ocean;in contrast,the stormsin the Canadian model scenario track closely alongthe East Coast (Figure 23). Observations indicatethat present storm tracks extend along the south-eastern coast of the US (Klein,1957),so,in this par-ticular region,the storm tracks are better located inthe Canadian model scenario than in the Hadleymodel scenario,although they are over-representedto the south. Both models project a decrease in thenumber of storms along this predominant East Coaststorm track (Figure 23),although some individualstorms appear to be more intense (Felzer andHeard,1999;Carnell and Senior, 1998;Lambert,1995). Because of the different baseline positions ofthe storm tracks,however, the effect of the reducednumber of storms is felt over the US only in theCanadian model scenario (Felzer and Heard,1999).Note that the increase in the number of storms overEast Coast land areas in the Hadley model scenario

is probably not statistically significant because thereare very few storms there to begin with. A decreas-ing number of storms would be a change from thehistorical pattern,which does not show anydecrease in East Coast storms over the past 100years,but instead shows an increase during the1960s (Hayden,1999). A separate study of theresults from the Canadian model also indicates thata higher CO2 concentration will alter wintertimevariability and the behavior of the Arctic Oscillation,affecting primarily the North Atlantic and Europeanregions (Monahan et al.,2000). Other studies showan entire range of possibilities for storm changes inthe North Atlantic (Meehl et al.,2000b),includingmore intense storms (Lunkeit et al.,1996),lessintense storms (Beersma et al.,1997),and a shift instorm tracks towards the northeast with no changein intensity (Schubert et al.,1998).

Changes in the tracks of storms and jet streams mayalso be the result of changes in tropical circulationdue to changes in the model projections for the ElNiño-Southern Oscillation (ENSO). ENSO is present-ly a major cause of inter-annual variations in tropicaland global circulation. During warm ENSO events(El Niño),the waters in the eastern and centralequatorial Pacific Ocean warm, changing the atmos- 55


Figure 22: Schematic illustrating wintertime changes in the jetstream, pressure systems, sea surface temperatures, and stormtracks over and adjacent to North America. The Canadian andHadley model scenarios both show: a southward-shifted jet streamover the eastern Pacific and Southwest; a southward-shifted andintensified Aleutian Low and weakened subtropical High in theWest; and warmer ocean surface temperatures off the coast ofCalifornia. The Canadian model scenario also shows a reduction inthe number of storms along the East Coast storm track; however,the Hadley model scenario does not show this reduction nor did itdevelop this observed storm center in its control simulation. Formore details, see Sousounis (1999).

Wintertime Changes in Jet Stream and Atmospheric Circulation

Intensified Aleutian Low

Weakened Subtropical High

Current Position

Shifted Position in 2090’s

Warmer OceanTemperatures Reduced Number

of Storms

Index (SOI) do provide an indication of how atmos-pheric pressure patterns may shift.

Indices for the Niño-3 and 4 regions in the PacificOcean,which record changes in the SST, showENSO cycles continuing to occur in both models asthe world warms,although around a higher averageoceanic temperature (D. Legler and J. O’Brien,per-sonal communication;see http://www.coaps.fsu.edu/~legler/NAST/Assess_ENSO.html). Examining theSOI results,the Canadian model scenario projects ashift towards a more persistent set of conditionsthat is similar to an El Niño state (He and Barnston,personal communication),while the Hadley modelscenario shows no change. In neither case do thesemodels project a significant change in the frequencyor amplitude of ENSO variability (Collins,2000). Arecent study using a model with sufficient tropicalresolution to more accurately reproduce ENSO var i-

pheric and oceanic circulations in the Pacificregion,which affects global weather patterns andthe position of the jet stream over North America.The potential effects of global warming on ENSOare not yet established with confidence,in partbecause of the limited ability of GCMs to simulateENSO variations over the 20th century. However,both oceanic and atmospheric indices can be usedto provide some indications of the types of changesthe models are projecting. In particular, althoughthe findings must be considered uncertain,the NiñoSST-based indices and the Southern Oscillation56


Wintertime Storm Counts

Figure 23: Wintertime (DJF) storm counts (Carnell and Senior, 1998;Lambert, 1995) from the (a) Canadian model scenario (1901-1910total); (b) Hadley model scenario (1990-2110 mean from unforcedcontrol run); (c) Canadian model scenario (2091-2100 total); (d)Hadley model scenario (2070-2100 mean from transient run); (e)Canadian model scenario delta (c-a); and (f) Hadley model scenariodelta (d-b). Units are number of winter storms per 145,000 km2. SeeColor Plate Appendix.

Canadian Model - Storms Modern>8

7

6

5

4

3

2

1

0

Hadley Model - Storms Modern>8

7

6

5

4

3

2

1

0

Canadian Model - Storms Future>8

7

6

5

4

3

2

1

0

Hadley Model - Storms Future>8

7

6

5

4

3

2

1

0

Canadian Model - Storms Future - Modern % Hadley Model - Storms Future - Modern %100%

80%

60%

40%

20%

0

-20%

-40%

-60%

-80%

-100%

100%

80%

60%

40%

20%

0

-20%

-40%

-60%

-80%

-100%

ability has shown increased ENSO amplitude (ElNiños and La Niñas are both more intense) as aresult of greenhouse warming (Timmermann et al.,1999). Other studies (Meehl et al.,2000b),however,show little change (or even a slight reduction) inENSO amplitude (Knutson et al.,1997),while theHadley model scenario shows an increase in ampli-tude only after CO2 levels have been quadrupled(Collins,2000). Because there are several frequen-cies of variability within the ENSO signal (Meehl etal.,2000b;Zhang et al.,1997;Lau and Weng,1999;Allan et al.,1996;Knutson et al.,1997),it is often dif-ficult to determine how ENSO is changing, evenwith a century-long time series. Given these modelresults,the stronger Aleutian Low and weaker sub-tropical high (Trenberth and Hurrell,1994) thatboth Assessment models project over the Pacificseem likely to result from either the El Niño-likeresponse in the two models or from the warmENSO phase (El Niño) response evident in theCanadian model scenario.

Many of the precipitation changes in the GCMs,par-ticularly during summer when the atmospheric cir-culation is weaker, appear to be the result of feed-backs involving the land surface. During winter,snow cover is the mechanism for this interaction,while during summer, soil moisture is most impor-tant. As warming occurs over land areas, evapora-tion of available moisture increases and the soilmoisture decreases;as the land dries out,this soilmoisture leads to a decrease in overall evaporationand therefore of the amount of precipitable water inthe atmosphere. This decrease,in turn, results infewer clouds,less precipitation,and increased warm-ing,completing the positive feedback loop. Thus,

while increased warming over the ocean is project-ed to result in increased precipitation,the increasedwarming over land is projected to lead to less pre-cipitation because of the limited moisture-holdingcapacity of the land. Differences in model projec-tions of changes in precipitation over land duringsummer may therefore result from differences in therespective land surface models used in each GCM.Soil moisture trends generally correlate with precipi-tation anomalies during winter. Although soil mois-ture trends (Figure 18) during summer also corre-late with the precipitation anomalies,there are evenbroader areas of decreased soil moisture due to thelarge increases in evapotranspiration. For example,both models show drying in the western GreatPlains during summer. Another example is Alaska,where increases in evaporation due to increasedsummer temperatures are projected to lead todecreased soil moisture even though precipitationincreases (Felzer and Heard,1999).

Snow cover also plays an important role in winter-time changes in climate. Given the degree of warm-ing across the US,the models project that the extentof snow cover is very likely to be significantlyreduced (Figure 24). As the snow line retreats pole-ward,a larger surface area is exposed to a loweralbedo surface,which increases the amount ofwarming,creating a large positive feedback. Whileboth the Canadian and Hadley models show thesnowline retreating towards the end of the 21st cen-tury, the reduction in snow cover over the US is pro-jected to be particularly dramatic in the Canadianmodel,where mean wintertime snow cover existsonly in the northern Rocky Mountains and northernGreat Plains. Although snow cover still remains in

57



Winter Average Snow Cover Difference - 2090s

Figure 24: Projections across the US of the decrease in winter average snow accumulation (inches) from 1961-1990 to 2090-2099based on results from the (a) Canadian model scenario and (b) Hadley model scenario. In these diagrams, the changes in snow depthcalculated as differences in the water equivalent of snow in kg/m2 have been converted to depth of dry snow (in inches), assuming a15 to 1 average ratio of snow depth to water equivalent (Judson and Doesken, 2000). See Color Plate Appendix.

>0”

-10

-20

-30

-40

-50

-60

-70

-80

humidity are projected to increase,as is the hydro-logic cycle of evaporation and precipitation. Manymodels project a greater frequency of extreme hightemperatures and decrease in frequency of extremelow temperatures as a result of increased green-house gas concentrations (Giorgi et al.,1998;Meehlet al.,2000b). Increased daily temperature variabili-ty in summer and decreased daily temperature vari-ability in winter is also likely (Mearns et al.,1995;Gregory and Mitchell,1995;Zwiers and Kharin,1998;Meehl et al.,2000b). As observed (Karl andKnight,1998) and modeled (Meehl et al.,2000b),reduced diurnal temperature range may result froma greater increase in minimum temperatures thanmaximum temperatures. Both observations (Gaffenand Ross,1998) and model results (Delworth et al.,1999),including the current scenarios (Figure 15),show an increase in the heat index,which is ameasure of the discomfort level due to warming.

Trends in one-day and multi-day precipitationevents over the US and other countries show anincrease in the number of days with the heaviestamounts of precipitation (Karl and Knight,1997,1998). The number of days annually with precipita-tion exceeding 2 inches (about 5 cm) has beenincreasing in the US (Karl et al.,1995a) and the fre-quency of the highest 1- to 7-day precipitationtotals has also been increasing (Kunkel et al.,1999).Increases have been largest for the Southwest,Midwest,and Great Lakes regions of the US.Projections from the Hadley and Canadian modelscenarios show an increase of heavy precipitationevents as the climate warms (Figure 25). Hulme etal.(1998) found some agreement between the pro-jected precipitation changes and recently observedtrends. In reviewing the Canadian model results,Zwiers and Kharin (1998) found that extreme tem-perature and precipitation events are very likely tooccur more frequently. Many other modeling stud-ies also show an increase in the heaviest precipita-tion events (Meehl et al.,2000b; Kothavala,1997;Hennessy et al.,1997;Durman et al.,2000;Giorgi etal.,1998). Studies on changes in climate extremesare summarized in recent workshop proceedings(Karl and Easterling,1999;AGCI,1999;Easterling etal.,2000b) and in Meehl et al.(2000b). Summerdrying in mid-continental regions due to increasedevaporation,sometimes coupled with decreasedprecipitation,has also been projected in many mod-els (Haywood et al.,1997;Gregory et al.,1997;Wetherald and Manabe,1999;Meehl et al.,2000b).

Studies with the GFDL hurricane model (Knutsonet al.,1998;Knutson and Tuleya,1999) also suggestthat the rate of precipitation during tropical storms

the northern Rocky Mountains,both models projectthat the amount of snow in this region will be dra-matically reduced. The sharply reduced extent ofsnow cover in the Canadian model scenario,whichmay have been initiated by the more zonal flowconditions leading to higher wintertime tempera-tures and fewer outbreaks of Arctic storms,allowsmore absorption of solar radiation,especially earlyin the year. This effect further diminishes snowcover and increases the warming in the Canadianmodel simulation.

Climate Extremes

While changes in average conditions and inter-annu-al variability are expected to have significant effectsfor some ecosystems and some parts of the econo-my, other parts of the economy are projected to beaffected more by potential changes in the frequencyor intensity of extreme events. It is likely that thefrequency of occurrence of exceeding certainthresholds and the intensity of extreme eventsmight change because temperatures and absolute

58


Hadley Model

Canadian Model

Figure 25: Bar chart showing projected changes in frequency ofvarious types of precipitation. Both the (a) Canadian and (b) Hadleymodel scenarios project increases in the frequency of heavy precip-itation events, intensifying the trend observed for the 20th century.Figure prepared by Byron Gleason of the National Climatic DataCenter based on the methods described in Karl and Knight (1998).

Projected Changes in Intensity of National Daily Precipitation

could increase due to the warmer conditions andthe increased amount of water vapor in the atmos-phere. Other studies confirm these results(Krishnamurti et al.,1998;Walsh and Ryan,1999;Meehl et al.,2000b). Two additional studies show adecrease in the frequency of hurricanes as a resultof global warming (Bengtsson et al.,1996;Yoshimura et al.,1999). Both the Canadian andHadley model scenarios project an increase ofheavy precipitation events as the climate warms.Ultimately there is a strong dependence of hurri-canes on ENSO (Meehl et al.,2000b;Knutson et al.,1998;Knutson and Tuleya,1999),indicating thathow ENSO changes is likely to be an importantindicator of how hurricanes will vary, especially forthe southeastern US.

Precipitation is the driving factor affecting stream-flow (Langbein,1949;Karl and Reibsame,1989) sothe observed and projected increase in the intensi-ty and frequency of heavy (the upper 5% per-centiles of all precipitation events) and extremeprecipitation (the highest annual 1-day precipita-tion events) have the potential to increase inlandflooding. Higher temperatures,conversely, have thepotential for exacerbating drying of the soil and,over time,of increasing drought frequency andintensity. Separating these two influences is chal-lenging. Nonetheless,analyses of changes indrought frequency and intensity (Karl et al.,1995a)reveal no trend in drought frequency, but they doreveal an increase in the area affected by severeand extreme moisture surplus. Streamflow dataanalyzed by Lins and Slack (1999) also reveal anincrease in low-stream flows,adding more confi-dence to the notion that drought frequency andintensity has not become more severe,despite theincrease in US average temperature. On the otherhand,Lins and Slack (1999) do not find an unusualnumber of statistically significant increases ofstreamflow, despite the fact that Karl and Knight(1998) show statistically significant increases ofprecipitation,including heavy and extreme events.New analyses indicate a strong relation betweenmulti-decadal increases in heavy and extreme pre-cipitation events and high and low streamflows,butwith considerable variability (Groisman et al.,1999,2000). These results indicate that part of this vari-ability is related to reductions in snow cover extentin the West,which have modified the peak streamflows and ameliorated the ef fect of increased heavyprecipitation. In these results,Groisman et al.(2000) find that,when averaged across watershedsand across the country, a clear relationshipbetween heavy precipitation and high streamflowevents emerges.

THE CLIMATIC EFFECTS OFSTABILIZING THE CARBONDIOXIDE CONCENTRATIONThe objective of the Framework Convention onClimate Change (FCCC),of which over 160 coun-tries including the US are signatories,is to stabilizethe atmospheric concentration of greenhouse gases“at a level that would prevent dangerous anthro-pogenic interference with the climate system.”Precise goals for stabilization of the CO2 concentra-tion have not been established. To provide informa-tion for the negotiating process,the IPCC consid-ered stabilization at concentrations of 350,450,550,650,750,and 1000 ppmv, plus a variety of temporalpathways to reach these goals (Wigley et al.,1997).Many alternative carbon emission and CO2 concen-tration pathways have been evaluated for achievingstabilization at 550 ppmv, which represents anapproximate doubling of the pre-industrial CO2 con-centration. Different end points and emission path-ways arise from different assumptions about thespeed at which emissions can or will be reducedbased on views about feasibility or optimality ofpolicies,measures,and technological changes.

To provide an estimate of the reduction in climatechange that might occur with CO2 stabilization,theNAST asked the National Center for AtmosphericResearch (NCAR) to use new models to carry outspecial simulations to provide an indication of thesize of the climatic change that would result fromstabilizing the CO2 concentration. NAST and NCARscientists chose to examine the climatic conse-quences of a reduced emission growth scenarioinvolving eventual stabilization at 550 ppmv. Thisemissions path would allow continued growth inemissions for a few decades into the 21st century,followed by rapid decreases in emissions. While thisemission path is a plausible alternative for investiga-tion of potential climatic impacts,it should not beinterpreted as the only way to achieve stabilization,as a prediction of what is most likely to happen,oras a preferred policy alternative. Reductions in theprojected warming would be greater from scenariosthat begin reducing emissions earlier in the 21st cen-tury than is assumed in the stabilization scenarioused here.

To carry out these simulations,two different climatemodels were used (Boville and Gent,1998;Washington et al.,2000). Having the results of onlyone modeling group (albeit with two similar mod-els) is somewhat limiting,especially because the 59


would occur in the 22nd century, modest effects dobecome apparent in the latter half of the 21st centu-ry (or could occur earlier if earlier actions are takento reduce the rate of rise of emissions). As indicatedin Figure 26,if the emissions pathways were tooccur as projected,global and US average tempera-tures would likely continue to rise significantly dur-ing the 21st century, even if actions were taken start-ing in the near future to limit the growth of emis-sions in order to move toward stabilizing atmos-pheric concentrations at 550 ppmv20. Basically, theNCAR results (Dai et al.,2001) suggest that, even ifsuch actions are taken,the warming in 2100 wouldstill likely be several degrees Fahrenheit (and othermore responsive models would suggest even more).With movement toward stabilization,the warming isprojected to be about half a degree Fahrenheit less(or about 10-15% lower) in 2100 than for the “no cli-mate policy”scenario used in the Hadley andCanadian models used in the Assessment. Figure 27shows the effects of the move toward stabilizationon temperature and precipitation patterns over theUS. In these model simulations (and other simula-tions may give different effects),the emissions cut-back begins to reduce the warming across thesouthern US and to change the resulting precipita-tion pattern slightly.

It should also be noted that the reduction in therate of rise of the CO2 concentration itself wouldalso have important effects. For forests and agricul-ture,increased CO2 stimulates growth and improveswater use efficiency under a range of conditions,sothat a lesser rise in the CO2 concentration wouldlikely reduce the increase in crop production andgrowth of natural biomass (see chapters onAgriculture and Forests) as well as reduce climaticstress on the various ecosystems. For coral reefs,the acidifying effects of CO2 cause reduced alkalini-ty of ocean waters, reducing calcification and weak-ening corals;therefore,limiting the rate of CO2

increase would help to ameliorate this situation (seeCoastal chapter and Kleypas et al.,1999).

CRUCIAL UNKNOWNS ANDRESEARCH NEEDSWhile much has been learned about the types of cli-mate changes that could occur over the 21st centuryas atmospheric concentrations of CO2 increase,much remains to be learned,especially about howthe variability and extremes of the climate willchange. Although the similarities in how the

baseline simulation reported here was not the samebaseline used in the Canadian and Hadley modelscenarios. However, these calculations do provideinteresting insights19. Similar stabilization runs havenow also been completed by the Hadley Centre(Mitchell et al.,2000).

While most of the differences in climatic conditionsthat would result from moving toward stabilization

60


Figure 26: Comparison of the time history of the increase in annual-average surface temperature for (a) the globe and (b) the US as pro-jected by two related models developed at the National Center forAtmospheric Research for an emission scenario where the green-house gas concentrations are allowed to rise without restriction(baseline) and for a case (stabilization) where steps are taken tolimit the rise in the CO 2 concentration to 550 ppmv (Dai et al., 1999;Washington et al., 2000). Results are also shown for a recentHadley model simulation (Mitchell et al., 2000). See Color PlateAppendix.

Global Mean Temperature Anomolies

US Mean Temperature Anomolies

a

b

19Detailed results from these model simulations are available athttp://www.nacc.usgcrp.gov/scenarios/.

Canadian and Hadley models represent changes inglobal scale features are encouraging,the differ-ences between their results on regional scales sug-gest that significant uncertainties remain. For exam-ple, even though the Canadian model scenario pro-duces a reasonable response to El Niño occurrencesacross North America,problems with the way theseGCMs simulate ENSO variability suggest that theprojected pattern of changes may not be definitive.Also,as illustrated by the different projections ofchanges in summer precipitation in the Southeast,there are often several processes that contribute tothe pattern of change that is seen,and these mayprogress differently. As illustrated by the discussionabout changes in storm tracks,often the sameprocess can lead to different projections of changeswhen imposed on a slightly different base state ofthe climate. In addition,the different representa-tions of land surface processes (as well as otherparameterizations) included in different GCMs canhave an important impact on projections of changesin regional precipitation. This dependence occursbecause precipitation,unlike atmospheric dynamics,is a highly localized feature of the climate,depend-ing on the interaction of many processes,some ofwhich are still represented in quite schematic ways.Given these many limitations,it is important to men-tion again that the model projections are not predic-tions,but that they instead should be viewed asinternally consistent scenarios of climatic changesthat might occur over the 21st century. As a result,they can,as indicated earlier, only provide indica-tions of the types of consequences that mightresult.

To build confidence in the projections, muchremains to be done. Further improvements in cli-mate models are needed,especially in the represen-tations of clouds,aerosols (and their interactionswith clouds),sea ice, hydrology, ocean currents,regional orography, and land surface characteristics.Improving projections of the potential changes inatmospheric concentrations of greenhouse gasesand aerosols is underway under the auspices of theIPCC (IPCC,2000) and model simulations based onthese revised emissions forecasts are expected toprovide improved estimates of future change. Inaddition to having results from more models avail-

61


Global Patterns of Surface TemperatureChanges - 21 st Century

Figure 27: Patterns across the US of projected changes in thetrends of annual mean surface temperature and precipitation for the21st century assuming an emissions profile that moves toward sta-bilization of the CO2 concentration at 550 ppmv in the 22nd century(STA) as compared to the baseline case (roughly case IS92a, orBAU, except projections in sulfur emissions are reduced in the CSMscenario). The projected differences in the changes that wouldgenerally be projected (case STA minus BAU) are based on resultsfrom: (a) NCAR CSM for annual mean temperature; (b) PCM forannual mean temperature: (c) NCAR CSM annual average monthlyprecipitation; and (d) PCM annual average monthly precipitation.Temperature trend differences are given as ˚F per 100 years.Precipitation trend differences are given in percent, with bothtrends calculated using a 1980-1999 baseline. Trends are derivedbased on a linear regression through each grid point. Results aredescribed in Dai et al. (1999) and Washington et al. (2000). SeeColor Plate Appendix.

CSM STA-BAU Average Annual TemperatureTrend

PCM STA-BAU Average Annual Temperature Trend

>1

.75

.50

.25

.00

-.25

-.50

-.75

-1

>1

.75

.50

.25

.00

-.25

-.50

-.75

-1

CSM STA-BAU Average Annual Precipitation Trend

PCM STA-BAU Average Annual Precipitation Trend

100%

80%

60%

40%

20%

0

-20%

-40%

-60%

-80%

-100%

100%

80%

60%

40%

20%

0

-20%

-40%

-60%

-80%

-100%

20Stabilizing the atmospheric CO2 concentration at 550 ppmv over the21st century would require keeping global average per capita emissionsof CO2 at roughly their present level of 1 tonne of carbon per year asglobal population increases by about 50% and developing nations raisetheir energy levels to enhance their standard-of-living. Accomplishingthis would require that all energy needs for the growing populationwould have to be met by an appropriate combination of reducing CO2emissions (e.g.,through higher efficiencies,use of less carbon intensivefuels,etc.), switching to energy sources not based on fossil fuels (e.g.,wind,solar, hydro,biomass, nuclear, etc.),providing more efficient ener-gy services,or reducing the emissions of other greenhouse gases.

affect us will provide important information on howwe will need to adapt to such changes.Use of historical climate data provides one basis forexploring the environment and society’s vulnerabili-ty to a changing climate. Records of the US climateindicate that the climate is apparently starting tochange in a manner consistent with the observedglobal-scale changes. These records also indicatethat there have been significant variations in the cli-mate,which have in turn had important effects onagriculture,water resources,and public health.There is no reason to believe from the historical orpaleoclimatic record or from model results thatsuch changes will not recur in the future. To help inanalyzing societal vulnerability to ongoing climatevariations,a range of historical information aboutthe climate of the 20th century has been providedfor the Assessment.

To explore how future climate may be affected bythe rising concentrations of greenhouse gases,model simulations have been used to provide quan-titative estimates. Although emission scenarios areuncertain and model simulations are still imperfect(e.g.,due to limitations in the representations ofimportant processes and feedbacks),two sets ofmodel results have been assembled to provide plau-sible projections of how conditions may changeover the US during the 21st century. These modelsproject quite significant warming across the US andsubstantial stress on water resources in severalregions. While the particular sets of model resultsdo not fully bound all of the possible futures,theydo provide a range of possible future conditions thatcan be used to start to explore the potential conse-quences of climate change for the US.

The results available for this Assessment thus pro-vide the basis for the most complete analysis yetundertaken and point to pathways for future analy-sis and research. Current understanding clearly indi-cates that the climate is changing and is very likelyto change significantly more in the future. At thesame time, much more work is needed over thecoming years to improve global and mesoscale pro-jections of future changes in CO2 concentration andof climate,to improve simulation of climate variabili -ty, to develop the means to project changes inextreme events,and to expand the statistical analysisand interpretation of existing and planned modelsimulations. Practical efforts should continue at thecommunity level to interpret the scientific data andclimate scenarios with the aim of providing usableinformation that integrates current needs with plan-ning for future community development.

able,ensembles of simulations from several modelruns are needed so that the statistical significance ofthe projections can be more fully examined. As partof these efforts,it is important to develop greaterunderstanding of how the climate system works(e.g.,of the role of atmosphere-ocean interactionsand cloud feedbacks),to refine model resolution,tomore completely incorporate existing knowledgeinto climate models,to more thoroughly test modelimprovements,and to augment computational andpersonnel resources in order to conduct and fullyanalyze a wider variety of model simulations,includ-ing mesoscale modeling studies.

While much remains to be done that will take signif-icant time, much can also be done at present toimprove the use and understanding of potential cli-mate change scenarios. For example,an intensifiedanalysis program is needed to provide greater under-standing of the changes and the reasons they occur.New efforts to examine the synoptic patterns of thechanges in the global models were started at thenational level in the analyses presented here. Sucheffort are also starting through region-specific stud-ies that combine analysis of the model results withthe insights available from analysis of historical cli-matology and past weather patterns (i.e.,synopticconditions). For example,Risbey et al.(1999) haveconstructed regional climate scenarios for two studyregions in North America (Chesapeake Bay andOklahoma/Great Plains) using a combination ofGCM output and dynamical reasoning. Otherapproaches being pursued involve use of mesoscalemodels that provide higher resolution of spatial con-ditions even though they can only provide simula-tions of shorter periods of time and smaller spatialscales.

SUMMARYThere are clear indications that the atmosphericconcentrations of CO2 and other greenhouse gasesand aerosols are being increased by human activi-ties. These changes in atmospheric composition,combined with the influences of other natural andhuman-induced changes,are changing the climate.Available model simulations,combined with ourunderstanding of the factors that have changed theEarth’s climate in the past,provide clear evidencethat global warming is occurring and that additionalwarming will result. There is clear evidence on theglobal scale that this warming is occurring,with thewarming during the 20th century in reasonableaccord with model simulations. Because climateaffects the environment and many of our naturalresources,considering how such changes might62


LITERATURE CITEDAGCI (Aspen Global Change Institute), Elements ofChange 1998, edited by S. J. Hassol and J. Katzenberger,Aspen Colorado,1999.

Alcamo, J.,E.Kreileman,M.Krol,R.Leemans, J. Bollen, J.Van Minnen,M.Schaeffer, S.Toet,and B.De Vries,Globalmodelling of environmental change: An overview ofImage 2.1.,pp.3-94 in Global Change Scenarios of the21st Century: Results from the Image 2.1 Model , edit-ed by J.Alcamo,R.Leemans,and E.Kreileman,Pergamon Press,Oxford,United Kingdom,1998.

Allan,R. J., J. Lindesay, and D. Parker, El Niño SouthernOscillation and Climate Variability, CSIROPublications,Melbourne,Australia,405 pp.,1996.

Andres,R. J., G. Marland,T. Boden,and S.Bischoff,Carbon dioxide emissions from fossil fuel combustionand cement manufacture,1751 to 1991,and an esti-mate for their isotopic composition and latitudinal dis-tribution,in The Carbon Cycle, edited by T. M.L.Wigleyand D. Schimel,Cambridge University Press,Cambridge,United Kingdom,312 pp.,2000.

Angell, J. K.,Difference in radiosonde temperaturetrend for the period 1979-1998 of MSU data and theperiod 1959-1998 twice as long, Geophysical ResearchLetters, 27, 2177-2180,2000.

Arrhenius,S.,On the influence of carbonic acid in theair upon the temperature of the ground, PhilosophicalMagazine, 41, 237,1896.

Arritt,R.W., D. C.Goering,and C. J.Anderson,The NorthAmerican monsoon system in the Hadley Centre cou-pled ocean-atmosphere GCM, Geophysical ResearchLetters, 27, 565-568,2000.

Barnett,T. P. K.,et al.,Detection and attribution ofrecent climate change: A status report, Bulletin of theAmerican Meteorological Society, 80, 2631-2659,1999.

Beersma, J. J.,K.M.Rider, G. J. Komen,E.Kaas,and V.V.Kharin,An analysis of extratropical storms in the NorthAtlantic region as simulated in a control and 2 x CO2

time-slice experiment with a high-resolution atmos-pheric model, Tellus, 49A, 347-361,1997.

Bengtsson,L.,M.Botzet,and M.Esch,Will greenhousegas-induced warming over the next 50 years lead tohigher frequency and great intensity hurricanes?,Tellus, 48A, 57-73,1996.

Berger,A.,Long-term variation of daily insolation andQuaternary climatic changes, Journal AtmosphericSciences, 35, 2362-2367,1978.

Berger,A.,The role of CO2, sea-level and vegetation dur-ing the Milankovitch forced glacial-interglacial cycles,in Geosphere-Biosphere Interactions and Climate,edited by L.Bengtsson,1999.

Berger,A.,and M. F. Loutre,Insolation values for the cli-mate of the last 10 million years, Quaternary ScienceReviews, 10, 297-317,1991.

Berger,A.,M. F. Loutre,and J. L.Melice, The 100 kyrperiod in the astronomical forcing, Scientific Report1999/6, Institut D’Astronomie et de Geophysique G.Lemaitre,Université Catholique de Louvain,1999.

Boer, G. J., G. M.Flato,and D. Ramsden,A transient cli -mate change simulation with historical and projectedgreenhouse gas and aerosol forcing: Projected climate

for the 21st century, Climate Dynamics, 16, 427-450,2000a.

Boer, G. J., G. M.Flato,M.C.Reader, and D. Ramsden,Atransient climate change simulation with historical andprojected greenhouse gas and aerosol forcing:Experimental design and comparison with the instru-mental record for the 20th century, Climate Dynamics,16, 405-425,2000b.

Boer, G. J., N.A.McFarlane,R.Laprise, J. D. Henderson,and J. P. Blanchet,The Canadian Climate Centre spectralatmospheric general circulation model, Atmosphere-Ocean, 22(4),397-429,1984.

Bove,M.C., J. B.Elsner, C.W. Landsea,X.Niu,and J. J.O’Brien,Effect of El Niño on US landfalling hurricanes,revisited, Bulletin of the American MeteorologicalSociety, 79, 2477-2482,1998.

Boville,B.A.,and P. R.Gent,The NCAR Climate SystemModel,Version One, Journal of Climate, 11(6),1115-1130,1998.

Cai,W.,and P. H.Whetton,Evidence for a time-varyingpattern of greenhouse warming in the Pacific Ocean,Geophysical Research Letters, 27, 2577-2580,2000.

Callendar, G. S.,The artificial production of carbondioxide and its influence on temperature, QuarterlyJournal of the Royal Meteorological Society, 64, 223,1938.

Carnell,R.E.,and C.A.Senior, Changes in mid-latitudevariability due to increasing greenhouse gases and sul-phate aerosols, Climate Dynamics, 14, 369-383,1998.

CDIAC (Carbon Dioxide Information Analysis Center),Oak Ridge National Laboratory, 2000.(Available athttp://cdiac.esd.ornl.gov:80/cdiac/)

Christoph,M., U. Ulbrich,and P. Speth,Midwinter sup-pression of Northern Hemisphere storm track activityin the real atmosphere and in GCM experiments,Journal of the Atmospheric Sciences, 54, 1589-1599,1997.

63


Easterling, D. R.,Variability and trends in temperaturethreshold exceedances and frost dates in the UnitedStates, Bulletin of the American MeteorologicalSociety, in review, 2000.

Easterling, D. R.,T. R.Karl,E.H.Mason, P.Y. Hughes, D. P.Bowman,R.C.Daniels,and T.A.Boden,United StatesHistorical Climatology Network (USHCN) MonthlyTemperature and Precipitation Data,Publication No.4500,Carbon Dioxide Information and Analysis Center,Oak Ridge National Laboratory, Oak Ridge,Tennessee,83 pp.with appendices,1996.

Easterling, D. R.,et al.,Maximum and minimum temper-ature trends for the globe, Science, 277, 364-367,1997.

Easterling, D. R., J. L.Evans, P.Ya.Groisman,T. R.Karl,K.E. Kunkel,and P.Ambenje,Observed variability andtrends in extreme climate events: A review, Bulletin ofthe American Meteorological Society, 81, 417-425,2000a.

Easterling, D. R., G.A.Meehl,C. Parmesan,S.A.Changnon,T. R.Karl,and L. O. Mearns,Climateextremes:Observations,modeling,and impacts,Science, 289, 2068-2074,2000b.

Etheridge,M.,L. P. Steele,R.L.Langenfelds,R. J. Francey,J.-M.Barnola,and V. I.Morgan,Historical CO2 recordsfrom the Law Dome DE08,DE08-2,and DSS ice cores,in Trends: A Compendium of Data on GlobalChange, Carbon Dioxide Information Analysis Center,Oak Ridge National Laboratory, US Department ofEnergy, Oak Ridge Tennessee,1998.

Felzer, B.,Hydrological implications of GCM results forthe U.S.National Assessment,pp.69-72 in Proceedingsof the Specialty Conference on PotentialConsequences of Climate Variability and Change toWater Resources of the United States, May 10-12,1999,Atlanta, Georgia,American Water ResourcesAssociation,Middleburg,Virginia,1999.

Felzer, B.,and P. Heard,Precipitation differencesamongst GCMs used for the U.S.National Assessment,Journal of the American Water Resources Association,35(6),1327-1339,1999.

Flato, G. M., G. J. Boer,W. G. Lee, N.A.McFarlane, D.Ramsden,M.C.Reader, and A. J.Weaver,The CanadianCentre for Climate Modeling and Analysis global cou-pled model and its climate, Climate Dynamics, 16, 451-467,2000.

Gaffen, D. J.,and R. J. Ross,Increased summertime heatstress in the U.S., Nature, 396, 529-530,1998.

Gates,W. L.,et al.,An overview of the results of theAtmospheric Model Intercomparison Project (AMIP I),Bulletin of the American Meteorological Society, 80,29-55,1999.

Clark, P. U.,R.B.Alley, and D. Pollard,NorthernHemisphere ice-sheet influences on global climatechange, Science, 286, 1104-1111,1999.

COHMAP (Climates of the Holocene Mapping Project),Climatic changes of the last 18,000 years: Observationsand model simulations, Science, 241, 1043-1052,1988.

Collins,M.,The El-Niño Southern Oscillation in the sec-ond Hadley Centre coupled model and its response togreenhouse warming, Journal of Climate, 13, 1299-1312,2000.

Conway,T. J., P. P.Tans,and L.S.Waterman,AtmosphericCO2 records from sites in the NOAA/CMDL air samplingnetwork,in Trends ‘93: A Compendium of Data onGlobal Change, edited by T.A.Boden,et al.,ORNL/CDIAC-65,Carbon Dioxide Information AnalysisCenter, Oak Ridge National Laboratory, US Departmentof Energy, Oak Ridge Tennessee,1994.

Crowley,T. J.,Causes of climate change over the past1000 years, Science, 289, 270-277,2000.

Crowley,T. J.,and G. R.North, Paleoclimatology, OxfordUniversity Press,New York,339 pp.,1991.

Dai,A.,K.E.Trenberth,and T. R.Karl,Effects of clouds,soil moisture,precipitation,and water vapor on diurnaltemperature range, Journal of Climate, 12, 2451-2473,1999.

Dai,A.,T. M.L.Wigley, B.A.Boville, J.T. Kiehl,and L.E.Buja,Climates of the 20th and 21st centuries simulatedby NCAR Climate System Model, Journal of Climate,14,485-519,2001.

Darwin,R. F.,World agriculture and climate change:Current questions, World Resource Review, 9, 17-31,1997.

Delworth,T. L., J. D. Mahlman,and T. R.Knutson,Changesin heat index associated with CO 2-induced global warm-ing, Climatic Change, 43, 369-386,1999.

Doherty, R.,and L. O. Mearns,A comparison of simula-tion of current climate from two coupled atmosphere-ocean global climate models against observations andevaluation of their future climates,National Institute forGlobal and Environmental Change (NIGEC),Boulder,Colorado,1999.

Dresler, P.V.,M.C.MacCracken,and A. Janetos,Nationalassessment of the potential consequences of climatevariability and change for the United States, WaterResources Update, 112, 16-24,1998.

Durman,C. F., J. M.Gregory, D. C.Hassell,and R. G. Jones,The comparison of extreme European daily precipita -tion simulated by a global and a regional climate modelfor present and future climates,Quarterly Journal of theRoyal Meteorological Society, in press,2000.

64


Giorgi, F.,L.Mearns,C.Shields,and L.McDaniel,Regional nested model simulations of present day and2 x CO2 climate over the central Great Plains of theUnited States, Climatic Change, 40, 457-493,1998.

Goody, R.M.,and Y. L.Yung, Atmospheric Radiation,Oxford University Press,Oxford,United Kingdom,1989.

Gregory, J. M.and J. F. B.Mitchell,Simulation of dailyvariability of surface temperature and precipitationover Europe in the current and 2 x CO2 climates usingthe UKMO climate model, Quarterly Journal of theRoyal Meteorological Society, 121, 1451-1476,1995.

Gregory, J. M., J. F. B.Mitchell,and A. J. Brady, Summerdrought in northern midlatitudes in a time-dependentCO2 climate experiment, Journal of Climate, 10, 662-686,1997.

Gregory, J. M.,and J. Oerlemans,Simulated future sea-level rise due to glacier melt based on regionally andseasonally resolved temperature changes, Nature, 391,474-476,1998.

Groisman, P.Ya.,et al.,Changes in the probability ofheavy precipitation: Important indicators of climaticchange, Climatic Change, 42, 243-283,1999.

Groisman, P.Ya.,R.W. Knight,and T. R.Karl,Heavy pre-cipitation and streamflow in the United States: Trendsin the 20th century, Bulletin of the AmericanMeteorological Society, in press, 2000.

Hansen, J., D. Johnson,A.Lacis,S.Lebedeff, P. Lee, D.Rind,and G. Russell,Climatic impact from increasingcarbon dioxide, Science, 213, 957-966,1981.

Hansen, J.,M.Sato,A.Lacis,R.Ruedy, I.Gegen,and E.Matthews,Climate forcings in the industrial era,inProceedings of the National Academy of Sciences, 95,12753-12758,1998.

Hayden,B. P.,Extratropical storms: Past,present,andfuture,pp.93-96 in Proceedings of SpecialtyConference on Potential Consequences of ClimateVariability and Change to Water Resources of theUnited States, May 1999,American Water ResourcesAssociation,1999.

Haywood, J. M.,R. J. Stouffer, R.T.Wetherald,S.Manabe,and V. Ramaswamy,Transient response of a coupledmodel to estimated changes in greenhouse gas and sul-fate concentrations, Geophysical Research Letters,24(11),1335-1338,1997.

Hegerl, G. C.,K.Hasselmann, U. Cubasch, J. F. B.Mitchell,E.Roeckner, R.Voss,and J.Waszkewitz,Multi-fingerprintdetection and attribution of greenhouse gas-andaerosol-forced climate change, Climate Dynamics, 13,613-634,1997.

Hengeveld,H. G.,Projections for Canada’s ClimateFuture,Special report CCD-00-01,Environment Canada,Downsview, Ontario,Canada,2000.

Hennessy, K. J., J. M.Gregory, and J. F. B.Mitchell,Changes in daily precipitation under enhanced green-house conditions, Climate Dynamics, 13, 667-680,1997.

Houghton,R.A.,Land-use change and the carbon-cycle,Global Change Biology, 1, 275-287,1995.

Houghton,R.A.,and J. L.Hackler, Continental scale esti-mates of the biotic carbon flux from land coverchange,1850-1980,ORNL/CDIAC-79,NDP-050,OakRidge National Laboratory, 144 pp.,1995.

Hoyt, D.V.,and K.H.Schatten,A discussion of plausiblesolar irradiance variations: 1700-1992, Journal ofGeophysical Research, 98, 18895-18906,1993.

Huang,S.,H. N. Pollack,and P.Y. Shen,Temperaturetrends over the past five centuries reconstructed fromborehole temperatures, Nature, 403, 756-758.

Hulme,M.,T. J. Osborn,and T. C. Johns,Precipitationsensitivity to global warming:Comparison of observa-tions with HadCM2 simulations, Geophysical ResearchLetters, 25(17),3379-3382,1998.

Hyde,W.T.,and T. J. Crowley, Probability of future cli-matically significant volcanic eruptions, Journal ofClimate, 13, 1445-1450,2000.

Imbrie, J.,et al.,On the structure and origin of majorglaciation cycles 1.Linear responses to Milankovitchforcing, Paleoceanography, 7, 701-738,1992.

Imbrie, J.,et al.,On the structure and origin of majorglaciation cycles 2.The 100,000 year cycle,Paleoceanography, 8, 699-735,1993.

Imbrie, J.,A.McIntyre,and A.Mix,Oceanic response toorbital forcing in the late Quaternary: Observationaland experimental strategies,pp.121-164 in Climateand the Geosciences , edited by A.Berger et al.,KluwerAcademic Publishers,Boston,Massachusetts,1989.

Indermuehle,A.,et al.,Holocene carbon-cycle dynamicsbased on CO2 trapped in ice at Taylor Dome,Antarctica,Nature, 398, 121-126,1999.

IPCC (Intergovernmental Panel on Climate Change),Climate Change: The IPCC Scientific Assessment, edit-ed by J.T. Houghton, G. J. Jenkins,and J. J. Ephraums,Cambridge University Press,Cambridge UnitedKingdom,365 pp.,1990.

IPCC (Intergovernmental Panel on Climate Change),Climate Change 1992: The Supplementary Report tothe IPCC Scientific Assessment, edited by J.T.Houghton,B.A.Callander, and S.K.Varney, CambridgeUniversity Press,Cambridge,United Kingdom,200 pp.,1992.

65


Karl,T. R.,and R.W. Knight,The 1995 Chica go heatwave: How likely is a recurrence?, Bulletin of theAmerican Meteorological Society, 78, 1107-1119,1997.

Karl,T. R.,and R.W. Knight,Secular trends of precipi-tation amount,frequency, and intensity in the UnitedStates, Bulletin of the American MeteorologicalSociety, 79, 231-241,1998.

Karl,T. R.,and W. E.Reibsame,The impact of decadalfluctuations in mean precipitation and temperatureon runoff: A sensitivity study over the United States,Climatic Change, 15, 423-447,1989.

Karl,T. R.,and K.Trenberth,The human impact on cli-mate, Scientific American, 281, 100-105,1999.

Karl,T. R.,H. F. Diaz,and G. Kukla,Urbanization: Itsdetection and effect in the United States climaterecord, Journal of Climate , 1, 1099-1123,1988.

Karl,T. R.,R.W. Knight,and B.Baker,The record break-ing global temperatures of 1997 and 1998: Evidencefor an increase in the rate of global warming?,Geophysical Research Letters, 27, 719-722,2000.

Karl,T. R.,R.W. Knight, D. R.Easterling,and R.Quayle,Indices of climate change for the United States,Bulletin of the American Meteorological Society , 77,279-292,1996.

Karl,T. R.,R.W. Knight,and N. Plummer,Trends inhigh-frequency climate variability in the twentiethcentury, Nature, 377, 217-220,1995a.

Karl,T. R.,R.W. Knight, D. R.Easterling,and R. G.Quayle.Indices of climate change for the UnitedStates, Bulletin of the American MeteorologicalSociety, 77, 279-292,1995b.

Keeling,C. D.,and T. P.Whorf,Atmospheric CO2

records from sites in the SIO air sampling network,inTrends: A Compendium of Data on Global Change,Carbon Dioxide Information Analysis Center, OakRidge National Laboratory, US Department of Energy,Oak Ridge,Tennessee,1999.

Kiehl, J.T.,and K.E.Trenberth,Earth’s annual globalmean energy budget, Bulletin of the AmericanMeteorological Society, 78, 197-208,1997.

Kittel,T. G. F., N.A.Rosenbloom,T. H. Painter, D. S.Schimel,and VEMAP participants,The VEMAP inte -grated database for modeling United States ecosys-tem/vegetation sensitivity to climate change, Journalof Biogeography, 22, 857-862,1995.

IPCC (Intergovernmental Panel on Climate Change),Climate Change 1995: The Science of ClimateChange, edited by J.T. Houghton,L. G. Meira Filho,B.A.Callander, N. Harris,A.Kattenberg,and K.Maskell,Cambridge University Press,Cambridge,UnitedKingdom,572 pp.,1996a.

IPCC (Intergovernmental Panel on Climate Change),Climate Change 1995: Impacts,Adaptations, andMitigation of Climate Change: Scientific-TechnicalAnalyses, edited by R.T.Watson,M.C.Zinyowera,andR.H.Moss,Cambridge University Press,Cambridge,United Kingdom,879 pp.,1996b.

IPCC (Intergovernmental Panel on Climate Change),Climate Change 1995: Economic and SocialDimensions of Climate Change, edited by E. J. Bruce,Hoesung Lee,and E.Haites,Cambridge UniversityPress,Cambridge,United Kingdom,464 pp.,1996c.

IPCC (Intergovernmental Panel on Climate Change), AnIntroduction to Simple Climate Models Used in theIPCC Second Assessment Report, by D. Harvey, J.Gregory, M.Hoffert,A. Jain,M.Lal,R.Leemans,S.Raper,T.Wigley, and J. deWolde,and edited by J.T. Houghton,L. G. Meira Filho, D. Griggs,and K.Maskell,WorldMeteorological Organization/United NationsEnvironment Programme,50 pp.,1997.

IPCC (Intergovernmental Panel on Climate Change),Special Report on Emissions Scenarios , N. Nakicenovic(lead author),Cambridge University Press,Cambridge,United Kingdom,599 pp.,2000.

Johns,T. C.,R.E.Carnell, J. F. Crossley, J. M.Gregory, J. F.B.Mitchell,C.A.Senior, S. F. B.Tett,and R.A.Wood,Thesecond Hadley Centre coupled ocean-atmosphereGCM: Model description,spinup and validation,Climate Dynamics, 13, 103-134,1997.

Jones, P. D.,M.New, D. E. Parker, S.Martin,and I. G.Rigor, Surface air temperature and its changes over thepast 150 years, Reviews of Geophysics, 37, 173-199,1999.

Jones, P. D., P.Ya.Groisman,M.Coughlan, N. Plummer,W.-C.Wang,and T. R.Karl,Assessment of urbanizationeffects in time series of surface air temperature overland, Nature, 347, 169-172,1990.

Joussaume,S.,et al.,Monsoon changes for 6000 yearsago: Results of 18 simulations from the PaleoclimateModelling Intercomparison Project (PMIP),Geophysical Research Letters, 26, 859-862,1999.

Judson,A.,and N. Doesken,Density of freshly fallensnow in the central Rocky Mountains, Bulletin of theAmerican Meteorological Society, 81, 1577-1587,2000.

Karl,T. R.,and D. R.Easterling,Climate extremes:Selected review and future research directions,Climatic Change, 42(10),309-325,1999.

66


Kittel,T. G. F., J.A.Royle,C.Daly, N.A.Rosenbloom,W. P.Gibson,H.H. Fisher, D. S.Schimel,L.M.Berliner, andVEMAP participants,A gridded historical (1895-1993)bioclimate dataset for the conterminous United States,pp.219-222 in Proceedings of the 10th Conference onApplied Climatology,American Meteorological Society,Boston,Massachusetts,1997.

Klein,W. H.,Principal tracks and mean frequencies ofcyclones and anticyclones in the NorthernHemisphere, Research Paper 40, US Weather Bureau,Washington DC,60 pp.,1957.

Kleypas, J.A.,R.W. Buddemeier, D.Archer, J. P. Gattuso,C.Langdon,and B. N. Opdyke,Geochemical consequencesof increased atmospheric carbon dioxide on coralreefs, Science, 284,118-120,1999.

Knutson,T. R.,and S.Manabe,Time-mean response overthe tropical Pacific to increased CO2 in a coupledocean-atmosphere model, Journal of Climate, 8, 2181-2199,1995.

Knutson,T. R.,and S.Manabe,Model assessment ofdecadal variability and trends in the tropical PacificOcean, Journal of Climate, 11, 2273-2296,1998.

Knutson,T. R.,and R.E.Tuleya,Increased hurricaneintensities with CO2-induced warming as simulatedusing the GFDL hurricane prediction system, ClimateDynamics, 15, 503-519,1999.

Knutson,T. R.,S.Manabe,and D. Gu,Simulated ENSO ina global coupled ocean-atmosphere model:Multidecadal amplitude modulation and CO 2 sensitivity,Journal of Climate, 10, 138-161,1997.

Knutson,T. R.,T. L.Delworth,K.W. Dixon,and R. J.Stouffer, Model assessment of regional surface tempera-ture trends (1949-1997), Journal of GeophysicalResearch, 104, 30981-30996,1999.

Knutson,T. R.,R.E.Tuleya,and Y. Kurihara,Simulatedincrease of hurricane intensities in a CO 2-warmed cli-mate, Science, 279, 1018-1020,1998.

Kothavala,Z.,Extreme precipitation events and theapplicability of global climate models to study floodsand drought, Mathematics and Computers inSimulations, 43, 261-268,1997.

Krishnamurti,T. N.,R.Correa-Torres,M.Latif, and G.Daughenbaugh,The impact of current and possiblyfuture sea surface temperature anomalies on the fre-quency of Atlantic hurricanes, Tellus, 50A, 186-210,1998.

Kunkel,K.E,K.Andsager, and D. R.Easterling,Long-termtrends in extreme precipitation events over the conter-minous United States and Canada, Journal of Climate,12, 2515-2527,1999.

Kutzbach, J. E., P. J. Guetter, P. J. Behling,and R.Selin,Simulated climatic changes: Results of the COHMAPclimate-model experiments,in Global Climates Sincethe Last Glacial Maximum, edited by H.E.Wright,Jr.,et al.,pp.24-93,University of Minnesota Press,Minnesota,1993.

Lambert,S. J.,The effect of enhanced greenhousewarming on winter cyclone frequencies and strengths,Journal of Climate, 8(5),1447-1452,1995.

Langbein,W. B.,Annual Runoff in the United States, USGeological Survey Circular No. 5, Department of theInterior,Washington,DC,1949 (reprinted 1959).

Lau,K.M.,and H.Weng,Interannual,decadal-to-inter-decadal and global warming signals in sea surface tem-perature during 1955-1997, Journal of Climate, 12,1257-1267,1999.

Lean, J.,and D. Rind,Climate forcing by changing solarradiation, Journal of Climate, 11, 3069-3094,1998.

Lean, J., J. Beer, and R.S.Bradley, Reconstruction of solarirradiance since 1620:Implications for climate change,Geophysical Research Letters, 22(23),3195-3198,1995.

Leatherman,S. P.,K.Zhang,and B.C.Douglas,Sea levelrise shown to drive coastal erosion, Transactions ofthe American Geophysical Union, EOS, 81, 55-57,2000(also see responses and reply in EOS, 81, 436-437 and439-441,2000).

Ledley,T. S.,E.T. Sundquist,S.E.Schwartz, D. K.Hall, J. D.Fellows,and T. L.Killeen,Climate Change andGreenhouse Gases, EOS, Transactions of the AmericanGeophysical Union, 80(39),September 28,1999,p.453,1999.(Available athttp://www.agu.org/eos_elec/99148e.html)

Legates, D. R.,Global and terrestrial precipitation: Acomparative assessment of existing climatologies:Areply, International Journal of Climatology, 17, 779-783,1997.

Legates, D. R.,and T. L.DeLiberty, Precipitation measure-ments biases in the United States, Water ResourcesBulletin, 29, 855-861,1993.

Legates, D. R.,and C. J.Wilmott,Mean seasonal and spa-tial variability in global surface air temperature,Theoretical and Applied Climatology, 41, 11-21,1990a.

Legates, D. R.,and C. J.Wilmott,Mean seasonal and spa-tial variability in gauge-corrected global precipitation,International Journal of Climatology, 10, 111-127,1990b.

Lins,H. F.,and J. R.Slack,Streamflow trends in theUnited States, Geophysical Research Letters, 26(2),227-230,1999.

67


Mitchell, J. F. B.,and T. C. Johns,On modification of glob-al warming by sulfate aerosols, Journal of Climate,10(2),245-267,1997.

Mitchell, J. F. B.,T. C. Johns, J. M.Gregory, and S.Tett,Climate response to increasing levels of greenhousegases and sulphate aerosols, Nature, 376, 501-504,1995.

Mitchell, J. F. B.,T. J. Johns,W. J. Ingram,and J.A.Lowe,The effect of stabilising atmospheric carbon dioxideconcentrations on global and regional climate change,Journal of Geophysical Research, 27, 2977-2980,2000.

Mitchell, J. F. B.,C.A.Wilson,and W. M.Cunnington,OnCO2 climate sensitivity and model dependence ofresults, Quarterly Journal of the Royal MeteorologicalSociety, 113, 293-322,1987.

Monahan,A.H., J. C.Fyfe,and G. M.Flato,A regime viewof Northern Hemisphere atmospheric variability andchange under global warming, Geophysical ResearchLetters, 27,1139-1142,2000.

Mote, P.,and A.O’Neill, Numerical Modeling of theGlobal Atmosphere in the Climate System, KluwerAcademic Publishers,Boston,Massachusetts,517 pp.,2000.

NAS (National Academy of Sciences), PolicyImplications of Greenhouse Warming:Mitigation,Adaptation, and the Science Base, Panel on PolicyImplications of Greenhouse Warming, NationalAcademy Press,Washington,DC,1992.

Neftel,A.,H. Friedli,E.Moor, H.Lötscher, H.Oeschger, U.Siegenthaler, and B.Stauffer, Historical CO2 record fromthe Siple Station ice core,in Trends: A Compendiumof Data on Global Change, Carbon DioxideInformation Analysis Center, Oak Ridge NationalLaboratory, US Department of Energy, Oak RidgeTennessee,1994.

Nihoul, J. C. J. (Ed.), Coupled Ocean-AtmosphereModels, Elsevier Science,Amsterdam,767 pp.,1985.

Noda,A.,K.Yamaguchi,S.Yamaki,and S.Yukimoto,Relationship between natural variability and CO2-induced warming pattern: MRI AOGCM experiment,pp.359-362 in Preprints volume, 10th Symposium onGlobal Change Studies,American MeteorologicalSociety, Boston,Massachusetts, 1999.

NRC (National Research Council), Carbon Dioxide andClimate: A Scientific Assessment, National Academy ofSciences,Washington,DC,22 pp.,1979.

NRC (National Research Council), Changing Climate:Report of the Carbon Dioxide Assessment Committee,National Academy Press,Washington,DC,496 pp.,1983.

Lunkeit, F.,M. Ponater, R.Sausen,M.Sogalla, U. Ulbrich,and M.Windelband,Cyclonic activity in a warmer cli-mate, Contributions to Atmospheric Physics, 69, 393-407,1996.

Mahlman, J. D.,Uncertainties in projections of human-caused climate warming, Science, 278, 1416-1417,1997.

Manabe,S.,and R.T.Wetherald,The effects of doublingthe CO2 concentration on the climate of a general cir-culation model, Journal of the Atmospheric Sciences,32, 3,1975.

Mann,M.E.,R.S.Bradley, and M.K.Hughes,Northernhemisphere temperatures during the past millennium:Inferences,uncertainties,and limitations, GeophysicalResearch Letters, 26(6),759-762,1999.

Mantua, N.,S.Hare,Y. Zhang, J.Wallace,and R. Francis,APacific interdecadal climate oscillation with impacts onsalmon production, Bulletin of the AmericanMeteorological Society, 78, 1069-1079,1997.

Marland, G.,T.A.Boden,R. J.Andres,A.L.Brenkert,andC. Johnston,Global, regional,and national CO2 emis-sions,in Trends: A Compendium of Data on GlobalChange, Carbon Dioxide Information Analysis Center,Oak Ridge National Laboratory, US Department ofEnergy, Oak Ridge,Tennessee,1999.

McFarlane, N.A., G. J. Boer, J. P. Blanchet,and M.Lazare,The Canadian Climate Centre second-generation gener-al circulation model and its equilibrium climate,Journal of Climate , 5, 1013-1044,1992.

Mearns,L. O.,I.Bogardi, F. Giorgi,I.Matyasovszky, andM. Palecki,Comparison of climate change scenariosgenerated from regional climate model experimentsand statistical downscaling, Journal of GeophysicalResearch, 104(D6),6603-6621,1999.

Mearns,L. O., F. Giorgi,L.McDaniel,and C.Shields,Analysis of the diurnal range and variability of dailytemperature in a nested modeling experiment:Comparison with observations and 2 x CO 2 results,Climate Dynamics, 11, 193-209,1995.

Meehl, G.A.,and W. M.Washington,El Niño-like climatechange in a model with increased atmospheric CO2

concentrations, Nature, 382, 56-60,1996.

Meehl, G.A., G. J. Boer, C.Covey, M.Latif, and R. J.Stouffer, Meeting summary: The Coupled ModelIntercomparison Project (CMIP), Bulletin of theAmerican Meteorological Society, 81, 313-318,2000a.

Meehl, G.A, F. Zwiers, J. Evans,T. Knutson,L.Mearns,andP.Whetton,Trends in extreme weather and climateevents: Issues related to modeling extremes in projec-tions of future climate change, Bulletin of theAmerican Meteorological Society, 81, 427-436,2000b.68


NRC (National Research Council), Capacity of USClimate Modeling, National Academy Press,Washington,DC,1998.

NRC (National Research Council), ReconcilingObservations of Global Temperature Change, NationalAcademy Press,Washington,DC,85 pp.,2000.

Palmer W.,Meteorological Drought, Research PaperNo. 45, US Weather Bureau,NOAA Library andInformation Services Division, Washington,DC,58 pp.,1965.

Parker, D. E.,T. P. Legg,and C.K. Folland,Interdecadalchanges of surface temperatures since the late 19th cen-tury, Journal of Geophysical Research , 99, 14373-14399,1994.

Petit, J. R.,et al.,Climate and atmospheric history of thepast 420,000 years from the Vostok ice core,Antarctica,Nature, 399, 429-436,1999.

Pisias, N. G.,and N. J. Shackleton,Modelling the globalclimate response to orbital forcing and atmosphericcarbon dioxide changes, Nature, 310, 757-759,1984.

Pitman,A.,R.Pielke,Sr.,R.Avissar, M.Claussen, J. Gash,and H.Dolman,The role of the land surface in weatherand climate: Does the land surface matter?, IGBPNewsletter, 39,pp.4-9,September 1999.

PSAC (President’s Science Advisory Council),Atmospheric Carbon Dioxide, Appendix Y4 inRestoring the Quality of Our Environment, Report ofthe Environmental Pollution Panel,The White House,Washington,DC,1965.

Quayle,R. G,T. C. Peterson,A.Basist,and C.Godfrey,Anoperational near-real-time global temperature index,Geophysical Research Letters, 26, 333-335,1999.

Risbey, J. S., P. J. Kushner, P. J. Lamb,R.Miller, M.C.Morgan,M.Richman, G. Roe, and J. Smith,Generatingregional climate scenarios by combining synoptic-cli-matological guidance and GCM output,US EPAResearch Report, Washington DC,88 pp.,1999 (alsosubmitted to Journal of Climate by a subset of theauthors and with dif ferent title).

Rothrock, D.A.,Y.Yu,and G.A.Maykut,Thinning of theArctic sea-ice cover, Geophysical Research Letters,26(23),3469-3472,1999.

Schimel, D, I. G. Enting,M.Heimann,T. M.L.Wigley, D.Raynaud, D.Alves,and U. Siegenthaler, CO2 and the car-bon cycle,in Climate Change 1994, Radiative Forcingof Climate Change and an Evaluation of the IPCC1992 Emission Scenarios, edited by J.T. Houghton,etal.,Cambridge University Press,Cambridge,UnitedKingdom,1995.

Schubert,M., J. Perlwitz,R.Blender, K. Fraedrich,and F.Lunkeit,North Atlantic cyclones in CO2-induced warmclimate simulations: Frequency, intensity, and tracks,Climate Dynamics, 14, 827-837,1998.

Shackleton, N. J., J. Le,A.C.Mix,and M.Hall,Carbon iso-tope records from Pacific surface waters and atmos-pheric carbon dioxide, Quaternary Science Reviews,11, 387-400,1992.

Smith,E.,Atlantic and East Coast hurricanes 1900-98: Afrequency and intensity study for the twenty-first cen-tury, Bulletin of the American Meteorological Society ,80, 2717-2720,1999.

Smith,S. J., N. Nakicenovic,and T. M.L.Wigley, Radiativeforcing in the IPCC SRES scenarios, Nature, in review,2000.

Sousounis, P.,A synoptic assessment of climate changemodel output: Explaining the differences and similari-ties between the Canadian and Hadley climate models,in Eleventh Symposium on Global Change Studies,80th American Meteorological Society AnnualMeeting, January 2000,American MeteorologicalSociety, Boston,Massachusetts,1999.

Stahle, D.W.,E.R.Cook,M.K.Cleaveland,M. D.Therrell,D. M.Meko,H. D. Grisino-Mayer, E.Watson,and B.H.Luckman,Tree-ring data document 16 th centurymegadrought over North America, Transactions of theAmerican Geophysical Union (EOS), 81, 121 and 125,2000.

Steadman,R. G.,The assessment of sultriness,part I: Atemperature-humidity index based on human physiolo-gy and clothing science, Journal of Climate andApplied Meteorology, 18, 861-873,1979.

Stott, P.A.,S. F. B.Tett, G. S. Jones,M.R.Allen,W. J. Ingram,and J. F. B.Mitchell,Attribution of twentieth centurytemperature change to natural and anthropogenic caus-es, Climate Dynamics, in press,2000.

Stouffer, R. J., G. Hegerl,and S.Tett,A comparison of sur-face air temperature variability in three 1000-year cou-pled ocean-atmosphere model integrations, Journal ofClimate, 13, 513-537,2000.

Tett,S. F. B., P.A.Stott,M.R.Allen,W. J. Ingram,and J. F. B.Mitchell,Causes of twentieth century temperaturechange, Nature, 399, 569-572,1999.

Thomson, D. J.,The seasons,global temperature,andprecession, Science, 268, 59-68,1995.

Timmermann,A.,Oberhuber, J.,Bacher,A.,Esch,M.,Latif, M.,and E.Roeckner, Increased El Niño frequencyin a climate model forced by future greenhouse warm-ing, Nature, 398, 694-697,1999.

69


Wigley,T. M.L.,A.K. Jain, F. Joos,B.S.Nyenzi,and P. R.Shukla,Implications of proposed CO2 emissions limita-tions,Technical paper for the Intergovernmental Panelon Climate Change,Geneva,Switzerland,41 pp.,1997.

Willson,R.C.,Total solar irradiance trend during solarcycles 21 and 22, Science, 277, 1963-1965,1997.

Woodhouse,C.,and J. Overpeck,2000 years of droughtvariability in the central United States, Bulletin of theAmerican Meteorological Society, 79, 2693-2714,1998.

Yoshimura, J.,M.Sugi,and A.Noda,Influence of green-house warming on tropical cyclone frequency simulat-ed by a high-resolution AGCM,pp.555-558 in Preprintsvolume, 10th Symposium on Global Change Studies ,American Meteorological Society, BostonMassachusetts,1999.

Zhang,Y., J. M.Wallace,and D. S.Battisti,ENSO-like inter-decadal variability: 1900-93, Journal of Climate, 10,1004-1020,1997.

Zwiers, F.W.,and V. Kharin,Changes in the extremes ofthe climate simulated by CCC GCM2 under CO2 dou-bling, Journal of Climate, 11, 2200-2222,1998.

Trenberth,K.E.,and G. Branstator, Issues in establishingcauses of the 1988 drought over North America,Journal of Climate , 5, 159-172,1992.

Trenberth,K.E.and J.W. Hurrell,Decadal atmosphere-ocean variations in the Pacific, Climate Dynamics, 9,303-319,1994.

USDOE (US Department of Energy), Projecting theClimatic Effects of Increasing Carbon Dioxide, editedby M.C.MacCracken and F. M.Luther, US Departmentof Energy,Washington DC,1985a.

USDOE (US Department of Energy), Detecting theClimatic Effects of Increasing Carbon Dioxide, editedby M.C.MacCracken and F. M.Luther, US Departmentof Energy,Washington DC,1985b.

USEPA (US Environmental Protection Agency),ThePotential Effects of Global Climate Change on theUnited States, EPA-230-05-89-050,edited by J. Smith andD.Tirpak,Washington DC,1989.

Van Loon,H.,and J. Rogers,The seesaw in winter tem-peratures between Greenland and Northern Europe.Part 1: General description, Monthly Weather Review,106, 296-310,1978.

VEMAP (Vegetation/Ecosystem Modeling and AnalysisProject) Members:Comparing biogeography and bio-geochemistry models in a continental-scale study of ter-restrial ecosystem responses to climate change andCO2 doubling, Global Biogeochemical Cycles, 9, 407-437,1995.

Walsh,K. J. E.,and B. F. Ryan,Idealized vortex studies ofthe effect of climate change on tropical cyclone inten -sities,pp.403-404 in Preprints volume, 23rd

Conference on Hurricanes and Tropical Meteorology,American Meteorological Society, Boston,Massachusetts,1999.

Washington,W. M.,and C.L. Parkinson, An Introductionto Three-Dimensional Climate Modeling, UniversityScience Books,Mill Valley, California,422 pp.,1986.

Washington,W. M.,et al., Parallel Climate Model (PCM)control and 1%/year CO2 simulations with a 2/3º oceanmodel and a 27 km dynamical sea ice model, ClimateDynamics, in press,2000.

Wetherald,R.T.,and S.Manabe,Detectability of summerdryness caused by greenhouse warming, ClimaticChange, 43, 495-511,1999.

Wigley,T. M.L.,and D. Schimel (Eds.), The CarbonCycle, Cambridge University Press,Cambridge UnitedKingdom,312 pp.,2000.

70


ACKNOWLEDGEMENTSMany of the materials for this chapter are based oncontributions from participants on and those working with the

Climate Scenario TeamRichard Ball,Department of Energy (retired)Tony Barnston,NOAA National Centers for

Environmental Prediction,Climate PredictionCenter

Eric Barron,Pennsylvania State UniversityDenise Blaha,University of New HampshireGeorge Boer, Canadian Centre for Climate Modelling

and Analysis,Victoria,BCRuth Carnell,Hadley Centre,Meteorological Office,

Bracknell,UKAiguo Dai,National Center for Atmospheric

ResearchChristopher Daly, Oregon State UniversityDavid Easterling,NOAA National Climatic Data

CenterBenjamin Felzer,National Center for Atmospheric

ResearchHank Fisher, National Center for Atmospheric

ResearchGreg Flato,Canadian Centre for Climate Modelling

and AnalysisByron Gleason,NOAA National Climatic Data CenterJonathan Gregory, Hadley Centre,Meteorological

Office,Bracknell,UKYuxiang He,NOAA National Centers for

Environmental Prediction,Climate PredictionCenter

Preston Heard,Indiana University - BloomingtonRoy Jenne,National Center for Atmospheric

ResearchDennis Joseph,National Center for Atmospheric

ResearchTom Karl,NOAA National Climatic Data Center

Tim Kittel,National Center for AtmosphericResearch

Richard Knight,NOAA National Climatic DataCenter

Steven Lambert,Canadian Centre for ClimateModelling and Analysis,Victoria,BC

Michael MacCracken,USGCRP/National AssessmentCoordination Office

Linda Mearns,National Center for AtmosphericResearch

John Mitchell,Hadley Centre,Meteorological Office,Bracknell,UK

James Risbey, Carnegie Mellon UniversityNan Rosenbloom,National Center for Atmospheric

ResearchJ.Andy Royle, U. S. Fish and Wildlife Service,Laurel

MDAnnette Schloss,University of New HampshireJoel B.Smith,Stratus ConsultingSteven J. Smith,Battelle Pacific Northwest National

LaboratoryPeter Sousounis,University of MichiganDavid Viner, Climatic Research Unit,Norwich,UKWarren Washington,National Center for

Atmospheric ResearchTom Wigley, National Center for Atmospheric

ResearchFrancis Zwiers,Canadian Centre for Climate

Modelling and Analysis,Victoria,BC

71


73

CHAPTER 2

VEGETATION AND BIOGEOCHEMICALSCENARIOSJerry Melillo1,2, Anthony Janetos3,2, David Schimel4,5, and Tim Kittel5


Chapter Summary

Introduction

Research Approach

Biochemistry Models

Biogeography Models

Databases

Vegetation in the Future

Biogeochemical Simulation Results

Biogeography Simulation Results

Literature Cited

Acknowledgments

1The Ecosystems Center, Marine Biological Laboratory; 2Coordinating author for the National Assessment SynthesisTeam, 3World Resources Institute, 4Max-Planck Institut für Biogeochemie, 5National Center for Atmospheric Research

CHAPTER SUMMARYEcosystems are communities of plants and animalsand the physical environment in which they exist.Ecologists often categorize ecosystems by theirdominant vegetation – the deciduous broad-leafedforest ecosystems of New England,the short-grassprairie ecosystems of the Great Plains,the desertecosystems of the Southwest. Concerns for contin-ued ecosystem health and performance stem fromtwo primary issues. Ecosystems of all types,fromthe most natural to the most extensively managed,produce a variety of goods and services that benefithumans. Examples of ecosystem services includemodification of local climate,air and water purifica-tion,landscape stabilization against erosion, floodcontrol,and carbon storage. Ecosystems are also val-ued for recreational and aesthetic reasons. Climatechange has the potential to affect the structure,function,and regional distribution of ecosystems,and thereby affect the goods and services they pro-vide.

For this Assessment,the Vegetation/EcosystemModeling and Analysis Project (VEMAP) was used togenerate future ecosystem scenarios for the conter-minous United States based on model-simulatedresponses to the Canadian and Hadley scenarios ofclimate change. The ecosystem scenarios were thenshared with Assessment participants to assist themin their evaluations of the potential sensitivities ofecosystems and ecosystem goods and services to cli-mate change.


74

Key Findings

Some of the key results from VEMAP for ecosys-tems in the absence of land-cover and land-usechanges are as follows:

• Over the next few decades climate change isvery likely to lead to increased plant productivi-ty and increased terrestrial carbon storage formany parts of the country, especially those thatget moderately warmer and wetter. Areaswhere soils dry out during the growing season,such as the Southeast for the climate simulatedwith the Canadian model,are very likely to seereduced productivity and decreases in carbonstorage.

• By the end of the 21st century, many regions ofthe country are likely to have experiencedchanges in vegetation distribution. Areas inwhich soil moisture increases are likely to main-tain or exhibit an increased woody component

of vegetative cover. Areas in which soil moisturedecreases are likely to lose woody vegetation.For example,in the Southeast,the climate simu-lated by the Canadian model causes soil dryingthat would lead to forest losses and savanna andgrassland expansion.

• Modeling of vegetation responses to climatechange is in the early stages of development. Nosingle model simulates all of the important fac-tors affecting vegetation responses to climate;model results must therefore be viewed withcaution. The complex,non-linear nature ofecosystems almost certainly means that we willbe surprised by some of the changes in ecosys-tem function and structure that climate changesset in motion. Keeping the magnitude of climatechange as small as possible and slowing its rateare the two things we can do to minimize thenegative impacts on natural ecosystems.

Chapter 2 / Vegetation and Biogeochemical Scenarios

75

VEGETATION AND BIOGEOCHEMICAL SCENARIOS

et al.,2000;Neilson et al.,2000). The models use acommon “baseline”data set and two potential climatescenarios. Common data are used to ensure that anyvariability in predicted responses is attributable tothe different structures and formulations of individ-ual ecological models rather than to input data.

For the National Assessment,the focus is on modeloutputs for two time periods:2025-2034 (near term)and 2090-2099 (long term). Outputs of the biogeo-chemistry models are used to consider near-termecological impacts,while outputs of the biogeogra-phy models are used to consider longer-termimpacts. This is based on the team’s expert judge-ment that biogeochemical changes will dominateecological responses to climate change in the nextfew decades,while species shifts will dominate eco-logical responses to climate change towards the endof the 21st century, as organisms attempt to migrateto occupy“optimal”climate space.

RESEARCH APPROACH

Biogeochemistry Models

The biogeochemistry models simulate the cycles ofcarbon, nutrients (e.g.,nitrogen),and water in terres-trial ecosystems which are parameterized accordingto life form (VEMAP Members,1995,Schimel et al.,2000). The models consider how these cycles areinfluenced by environmental conditions includingtemperature,precipitation,solar radiation,soil tex-ture,and atmospheric CO2 concentration. Theseenvironmental variables are inputs to general algo-rithms that describe plant and soil processes such ascarbon capture by plants with photosynthesis,decomposition,soil nitrogen transformations mediat-ed by microorganisms,and water flux between landand the atmosphere in the processes of evaporationand transpiration. Common outputs from biogeo-chemistry models are estimates of net primary pro-ductivity, net nitrogen mineralization, evapotranspira-tion fluxes (e.g.,PET, ET),and the storage of carbonand nitrogen in vegetation and soil. In the VEMAP IIactivity, three biogeochemistry models were used:BIOME-BGC (Hunt and Running,1992;Running andHunt,1993),CENTURY (Parton et al.,1987,1988,1993),and the Terrestrial Ecosystem Model (TEM)

INTRODUCTIONThis chapter is designed to report the results of theVegetation/Ecosystem Modeling and Analysis ProjectII (VEMAP II);a project that has provided data aboutterrestrial ecosystem responses to climate change toAssessment participants. The chapter is not meant tobe a comprehensive,in-depth analysis of climateimpacts on all aspects of terrestrial ecosystem struc-ture and function. The chapter has two focus areas –biogeochemistry and plant biogeography in naturalterrestrial ecosystems. While animal communitiesare mentioned brief ly in the chapter, they were notconsidered in the VEMAP II analysis and so are notfocused on in this chapter. These scenarios for vege-tation and biogeochemical change can serve as back-ground for analyses of changes to fauna and biologi-cal diversity by contributing broad-scale informationon habitat changes.

VEMAP II is an international,collaborative effort sup-ported by several US Global Change ResearchProgram agencies and sponsored by the InternationalGeosphere-Biosphere Program (IGBP) to conduct ananalysis of the potential effects of climate change onecosystem processes and vegetation distributionwithin the continental United States. Modelingresults to date indicate that natural terrestrial ecosys-tems are sensitive to changes in global surface tem-perature,precipitation patterns,atmospheric carbondioxide (CO2) levels,and other climate parameters.Major ecological characteristics to be affectedinclude the geographic distribution of dominantplant species,productivity of plants,biodiversitywithin natural ecosystems,and basic ecologicalprocesses and their feedbacks to the climate system.

Two types of models that have been used in VEMAPII to examine the ecological effects of climatechange are biogeochemistry models and biogeogra-phy models. Biogeochemistry models projectchanges in basic ecosystem processes such as thecycling of carbon, nutrients,and water (ecosystemfunction),and biogeography models simulate shiftsin the geographic distribution of major plant speciesand communities (ecosystem structure).

VEMAP II involves a comparison of three biogeo-chemistry and three biogeography models (Schimel


76


77

and soils for non-wetland ecosystems of the globe(Tian et al.,1999). This model requires monthly cli -matic data along with soil and vegetation-specificparameters to estimate monthly carbon and nitro-gen fluxes and pool sizes. The model includes algo-rithms from the water balance model of Vörösmartyet al.(1989) to calculate potential and actual evapo-transpiration,soil moisture,and drainage. Estimatesof net primary production and carbon storage bythis version of TEM have been evaluated in previousapplications of the model at both regional and glob-al scales (Xiao et al.,1998;Tian et al.,1998,1999,2000;Kicklighter et al.,1999;Prinn et al.,1999;Reilly et al.,1999;McGuire et al.,2000).

BIOGEOGRAPHY MODELS The models used to estimate biogeographic respons-es to climate change in VEMAP II include LPJ, MAPSS(Mapped Atmosphere-Plant-Soil System) and MC1.These three models project the local dominance ofvarious terrestrial vegetation forms based on (1)ecophysiological constraints,which determine thebroad distribution of major categories of woodyplants,and (2) response limitations,which deter-mine specific aspects of community composition,such as the competitive balance of trees and grass-es. Though similar in some respects,these modelssimulate potential evapotranspiration and direct CO2

effects differently, and as a result they show varyingsensitivities to temperature,CO2 levels,and otherfactors. Two of the models,LPJ and MC1 have bio-geochemistry modules while the third,MAPSS,doesnot. Both LPJ and MC1 are dynamic vegetationmodels,while MAPSS is an equilibrium model.

LPJThe LPJ-Model was constructed in a modular frame-work. Individual modules describe key ecosystemprocesses,including vegetation establishment,resource competition, growth,and mortality (Sitch,2000). Vegetation structure and composition isdescribed by nine plant functional types (PFTs)which are distinguished according to their plantphysiological (C3, C4 photosynthesis),phenological(deciduous, evergreen) and physiognomic (tree,grass) attributes. The model is run on a grid cellbasis with input of soil texture,monthly fields oftemperature,precipitation,and percentage sunshinehours. Each grid cell is divided into fractions cov-ered by the PFTs and bare ground. The presenceand fractional coverage of an individual PFTdepends on its specific environmental limits,and onthe outcome of resource competition with theother PFTs.

(Melillo et al.,1993;McGuire et al.,1997;Tian et al.,1999). The similarities and differences among themodels are summarized in Table 1. A detailed inter-comparison of these biogeochemistry models hasrecently been published (Pan et al.,1998). The capa-bilities and limitations of the models are identified inthis intercomparison. A comparison of model resultsto field data for the Mid-Atlantic region of the north-eastern US is presented in Jenkins,et al.(2000).

BIOME-BGCThe BIOME-BGC (BioGeochemical Cycles) model isa multi-biome generalization of FOREST-BGC, amodel originally developed to simulate a forest standdevelopment through a life cycle (Running andCoughlan,1988;Running and Gower, 1991). Themodel requires daily climate data and the definitionof several key climate, vegetation,and site conditionsto estimate fluxes of carbon,nitrogen,and waterthrough ecosystems (Table 4 in VEMAP Members,1995). Allometric relationships are used to initializeplant and soil carbon (C) and nitrogen (N) poolsbased on the leaf pools of these elements (Vitouseket al.,1988). Components of BIOME-BGC have pre-viously undergone testing and validation,includingthe carbon dynamics (McLeod and Running,1988;Korol et al.,1991;Hunt et al.,1991;Pierce,1993;Running,1994) and the hydrology (Knight et al.,1985;Nemani and Running,1989;White andRunning,1995).

CENTURYThe CENTURY model is a general model of plant-soilnutrient cycling which has been used to simulatecarbon and nutrient (nitrogen,phosphorus,and sul-fur) dynamics for different types of ecosystemsincluding grasslands, agricultural lands, forests,andsavannas (Parton et al.,1987,1993;Metherell,1992).For VEMAP, only carbon and nitrogen dynamics areincluded. The model uses monthly temperature andprecipitation data as well as atmospheric CO2 and Ninputs to estimate monthly stocks and fluxes of car-bon and nitrogen in ecosystems. The CENTURYmodel also includes a water budget submodel whichcalculates monthly evapotranspiration,transpiration,water content of the soil layers,snow water content,and saturated flow of water between soil layers. TheCENTURY model incorporates algorithms thatdescribe the impact of fire, grazing,and storm distur-bances on ecosystem processes (Ojima et al.,1990;Sanford et al.,1991;Holland et al.,1992;Metherell,1992).

TEM The Terrestrial Ecosystem Model (TEM version 4.1)describes carbon and nitrogen dynamics of plants


78

References

Responses ofPlant Physiology

CO2

Temperature

Moisture regime

Solar radiation

Responses ofSoil Processes

CO2

Temperature

Precipitation

Solar radiation

DisturbanceRegimes

Biome-BGCRunning and Hunt (1993)

Reduction in canopy conduc-tance and leaf N concentra-tion;and increases in intercel-lular CO2 concentration,pro-duction and water-use effi-ciency

Optimum temperature forphotosynthesis;maintenancerespiration increases withtemperature; growth respira-tion increases with photosyn-thesis

Canopy conductance increas-es with enhanced soil mois -ture and reduced vapor pres-sure deficit

Photosynthesis increases withenhanced photosyntheticallyactive radiation (PAR)

Soil moisture increases withreduced canopy conduc-tance;decompositiondecreases with lower N con-centration in litterfall

with increases in tempera-ture:1) decompositionincreases;2) soil moisturedecreases:and 3) net N min-eralization increases

Soil moisture increase withenhanced precipitation;opti-mum soil moisture fordecomposition

Soil moisture decreases withenhanced solar radiation

Prescribed mortality

CenturyParton et al.(1994)

Reductions in transpirationand leaf N concentration;andprescribed increases inpotential production

Optimum temperature of pro-duction

Potential production increas-es with enhanced soil mois -ture

None

Soil moisture increases withreduced transpiration;decom-position decreases withlower N concentration in lit -terfall



None

Scheduled fire regimes

TEMTian et al.(1999)

Increases in intercellular CO2

concentration and produc-tion

Optimum temperature ofgross primary production(GPP);maintenance respira-tion increases with tempera-ture; growth respirationincreases with GPP

GPP increases with enhancedevapotranspiration;phenolo-gy modified with enhancedevapotranspiration

GPP increases with enhancedPAR

Decomposition decreaseswith lower N concentrationin litterfall



Soil moisture decreases withenhanced solar radiation

Implicitly implementedthrough litterfall fluxes

Table 1. Key Characteristics of the Three Biogeochemical Models used in VEMAP II.


79

MC1MC1 consists of three linked modules simulatingbiogeography, biogeochemistry, and fire disturbance(Lenihan et al.,1998;Daly et al.,2000). The mainfunctions of the biogeography module are:(1) tosimulate the composition of deciduous/evergreen,needleleaf/broadleaf tree and C3/C4 grass life-formmixtures from climatic thresholds;and (2) to classifythose woody and herbaceous life forms into differ-ent vegetation classes based on their biomass (orleaf area index) simulated by the biogeochemistrymodule.

The biogeochemistry module,which is based on theCENTURY model (Parton et al.,1987),simulatesmonthly carbon and nutrient dynamics for a givenlife-form mixture. It was configured to always allowtree-grass competition. Above- and below-groundprocesses are modeled in detail,and include plantproduction,soil organic matter decomposition,andwater and nutrient cycling. Nitrogen (N) demand isalways assumed to be met in this study and neverlimited by local conditions since there were no soilN data available to initialize and calibrate the model.

Parameterization of this module is based on the lifeform composition of the ecosystems,which isupdated annually by the biogeography module. Thefire module simulates the occurrence,behavior andeffects of severe fire. Allometric equations, keyed tothe life-form composition supplied by the biogeogra-phy module,are used to convert above-ground bio-mass to fuel classes. Fire effects (i.e.,plant mortalityand live and dead biomass consumption) are esti-mated as a function of simulated fire behavior (i.e.,fire spread and fire line intensity) and vegetationstructure. Fire effects feed back to the biogeochem-istry module to adjust the levels of the carbon andnutrient pools. A detailed description of the modelcan be found in Daly et al.(2000).

Simulated grazing is species-independent and onlyoccurs in the model between April and September.Only grasses are consumed and there is no treedeath assumed due to either consumption or tram-pling by herbivores. A fraction of the material con-sumed by the grazers (C and N) is returned tothe site.

DATABASESTo meet the various input requirements of the bio-geochemistry and biogeography models and ensurea common starting point for the VEMAP II simula-tions,the “baseline”database was created to incor-

The two-layer soil water balance model is basedon Haxeltine and Prentice (1996). Moisture ineach layer, expressed as a fraction of water hold-ing capacity, is updated daily. Percolation from theupper to the lower layer, and absolute water hold-ing capacity are soil texture dependent.

Establishment and mortality are modeled on anannual basis. Plant establishment,in terms of addi-tional PFT individuals,depends on the fraction ofbare ground available for seedlings to successfullyestablish. Natural mortality is taken as a functionof PFT vigor, and corresponds to an annual reduc-tion in the number of PFT individuals. Dead bio-mass enters the litter pool,and the soil pools.Mortality also occurs due to disturbance(Thonicke et al.,2000).

MAPSSThe MAPSS (Mapped Atmosphere-Plant-SoilSystem) model begins with the application of eco-physiological constraints to determine whichplant types can potentially occur at a given loca-tion. A two-layer hydrology module (includinggravitational drainage) with a monthly time stepthen allows simulation of leaf phenology, leaf areaindex (LAI) and the competitive balance betweengrass and woody vegetation. A productivity indexis derived based on leaf area duration and evapo-transpiration. This index is used to assist in thedetermination of leaf form,phenology, and vegeta-tion type,on the principle that any successfulplant strategy must be able to achieve a positiveNet Primary Production (NPP) during its growingseason.

The LAI of the woody layer provides a light-limita-tion to grass LAI. Stomatal conductance is explic-itly included in the water balance calculation,andwater competition occurs between the woodyand grass life forms through different canopy con-ductance characteristics as well as rooting depths.The direct effect of CO2 on the water balance issimulated by reducing maximum stomatal conduc-tance. The MAPSS model is calibrated againstobserved monthly runoff, and has been validatedagainst global runoff (Neilson and Marks,1995). Asimple fire model is incorporated to limit shrubsin areas such as the Great Plains (Neilson,1995).

The forest-grassland ecotone is reproduced byassuming that closed forest depends on a pre-dictable supply of winter precipitation for deepsoil recharge (Neilson et al.,1992). An index isused that decrements the woody LAI as the sum-mer dependency increases.

Table 2. Simulated Changes in Annual Net Primary Production due to Changes in Climate plus CO2and Climate only in the Conterminous United States.

Changes

Hadley Climate Canadian Simulation Climate

Simulation

Biome-BGC 2800 Climate + CO2 +439 (15.7%) +222 (7.9%)Climate +98 (3.6%) -274 (-10.0%)

CENTURY 3300 Climate + CO2 +177 (7.1%) +72 (2.9%)Climate +109 (4.4%) +3 (0.1%)

TEM 3500 Climate + CO2 +539 (15.4%) +397 (11.3%)Climate +221 (6.6%) -102 (-3.1%)

21st century in VEMAP II and so shifts in croplandareas and expansion of urban areas is not included.

Climate change scenarios are based on two atmos-pheric general circulation model (GCM) exper i-ments – one conducted at the Hadley Centre forClimate Prediction and Research of theMeteorological Office of the United Kingdom(HadCM2 version) (henceforth,Hadley) and theother at the Canadian Centre for Climate Modellingand Analysis (henceforth,Canadian). These scenar-ios were selected because they are representativeof the higher and lower halves of the range of tem-perature sensitivity among the “transient”GCMsavailable at the beginning of VEMAP II.

Because elevated CO2 may directly affect plantsindependently of whether it causes any change inclimate,VEMAP II included a partial factorial experi-mental design in which simulations were run withboth climate and CO2 changing through time andthen only climate changing through time. Both thebiogeochemistry and biogeography models wererun with both transient climate and CO2 and withtransient climate alone.

For the biogeochemistry models,several aspects ofcarbon cycle changes were analyzed includingchanges in annual net primary production,and inannual net carbon storage. For the biogeographymodels,the focus was on changes in the area ofmajor vegetation assemblages.

porate “current”climate parameters (includingatmospheric CO2 of 354 ppmv in 1990), existingsoil properties,a uniform vegetation classification,and two climate-change scenarios. Key databasedesign criteria include temporal consistency, withdaily and monthly climate sets having the samemonthly average. The database is also spatially con-sistent with, for example, climate and vegetationreflecting topographic effects. And finally, the data-base is physically consistent,with relations main-tained among climate variables and among soilproperties in soil profiles.

The database covers the coterminous United Stateswith a spatial resolution of 0.5°. The coterminousUnited States is made of about 3100 of the 0.5° x0.5° grid cells. The baseline vegetation is assumedto be in equilibrium under current climate. The cur-rent vegetation distribution is determined by firstdefining a “potential”vegetation distribution basedon ecophysiological and resource constraints.

VEGETATION IN THEFUTURECurrent cropland and urban areas are defined and acropland and urban “mask”is applied to the poten-tial vegetation distribution to define the extent ofcurrent natural vegetation. This same unchangedcropland and urban mask is used in throughout the


80

Models

ModeledCurrent NPP

Factorsaffecting NPP

Changes are given as deviations from “current” NPP as both absolute (Tg C/yr) and relative (%) values. Simulations arefor the period 2025-2034.

81


BIOGEOCHEMICALSIMULATION RESULTSThe three biogeochemistry models estimate conti-nental scale Net Primary Production (NPP) in natu-ral ecosystems for contemporary climate and CO2.For NPP, estimates range from 2.8 Pg C/yr to 3.5 PgC/yr. In the near term (2025-2034),all three modelsproject small increases in continental NPP for bothclimate simulations when climate and CO2 effectsare considered (Table 2). For the scenario used,theCO2 concentration in 2025-2034 averaged about 425ppmv. The magnitude of the CO2 fertilization effectin the decade 2025-2034 ranges from a low of about3% in CENTURY to a high of 18% in Biome-BGC.These sensitivities to CO2 differ from experimentalresults,in part,because most field experiments aredone at doubled pre-industrial CO2 (about 300ppmv CO2),higher than the projected levels in2025-2034 in the mid-range IPCC emissions scenarioused in this assessment. For the near-term climatesimulated by the Canadian model,both Biome-BGCand TEM suggest that without a CO2 fertilizationeffect, average annual NPP for the period 2025-2034would decline relative to current average annualNPP. This is an important point since the exact mag-nitude of the CO2 fertilization effect on NPP isuncertain for many natural ecosystems,especiallyforests.

Annual net carbon storage at the continental level isprojected by all three biogeochemistry models toincrease in the near term for both climate simula-tions,when climate and CO2 effects are considered(Table 3). The biogeochemistry models estimate

Table 3. Simulated Annual Net Carbon Storage due to Changes in Climate and CO2 in the MajorRegions of the Conterminous United States for “Today” and the Period 2025-2034 in Tg C/yr.

Current Future (2025-2034)

Canadian Hadley

Northeast 3 9 13

Southeast 14 -4 34

Midwest 6 17 27

Great Plain 14 16 16

West 22 41 16

Northwest 7 17 11

Conterminous US Total 66 96 117

Changes in Vegetation Carbon

Hadley Model 2030s

Canadian Model 2030s

Figure 1. The maps above show projections of relative changes invegetation carbon between 1990 and the 2030s for two climatescenarios. Under the Canadian model scenario, vegetation carbonlosses of up to 20% are projected in some forested areas of theSoutheast in response to warming and drying of the region by the2030s. A carbon loss by forests is treated as an indication thatthey are in decline. Under the same scenario, vegetation carbonincreases of up to 20% are projected in the forested areas in theWest that receive substantial increases in precipitation. Outputfrom TEM (Terrestrial Ecosystem Model) as part of the VEMAP II(Vegetation Ecosystem Modeling and Analysis Project) study. See Color Plate Appendix

>10% decrease

up to 10% decrease

no change

up to 10% increase

>10% increase

CurrentRegion

Results are the mean of three biogeochemistry models.

Northeast• Under both simulated climates, forests remain

the dominant natural vegetation,but the mix offorest types changes. For example,winter-decid-uous forests expand at the expense of mixedconifer-broadleaf forests.

• Under the climate simulated by the Canadianmodel,there is a modest increase in savannasand woodlands.

Southeast• Under the climate simulated by the Hadley

model, forest remains the dominant natural vege-tation,but once again the mix of forest typeschanges.

• Under the climate simulated by the Canadianmodel,all three biogeography models show anexpansion of savannas and grasslands at theexpense of forests. For two of biogeographymodels,LPJ and MAPSS,the expansion of thesenon-forest ecosystems is dramatic by the end ofthe 21st century. Both drought and fire play animportant role in the forest breakup.

Midwest• Under both simulated climates, forests remain

the dominant natural vegetation,but the mix offorest types changes.

• One biogeography model,LBJ, simulates a modestexpansion of savannas and grasslands.

Great Plains• Under the climate simulated by the Hadley

model,two biogeography models project anincrease in woodiness in this region,while thethird projects no change in woodiness.

• Under the climate simulated by the Canadianmodel,the biogeography models project eitherno change in woodiness or a slight decrease.

West• Under the climate simulated by both the Hadley

and Canadian models,the area of desert ecosys-tems shrinks and the area of forest ecosystemsgrows.

Northwest• Under both simulated climates,the forest area

grows slightly.

that today, the average carbon storage rate of 66Tg/yr. For the climate simulated with the Hadleymodel over the period 2025-2034,the biogeochem-istry models estimate an average carbon storage rateof 117 Tg/yr, almost a 100% increase relative to pres-ent. For the climate simulated with the Canadianmodel for the same period,the biogeochemistrymodels estimate an average carbon storage rate of96 Tg/yr. One particularly interesting resultbecomes apparent when the annual carbon storagedata are analyzed by regions (Table 3). For the cli-mate simulated with the Canadian model,the meanprojection of the biogeochemistry models is thatthe southeastern ecosystems will loose carbon inthe near term (Figure 1). This ecological response isconsistent with the hot,dry climate conditions themodel projects for this region during the period of2025-2034.

BIOGEOGRAPHYSIMULATION RESULTSFor both the Hadley and Canadian climate scenar-ios,the biogeography models project shifts in thedistribution of major vegetation types as plantspecies move in response to climate change (Figure2). An implicit assumption in the biogeographymodels is that vegetation will be able to move freelyfrom location to location;an assumption that maybe at least in part unwarranted because of the barri-ers to plant migration that have been put in placeon landscapes through agricultural expansion andurbanization.

The projected changes in vegetation distributionwith climate change vary from region to region(Figure 3a-f; Tables 4-9). Some of the major changesas simulated by the biogeography models for the sixNational Assessment regions of the coterminous UScan be summarized as follows:


82


83

Ecosystem ModelsCurrent Ecosystems

Canadian Model

Hadley Model

Figure 2. The models used to estimate biogeo-graphic responses to climate change in VEMAPII include LPJ, MAPSS, and MC1. These threemodels predict the local dominance of variousterrestrial vegetation forms based on: (1) eco-physiological constraints, which determine thebroad distribution of major categories of woodyplants; and (2) response limitations, which deter-mine specific aspects of community composi -tion, such as the competitive balance of treesand grasses. Though similar in some respects,these models simulate potential evapotranspira-tion and direct CO2 effects differently, and as aresult they show varying sensitivities to temper-ature, CO2 levels, and other factors. Two of themodel models, LPJ and MC1 have biogeochem-istry modules, while the third, MAPPS, does not.For both the Hadley and Canadian climate sce-narios, the biogeography models project shiftsin the distribution of major vegetation types asplant species move in response to climatechange. The projected changes in vegetationdistribution with climate change vary fromregion to region. (Source: VEMAP, 1998). See Color Plate Appendix

TundraTaiga / TundraConifer ForestNortheast Mixed ForestTemperate Deciduous ForestSoutheast Mixed ForestTropical Broadleaf ForestSavanna / WoodlandShrub / WoodlandGrasslandArid Lands

84

Figure 3(a) Under both simulated climates,forests remain the dominant natural vegeta-tion, but the mix of forest types changes. Forexample, winter-deciduous forests expand atthe expense of mixed conifer-broad-leavedforests. Under the climate simulated by theCanadian model, there is a modest increasein savannas and woodlands. See Color PlateAppendix.

Figure 3(b) Under the climate simulated bythe Hadley model, forest remains the domi-nant natural vegetation, but once again themix of forest types changes. Under the cli-mate simulated by the Canadian model, allthree biogeography models show an expan-sion of savannas and grasslands at theexpense of forests. For two of biogeographymodels, LPJ and MAPSS, the expansion ofthese non-forest ecosystems is dramatic bythe end of the 21 st century. Both drought andfire play an important role in the forestbreakup. See Color Plate Appendix.

LPJ, MC1 and MAPSSEstimates

85

Figure 3(c) Under both simulated climates,forests remain the dominant natural vegeta-tion, but the mix of forest types changes.One biogeography model, LBJ, simulates amodest expansion of savannas and grass-lands. See Color Plate Appendix.

Figure 3(d) Under the climate simulated bythe Hadley model, two biogeography modelsproject an increase in woodiness in thisregion, while the third projects no change inwoodiness. Under the climate simulated bythe Canadian Model, the biogeography mod-els project either no change in woodiness ora slight decrease. See Color Plate Appendix.

86

Figure 3(e): Under both simulated climates,the forest area grows slightly. See Color PlateAppendix.

Figure 3(f). Under the climate simulated byboth the Hadley and Canadian models, thearea of desert ecosystems shrinks and thearea of forest ecosystems grows. See ColorPlate Appendix.


87


88

LITERATURE CITEDBachelet, D.,R. P. Neilson, J. M.Lenihan,and R. J. Drapek,Climate change effects on vegetation distribution andcarbon budget in the US, Ecosystems, in review, 2000.

Daly C., D. Bachelet, J. M.Lenihan,R. P. Neilson,W.Parton,and D. Ojima,Dynamic simulation of tree-grassinteractions for global change studies, EcologicalApplications, 10(2) 449-469,2000.

EOS Webster, University of New Hampshire,http://eos-webster.sr.unh.edu/,2000.

Haxeltine, A.,I.C.Prentice,BIOME3: An equilibriumterrestrial biosphere model based on ecophysiologicalconstraints, resource availability, and competitionamong plant functional types, Global BiogeochemicalCycles, 10(4),693-709,1996.

Holland,E.A.,W. J. Parton, J. K.Delting,and D. L.Coppock,Physiological response of plant population toherbivory and their consequences for ecosystem nutri-ent flow, American Naturalist, 140, 685-706,1992.

Hunt,E.R.,Jr, F. C.Martin,and S.W. Running,Simulatingthe effect of climatic variation on stem carbon accumu-lation of a ponderosa pine stand: Comparison withannual growth increment data, Tree Physiology, 9,161-172,1991.

Hunt,E.R.,Jr.,and S.W. Running,Simulated dry matteryields for aspen and spruce stands in the NorthAmerican boreal forest, Canadian Journal of RemoteSensing, 18,126-133,1992.

Jenkins, J. C.,R.A.Birdsey, and Y. Pan,Biomass and NPPestimations for the mid-Atlantic region (USA) usingplot-level forest inventory, Ecological Applications, inreview, 2000.

Kicklighter, D.W.,et al.,A first-order analysis of thepotential role of CO2 fertilization to affect the globalcarbon budget: A comparison study of four terrestrialbiosphere models, Tellus ,51B, 343-366,1999.

Knight, D. H.,T. J. Fahey, and S.W. Running, Factorsaffecting water and nutrient outflow from lodgepolepine forests in Wyoming, Ecological Monographs, 55,29-48,1985.

Korol,R.L.,S.W. Running,K.S.Milner, and E.R.Hunt,Jr.,Testing a mechanistic carbon balance model againstobserved tree growth, Canadian Journal of ForestResearch, 21,1098-1105,1991.

Lenihan, J. M.,C.Daly, D. Bachelet,and R. P. Neilson,Simulating broad-scale fire severity in a dynamic globalvegetation model, Northwest Science, 72,91-103,1998.

McGuire,A. D., J. M.Melillo, D.W. Kicklighter,Y. Pan,X.Xiao, J. Helfrich,B.Moore, III,C. J.Vorosmarty, and A.L.Schloss,Equilibrium responses of global net primaryproduction and carbon storage to doubled atmosphericcarbon dioxide: Sensitivity to changes in vegetationnitrogen concentration, Global Biogeochemical Cycles,11,173-189,1997.

McGuire A. D., J. S.Clein, J. M.Melillo, D.W. Kicklighter,R.A.Meier, C. J.Vorosmarty, and M.C.Serreze,Modelingcarbon responses of tundra ecosystems to historicaland projected climate: Sensitivity of pan-Arctic carbonstorage to temporal and spatial variation in climate,Global Change Biology, in press,2000.

McLeod,S.,and S.W. Running,Comparing site qualityindices and productivity of ponderosa pine stands inwestern Montana, Canadian Journal of ForestryResearch, 18,346-352,1988.

Melillo, J.M.et al. Global climate change and terrestrialnet primary production, Nature 63,234-240,1993.

Metherall,A.K.,Simulation of soil organic matterdynamics and nutrient cycling in agroecosystems,Ph.D.dissertation,Colorado State University, Ft.Collins,Colorado,1992.

Neilson,R. P.,and R. J. Drapek, Potentially complex bios-phere responses to transient global warming, GlobalChange Biology, 4,505-521,1998.

Neilson,R. P., G.A.King,and G. Koerper,Toward a rule-based biome model, Landscape Ecology, 7,27-43,1992.

Neilson,R. P.,A model for predicting continental-scalevegetation distribution and water balance, EcologicalApplications, 5,362-385,1995.

Neilson,R. P.,and D. Marks,A global perspective ofregional vegetation and hydrologic sensitivities fromclimate change, Journal of Vegetation Science 5,715-730,1995.

Neilson,R.P., D. Bachelet, J.M.Lenihan and R.J. Drapek,The VEMAP Models and Potentially ComplexBiosphere—Atmosphere Feedbacks: Is There aThreshold Global,paper presented at AGU, Spring2000.

Nemani,R.R.,and S.W. Running,Testing a theoreticalclimate-soil-leaf area hydrologic equilibrium of forestsusing satellite data and ecosystem simulation,Agriculture and Forest Meteorology, 44,245-260,1989.

Ojima, D.S.,W. J. Parton, D. S.Schimel,and C.E.Owensby, Simulated impacts of annual burning onprairie ecosystems,in Fire in North AmericanTallgrass Prairies, edited by S.L.Collins and L.L.Wallace,175 pp.,University of Oklahoma Press,Norman,Oklahoma,1990.


89

Schimel, D.,et al.,Contribution of increasing CO2 andclimate to carbon storage by ecosystems in the UnitedStates, Science, 287,2004-2006,2000.

Sitch,S.,The role of vegetation dynamics in the controlof atmospheric CO2 content,doctoral dissertation,Department of Ecology, Plant Ecology, Lund University,Lund,Sweden,2000.

Thonicke,K.S.,S.Venevsky, S.Sitch,and W. Cramer,Therole of fire disturbance for global vegetation dynamics:Coupling fire into a dynamic Global Vegetation Model,Global Change Biology, in review, 2000.

Tian,H., J. M.Melillo, D.W. Kicklighter,A. D. McGuire, J.V.K.Helfrich,III,B.Moore,III,and C. J.Vörösmarty, Effectof interannual climate variability on carbon storage inAmazonian ecosystems, Nature, 396,664-667,1998.

Tian,H., J. M.Melillo, D.W. Kicklighter,A. D. McGuire,and J. Helfrich,The sensitivity of terrestrial carbon stor-age to historical climate variability and atmosphericCO2 in the United States, Tellus 51B,414-452,1999.

Tian,H., J. M.Melillo, D.W. Kicklighter,A. D. McGuire,B.Moore,III,and C. J.Vorosmarty, Climatic and bioticcontrols on interannual variations of carbon storage inundisturbed ecosystems of the Amazon Basin, GlobalEcology and Biogeography, in press,2000.

VEMAP Members,Vegetation/ecosystem modeling andanalysis project: Comparing biogeography and biogeo-chemistry models in a continental-scale study of terres-trial ecosystem responses to climate change and CO2

doubling, Global Biogeochemical Cycles , 4,407-437,1995.

Vitousek, P. M.,T. Fahey, D.W. Johnson,and M. J. Swift,Element interactions in forest ecosystems: Succession,allometry, and input-output budgets, Biogeochemistry,5,7-34,1988.

Vörösmarty, C. J.,et al.,Continental scale model ofwater balance and fluvial transport: An application toSouth America, Global Biogeochemical Cycles 3,241-265,1989.

White, J. D.,and S.W. Running,Testing scale dependentassumptions in regional ecosystem simulations, JournalVegetation Science, 5,687-702,1995.

Xiao,X., J. M.Melillo, D.W. Kicklighter,A. D. McGuire,R.G. Prinn,C.Wang, P. H.Stone,and A.Sokolov,Transientclimate change and net ecosystem production of theterrestrial biosphere, Global Biogeochemical Cycles,12,345-360,1998.

Pan,Y.,et al.,Modeled responses of terrestrial ecosys-tems to elevated atmospheric CO2: A comparison ofsimulations by the biogeochemistry models of theVegetation/Ecosystem Modeling and Analysis Project(VEMAP), Oecologia, 114,389-404,1998.

Parton,W. J., D. S.Schimel,C.V. Cole,and D. S.Ojima,Analysis of factors controlling soil organic levels ofgrasslands in the Great Plains, Soil Science Society ofAmerica Journal, 51,1173-1179,1987.

Parton,W. J., J.W. B.Stewart,and C.V. Cole,Dynamics ofC, N, P, and S in grassland soils: A model,Biogeochemistry, 5,109-131,1988.

Parton,W. J.,et al.,Observations and modeling of bio -mass and soil organic matter dynamics for the grass-land biome worldwide, Global Biogeochemical Cycles,7,785-809,1993.

Pierce,L.L.,Scaling ecosystem models from watershedsto regions: Tradeoffs between model complexity andaccuracy, Ph.D. dissertation,School of Forestry,University of Montana,146 pp.,1993.

Prinn,R.,et al.,Integrated global system model for cli-mate policy assessment: Feedbacks and sensitivitystudies, Climatic Change, 41(3/4),469-546,1999.

Reilly, J.,R.Prinn, J. Harnisch, J. Fitzmaurice,H. Jacoby, D.Kicklighter, P. Stone,A.Sokolov, and C.Wang,Multi-gasassessment of the Kyoto Protocol, Nature, 401,549-555,1999.

Running,S.W.,and J. C.Coughlan,A general model offorest ecosystem processes for regional applications,I.Hydrologic balance,canopy gas exchange,and primaryproduction processes, Ecological Modelling, 42,125-154,1988.

Running,S.W.,and G.T. Gower, FOREST-BGC,a generalmodel of forest ecosystem processes for regional appli -cations,II.Dynamic carbon allocation and nitrogenbudgets, Tree Physiology, 9,147-160,1991.

Running,S.W.,and E.R.Hunt,Jr.,Generalization of aforest ecosystem process model for other biomes,BIOME-BGC,and an application for global-scale models,in Scaling Processes Between Leaf and LandscapeLevels, edited by J. R.Ehleringer and C. Field,pp.141-158,Academic Press,San Diego,California,1993.

Running,S.W.Testing FOREST-BGC ecosystem processsimulations across a climatic gradient in Oregon,Ecological Applications 4,238-247,1994.

Sanford,R.L.,Jr.,W. J. Parton, D. S.Ojima,and D. J. Lodge,Hurricane effects on soil organic matter dynamics andforest production in the Luquillo Experimental Forest,Puerto Rico: Results of simulation modeling,Biotropica, 23,364-372,1991.


90

ACKNOWLEDGMENTSMany of the materials for this chapter are based oncontributions from participants on and those work-ing with the

Ecosystem Scenario TeamTimothy G. F. Kittel*,National Center for

Atmospheric ResearchJerry Melillo*,Marine Biological LaboratoryDavid S.Schimel*,Max-Planck-Institute for

Biogeochemistry, Jena,GermanySteve Aulenbach,National Center for Atmospheric

ResearchDominique Bachelet,Oregon State UniversitySharon Cowling,Lund University, SwedenChristopher Daly, Oregon State UniversityRay Drapek,Oregon State UniversityHank H. Fisher, National Center for Atmospheric

ResearchMelannie Hartman,Colorado State UniversityKathy Hibbard,University of New HampshireThomas Hickler, Lund University, SwedenCristina Kaufman,National Center for Atmospheric

ResearchRobin Kelly, Colorado State UniversityDavid Kicklighter, Marine Biological LaboratoryJim Lenihan,Oregon State UniversityDavid McGuire, U.S.Geological Survey and

University of Alaska,FairbanksRon Neilson,USDA Forest ServiceDennis S.Ojima,Colorado State UniversityShufen Pan,Marine Biological LaboratoryWilliam J. Parton,Colorado State UniversityLouis F. Pitelka,University of Maryland Appalachian

LaboratoryColin Prentice,Max-Planck-Institute for

Biogeochemistry, Jena,GermanyBrian Rizzo,University of VirginiaNan A.Rosenbloom,National Center for

Atmospheric ResearchJ.Andy Royle, U. S.Department of the InteriorSteven W. Running,University of MontanaStephen Sitch, Potsdam Institute for Climate Impact

Research,GermanyBen Smith,Lund University, SwedenThomas M.Smith,University of VirginiaMartin T. Sykes,Lund University, SwedenHanqin Tian,Marine Biological LaboratoryJustin Travis,Lund University, SwedenPeter E.Thornton,University of MontanaF. Ian Woodward,University of Sheffield,UK

* Assessment Team chair/co-chair


91

CHAPTER 3

THE SOCIOECONOMIC CONTEXT FORCLIMATE IMPACT ASSESSMENTEdward A. Parson1 and M.Granger Morgan2 served as Coordinating Authors for theNational Assessment Synthesis Team with contributions from: Anthony Janetos3,Linda Joyce4, Barbara Miller5, Richard Richels6, and Tom Wilbanks7


Climate Impacts and their Assessment

Climate Impacts in Socioeconomic Context:

Lessons from History

Adaptation and Vulnerability

Socioeconomic Scenarios in Impact Assessment:

Coping with Complexity

Multiple Stresses

Thresholds,Breakpoints,and Surprises

Integrated Assessment

Thinking about the Future

Appendix 1:Three Scenarios of Future Socioeconomic Conditions

Appendix 2:Template for Developing Socioeconomic Scenarios

Literature Cited

1 John F. Kennedy School of Government,Harvard University; 2 Dept.of Engineering and Public Policy, Carnegie-Mellon

University; 3 World Resources Institute; 4 US Forest Service; 5 World Bank; 6 EPRI; 7 Oak Ridge National Laboratory 93

with most climate scenarios bringing mixed effects:benefits to some people,places,and sectors,andharm to others. A system is more or less sensitive toclimate depending on whether a specified change inclimate brings large or small impacts.

The simplest framework for assessing climateimpacts involves specifying the climate change andclimate baseline,and attempting to infer impactsdirectly. The state of the society or economy thatbears the climate change is not considered (Kates,1985). Although this framework has been widelycriticized as too simplistic,it is adequate for studiesof some important impacts,which can be describedwithout detailed or explicit consideration of socioe-conomic conditions. In particular, assessments thatonly describe climate’s first-order effects on environ-mental characteristics,or biological or physicalresources whose importance to society is clearlyevident,can be conducted without explicit consid-eration of socioeconomic context. Assessments ofthis type might, for example,attempt to calculatethe effects of specified climate change on the rangeof sugar maple trees in New England,the productivi-ty of loblolly pine forests in Georgia,the expectedwheat yield in Kansas,the mean annual runoff inthe Colorado basin,the average July heat index inChicago,or the expected frequency and intensity ofstorms in North Carolina. Most assessments of cli-mate impacts conducted to date have followed thisframework. With a few exceptions,this Assessmenthas from necessity followed the same practice.

Conducting assessments using this approach is diffi-cult. It requires projecting future behavior of theclimate system,and of managed and unmanagedecological systems. These projections are challeng-ing because the systems are highly complex,interac-tive,and uncertain,and because we do not under-stand all the factors that control their operation.

But this approach, challenging as it is and useful asit can be,is not sufficient for a full assessment of cli-mate impacts that seeks to identify, describe,andvalue their effects on people,economies,and soci-eties. Climate variability and change occur in asocial and economic context that contributes todetermining impacts. In some cases,socioeconomicconditions may mediate or alter even first-order bio-physical impacts such as the examples listed above,so socioeconomic information will be necessary todescribe and assess even these impacts. The effectof a specified climate change on wheat yields orpine productivity will depend on how the farm orforest is managed,as well as on how the climatechanges. The heat index in Chicago is strongly influ -

CLIMATE IMPACTS ANDTHEIR ASSESSMENT

It is obvious,from history and everyday observation,that weather and climate can have impacts on peo-ple. Human impacts can arise from weather and cli-mate events at many scales:from individual extremeevents such as hurricanes or ice storms;from anom-alous seasons such as an unusually cold winter ordry summer;or from multi-year departures from nor-mal climate conditions,such as the drought of the1930s.

Although particular climate impacts may be clear,their mechanisms of causation can be complex andthe degree of influence climate has on humanaffairs in aggregate remains controversial. The viewthat climate determines major historical events andthe character of societies and economies,which hasbeen periodically expressed since antiquity andenjoyed perhaps excessive respect in the early 20th

century (e.g.,Huntington,1915),has fallen into per-haps excessive disrepute,although it has never beenfully refuted. More persuasive arguments for signifi-cant climatic influence on particular historicalevents or characteristics of societies continue to beadvanced (e.g.,Myrdal,1972;Bryson et al.,1974;Lambert,1975;Schneider, 1984;Diamond,1997;Sachs,1999),but it remains the case that the aggre-gate degree and mechanisms of climatic influenceon human affairs are not fully understood(Riebsame,1985).

Given an assumed state of America’s society andeconomy, the impacts of a specified weather or cli -mate event are the changes it induces in matters ofhuman concern. Defining climate impacts aschanges implies an alternative,the baseline climateagainst which changes are measured. For studyingthe impacts of climate change,the baseline is nor-mally assumed to be continuation of the climate ofthe past few decades. Describing impacts alsorequires specifying the perturbed climate,whoseeffects relative to the baseline are to be measured.Methods for specifying such hypothesized changesin climate,through model projections and historicalanalogs,are discussed in Chapter 1.

A specified climate change may have multipleimpacts. For example,an unusually warm wintercan have diverse impacts on home heating bills,driving safety, recreational opportunities,ski areaprofitability, and the over-wintering of household orcrop pests. Impacts may be beneficial or harmful,

94


95

Chapter 3 / Socioeconomic Context for Climate Impact Assessment

Reasonable judgments can be drawn about whatkinds of futures are more or less likely, but causallaws of society and history – if they should exist atall – are not known.

The central place of socioeconomic conditions indetermining impacts requires that they be consid-ered,and for many analyses,be explicitly projected.But the profound limits to our knowledge of the fac-tors that determine socioeconomic change requirethat explicit acknowledgment of uncertainty be cen-tral to such projections. This requirement cannot bemet by assuming any single socioeconomic future.Rather, multiple scenarios representing a plausiblerange of alternative socioeconomic futures are need-ed,ideally with explicit quantification of judgmentsabout uncertainty. The sensitivity of results to alter-native assumptions should also be examined. In par-ticular, the charge to not assume just one socioeco-nomic future applies to the widespread practice ofstudying the impacts of future climate changes as ifthey were imposed on today’s society. Although ithas long been recognized that this practice intro-duces serious biases to impact assessment,and sev-eral major studies have demonstrated the alternativeof explicit,coherent socioeconomic projections(e.g.,Rosenberg,1993),the practice remains wide-spread. This practice,often advocated in order toavoid criticism for engaging in speculation,is equiv-alent to assuming that the future society that willbear the impacts of climate change will resemblethe present in all relevant ways – an assumptionthat may be acceptable for near-term assessments,but grows increasingly unacceptable as the timehorizon lengthens. To see how wrong this assump-tion is likely to be,one need only compareAmerica’s society and economy of today to that of100,50,or even 25 years ago.

CLIMATE IMPACTS INSOCIOECONOMIC CONTEXT

Lessons from History

Looking backward a century underscores the extentto which impacts of climate depend on socioeco-nomic conditions. It also shows the severity of thechallenge posed by attempting to project socioeco-nomic conditions up to a century in the future. Atthe turn of the 20th century, most of the US work-force was employed on farms;aircraft,electronics,and antibiotics had not been invented;aluminumwas a semi-precious metal;the automobile existed

enced by the urban heat island effect,whichdepends on the size,density, and surface characteris-tics (e.g.,building, roofing,and paving materials) ofthe city. Runoff in the Colorado basin under a speci-fied climate can be altered by large-scale land-usechange in the basin,as well as by water engineeringprojects. A specified runoff event may cause a floodor may not,depending on the infrastructure pres-ent. Winnipeg survived the Red River flood thatdestroyed Grand Forks,because a large emergencyflood channel had been constructed aroundWinnipeg decades earlier.

More fundamentally, the impacts of climate changethat matter to people are not limited to direct bio-physical impacts,but can also include many indirecteffects on such factors as health,income,andemployment;the price, availability, and quality ofgoods and services;property values and losses;recreational opportunities;the character of the land-scape;and the political,social,and economic charac-ter of their community – as well as the direct effectsof weather and climate on people’s experience.Such impacts are not exclusively caused by weatheror climate,but are mediated by many characteristicsof the economy and society. They can only bemeaningfully defined relative to specified individualand collective perceptions,interests,and values,which in turn may themselves be subject to change.For example,what is the value of fall foliage in NewEngland,and what would be the impacts if itchanged? The settlement patterns and demographicstructure of the population,the prosperity andstructure of the economy, the technologies availableand in use,the patterns of land and natural resourceuse,and the institutions and policies in place will allcontribute to how – and how much – climate willmatter to people,and what they can and might wishto do about it. Climate conditions and societal con-ditions jointly cause climate impacts (Kates,1985).

Because of this joint causation,making a coherentassessment of climate impacts requires careful,sys-tematic assumptions about future socioeconomicconditions as well as future climatic conditions.However challenging it is to model and projectfuture climate,projecting future socioeconomic con-ditions is even more so. As is the case for climateand ecosystems,the nation’s economy, society,pat-terns of resource use,technology, and land use,areshaped by highly complex,interactive,and uncer-tain processes. But while most aspects of climateprojection are based on well understood physicalprocesses,our understanding of the basic structureand causal factors operating in socioeconomic sys-tems and their evolution is vastly more limited.

oping substantial predictive skill on weekly andeven seasonal intervals. Other examples includebetter roads and automobiles,navigation and instru-ment systems for aircraft and shipping,broadcastingand other forms of wireless communication,air con-ditioning and improved heating technology, newconstruction materials and techniques that haveallowed construction of huge indoor spaces,andtechnologies that have made many forms of outdoorsport and recreation (e.g.,skiing and climbing) saferand more accessible.

Technology can also increase society’s vulnerabili-ty to climate,particularly to extreme climate orweather events. This can happen because modernsocieties are organized around the available tech-nologies,and become dependent on them.Contemporary American society relies in criticalways on electric power, transportation,and com-munications systems,all of which can be disrupt-ed by extreme events if systems have not beenadequately designed to deal with them. Large-scale loss of power lines in an ice storm can havecatastrophic effects on a modern industrial socie-ty, even though all societies,including early industri-al ones,functioned without widespread electricalservice only a century ago.

US population has not simply grown in the past cen-tury, it has also shifted markedly in its demographicstructure and its distribution around the country.These trends have also shaped patterns of sensitivityto climate. For example,the US population is grow-ing older. The fraction of Americans aged 65 or overhas increased from 1 in 25 in 1900 to 1 in 8 in2000. Older people are physiologically more vulner-able to heat stress. Without adaptive measures, amore aged society will be more vulnerable toincreases in heat-related illness and death under awarmer climate. A warmer climate may also bring areduction in cold-related mortality, a trend that willalso interact with the aging of the population,although the effect of temperature changes on mor-tality appears to be weaker for cold conditions thanfor hot. Recent migration to the South andSouthwest demonstrates that many older Americansprefer warmer climates,although the nearly univer-sal spread of one technology – air conditioning –has played an essential role in allowing the rapidgrowth of these regions. At the same time, rapidpopulation and economic growth in arid parts ofthe Southwest has sharply increased vulnerability towater shortages.

96


only as a primitive novelty;and the predominantform of transportation – and the predominant urbanenvironmental problem – was the horse. Over theintervening century, the population of the UnitedStates nearly quadrupled,from 76.2 million to 275million (US Bureau of the Census,1998),while USreal GDP increased more than thirty-fold,from justunder $300 Billion to about $9.5 Trillion (1996 dol-lars,Bureau of Economic Analysis,2000) – corre-sponding to a nearly ten-fold increase in real percapita income.

These increases in material welfare,and the processof industrial transformation that drove them,havehad profound effects on the nation’s relationshipand sensitivity to climate. As first the industrial sec-tor and later the services sector grew to dominatethe American economy, fewer Americans’livelihoodshave been directly tied to climate. Moreover,wealthier nations – like wealthier individuals – arein general better able to cope with the negativeimpacts of climate variability and change,and betterable to take advantage of the opportunities theypresent. Wealthy societies can spare resources tosupport adaptation,can better afford to makerequired changes in technology and infrastructure,and can more easily endure climate-related losses.Within societies, climatic harms and opportunitieswill not be equally distributed among individualsand communities:some will face greater burdensthan others. Moreover, high rates of economic andpopulation growth can themselves impose stresseson natural systems,through rising pollution (includ-ing greenhouse gases),congestion,and demands forland and resources,potentially increasing these sys-tems’vulnerability to climatic stresses.

Much of our recent prosperity has been fueled bynew technology. Although technological change canalso carry significant social and environmental costs,in aggregate it has greatly contributed to Americans’increased material well-being over the 20th century.For example,in the past decade,computers andnew communication technologies have transformedmany activities,bringing increases in productivity aswell as new products and services.

Technology affects society’s relationship to climatein many ways. Technological change will stronglyinfluence the success of future efforts to controlgreenhouse gas emissions. Many technologicalchanges,large and small,have reduced Americans’vulnerability to weather and climate in a host ofways. A striking example has been weather and cli-mate forecasting,which with increasing understand-ing of large-scale patterns of variability is now devel-

97


America is also becoming more urban. Over the20th century the fraction of Americans living incities increased from 40% to more than 75% (USBureau of the Census,1999). Urbanization affectsclimate vulnerabilities and capacity for adaptation inmultiple and complex ways. City dwellers dependless on climate-sensitive activities for their liveli-hoods,and have more resources and social supportsystems close at hand. But the dense concentrationof people and property in coastal or riverside met-ropolitan areas,dependent on extensive fixed infra-structure such as water, sewer, and energy utilities,and roads,tunnels,and bridges (which are aging andoverburdened in many US cities),can increase vul-nerability to extreme events such as floods,stormsurges,and heat waves. Combined with other urbanstresses such as congestion,pollution,and the urbanheat island effect, climate change could significantlyharm urban quality of life and health.

Americans are also moving to the coasts. Some 53%of the total US population now live in the 17% ofland area that comprises the coastal zone,and thelargest continuing population increases for severaldecades are projected to be in coastal areas. Thistrend is exacerbating wetland loss and coastal-zonepollution. In addition,locating more people andmore valuable property in low-lying coastal areasincreases vulnerability to storms,storm surges,coastal erosion,and sea-level rise – as severe recentlosses in Florida,Georgia,and the Carolinas,as wellas a century of damage trends,all confirm(Changnon et al.,2000).

Observing past patterns of climate impacts revealshow America’s vulnerability to climate and its capac-ity for adaptation have depended on many highlydetailed and specific characteristics of its economyand society. For particular communities or activi-ties,the most important factors shaping climate vul-nerability might be as diverse as local zoning ordi-nances,housing styles,or building codes;popularforms of recreation;the age and degree of specializa-tion of capital in particular industries; world marketconditions;and the distribution of income. Forexample,the vulnerability of American agriculture topast climate extremes has been shaped by a host ofsocioeconomic factors,including the size and struc-ture of farm families, agricultural practices and avail-able technologies,markets for alternative crops,available capacity for storage and transport, ground-water accessibility, local and nationwide markets forcapital and labor, bank lending practices and thenationwide organization of banking and capital mar-kets,global trade rules,and public policies.

Over the 21st century, population and demographicstructure,settlement patterns,economic output andstructure,technology, policy, and other social andeconomic factors will continue to affect the easewith which American society can adapt to,or takeadvantage of, climate variability and change.Continuing income growth and continuing develop-ment of new technologies remain likely, in aggre-gate,to reduce our vulnerability to climate. But asin the 20th century, specific climate impacts and vul-nerabilities in the 21 st century are likely to remaindependent on many detailed and specific character-istics of America’s society, with the particular factorsthat turn out to be most important not evident inadvance. Moreover, the changes in these factorsover the 21st century are likely to be at least as great,and at least as unpredictable in their details,as thechanges that took place over the 20th century.

ADAPTATION ANDVULNERABILITYPeople need not merely suffer the climate condi-tions they face,but can change their practices,insti-tutions,or technology to take maximum advantageof the opportunities the climate presents and tolimit the harms they suffer from it. Through suchadaptations, people and societies (like ecosystems)adjust to the average climate conditions,and thevariability of conditions they have experienced inthe recent past. Present climates are not tuned tomaximize human welfare,of course,so some poten-tial changes might be purely benign (e.g.,if therewas a reduction in maximum hurricane windspeeds). But when habits,livelihoods,capital stock,and management practices are finely tuned to cur-rent climate conditions,the direct effect of manytypes of change in these conditions,particularly ifthe change occurs rapidly, is more likely to be harm-ful and disruptive than beneficial.

But just as societies adapt to the present climate,they can also adapt to changes in it. Adaptation canbe intentional or not,and can be undertaken eitherin anticipation of projected changes or in reactionto observed changes. Society’s capacity to adapt tofuture climate change is a crucial uncertainty indetermining what the actual consequences of cli-mate change will be. Societies and economies arevulnerable to climate change if they face substantialunfavorable impacts,and have limited ability toadapt. Like impacts themselves,the set of optionsand resources available to adapt to change,and theability of particular individuals,communities,and

98


societies to adopt them,depend on complex sets oflinked social and economic conditions. Such factorsas wealth,economic structure,settlement patterns,and technology play strong roles in determining vul-nerability to specified climate conditions (Downinget al.,2000).

Human societies and economies have demonstratedgreat adaptability to wide-ranging environmentaland climatic conditions found throughout theworld,and to historical variability. Wealthy industri-al societies like the US function quite similarly insuch divergent climates as those of Fairbanks,Alaskaand Orlando,Florida. While individual adaptabilityalso contributes,it is principally social and econom-ic adaptations in infrastructure,capital,technology,and institutions that make life in Orlando andFairbanks so similar that individual Americans canmove between them (in either direction) with atmost moderate discomfort.

But adaptability has limits, for societies as for indi-viduals,and individuals’ ability to move throughlarge climate differences tells us little about theselimits. Moving between Fairbanks and Orlando mayonly be uncomfortable,but rapidly imposing the cli-mate of either place on the other would be very dis-ruptive. The countless ways that particular localsocieties have adapted to current conditions andtheir history of variability can be changed,but notwithout cost,not all with equal ease,and notovernight. The speed of climate change,and its rela-tionship to the speed at which skills,habits,resource-management practices,policies,and capitalstock can change,is consequently a crucial contribu-tor to vulnerability. Moreover, however wisely wemay try to adjust long-lived decisions to anticipatecoming climate changes, we will inevitability remainlimited by our imperfect projections of the comingchanges. Effective adaptation may depend as muchon our ability to devise responses that are robust tovarious possible changes,and adjustable as we learnmore,as on the quality of our projections at any par-ticular moment. While societies have shown sub-stantial adaptability to climate variability, the chal-lenge of adapting to a climate that is not stable,butevolving at an uncertain rate,has never been testedin an industrialized society.

While adaptation measures can help Americansreduce harmful climate impacts and take advantageof associated opportunities,one cannot simplyassume that adaptation will make the aggregateimpacts of climate change negligible or beneficial.Nor can one assume that all available adaptationmeasures will necessarily be taken. Even for such

well-known hazards as fire, flood,and storms,peo-ple often fail to adopt inexpensive and easy risk-reduction measures in their choices of buildingsites,standards,and materials – sometimes withgrave consequences. In this first NationalAssessment,potential climate adaptation optionswere identified,but their feasibility, costs,effective-ness,and the likely extent of their actual implemen-tation were not assessed. Careful assessment ofthese will be needed.

SOCIOECONOMICSCENARIOS IN IMPACTASSESSMENT

Coping with Complexity

One way to assemble the socioeconomic assump-tions needed for impact assessment is to constructscenarios. Scenarios are coherent,internally consis-tent,and plausible descriptions of possible futurestates of the world,used to inform investigations offuture trends,potential decisions,or consequences(IPCC,1994). Scenarios can be simple or complex,quantitative or qualitative,stochastic or determinis-tic,and can provide variable levels of detail accord-ing to their purpose. In most usage,scenarios areexogenous to the analysis:they describe aspects ofthe world that must be specified for the analysis toproceed,but which are simply assumed,not calcu-lated within the analysis.

In assessments of climate change,the craft of devel-oping and applying scenarios is most advanced forthe scenarios of future greenhouse-gas emissionsused to drive climate models. Scenarios for thelargest sources of emissions can be developed byprojecting a few aggregate characteristics of thenation or region being considered,such as popula-tion,economic growth,and changes in the energyintensity and carbon intensity of economic output(Nakicenovic and Stewart,2000). While projectionsof these variables may have wide uncertaintyranges,they can be based on widely available consis-tent historical data and their complexity is not over-whelming. Moreover, because emission trendsdepend jointly on trends in population,economicgrowth,and technological change,it is possible togenerate a wide range of emissions futures whileconsidering only a narrow, largely benign range ofpopulation and economic futures, by making widelydivergent assumptions of technical change.

99


The decentralized track was to be used when partic-ular analyses required specifying future values ofmore specific or local socioeconomic characteris-tics. In such cases,the relevant assessment teamswere asked to develop and document the requiredassumptions themselves. A common template wasprovided to guide teams in developing scenarios,which involved identifying two or three key charac-teristics they judged to have the most direct effectson the impact of interest,constructing uncertaintyranges for these characteristics,and varying themjointly through their ranges. In addition,two back-ground papers were provided that reviewed alterna-tive methods and attempts at projecting futuretrends in technology and institutions (Patt et al.,1998;Wilbanks,1998). The template for the decen-tralized track is described in Appendix 2 of thischapter.

Teams were also requested to attempt an alterna-tive, exploratory approach to projecting impacts in2100,which would avoid the need for 100-yearsocioeconomic projections. This exploratoryapproach involved reversing the relationshipbetween assumed socioeconomic futures and cli -mate impacts. The standard approach used through-out the Assessment involved assessing the impact ofa specific climate scenario under a specific futuresocioeconomic scenario. Instead,this alternative,exploratory approach involved specifying only afuture climate scenario,and trying to identify plausi-ble socioeconomic conditions that would make forlarge variation in the impacts of this specified cli-mate. For the region or sector in question,whatpotential future socioeconomic conditions mightmake this climate change seriously harmful? Whatconditions might make it insignificant? What condi-tions might make it greatly beneficial? The purposeof this alternative approach was to engage teams ina more open-ended process of thinking throughpotential socioeconomic futures,to scout for poten-tial vulnerabilities and opportunities that mightescape notice in a more conventionally structuredinquiry.

In this first Assessment,the region and sector teamsmade very limited use of the socioeconomic scenar-ios and template provided. In some cases,such asthe Human Health sector, teams judged the state ofknowledge in their domain insufficient to supportany prospective,scenario-based analysis. In severalother cases,analyses projected only first-order bio-physical impacts such as changes in forest produc-tivity or streamflow, for which no socioeconomicassumptions were needed. The few analyses thatattempted to project impacts further down the

Developing scenarios for assessment of impacts is afundamentally different and more complex problem,on which less experience is available. No simpleaggregate technical coefficients are known,analo-gous to energy intensity or carbon intensity in emis-sions scenarios,which would largely define impacts.Indeed,impacts and vulnerability are likely todepend on highly specific,detailed,often local char-acteristics of particular communities or activities, forwhich reliable consistent data are unlikely to beavailable – if we even knew what the relevant char-acteristics were. Also in contrast to emission scenar-ios,scenarios for impact assessment must considerthe possibility of sustained low economic growth aswell as high,since income and wealth are likely tobe important determinants of vulnerability andadaptive capacity.

A working group of the NAST was charged withdeveloping scenarios for the socioeconomicassumptions necessary for the Assessment. Becauseof the complexity and diversity of the socioeconom-ic characteristics that might be important determi-nants of impacts and vulnerability, and because ofthe highly decentralized nature of the NationalAssessment process,this working group judged itinfeasible to attempt to develop fully detailedsocioeconomic scenarios centrally. To do so wouldamount to trying to predict a century of Americanhistory. Moreover, such an attempt would be inap-propriate because the determinants of impacts arelikely to vary among regions,and identifying themost important ones is likely to require detailedregional expertise. Rather, the working groupattempted to balance the Assessment’s competingneeds – to reflect regional concerns and expertisewhile maintaining enough consistency to allownational-level synthesis – by recommending a two-tracked approach to scenario development,partlycentralized and partly decentralized.

The centralized track comprised a few key socioeco-nomic variables likely to influence many domains ofimpact,such as population,economic output,andemployment. For these,where nationwide consis-tency was most important,the working group devel-oped three internally consistent socioeconomic sce-narios,which were used in all region and sectorstudies in the Assessment. The three scenariosspanned a wide range of high- and low-growthfutures. Projections of population,income,andemployment were provided in substantial detailthrough 2050 – by county and by thirteen econom-ic sectors – and at the national level through 2100.These scenarios are described in Appendix 1 of thischapter.

causal chain to effects on humans only required,orcould only effectively use,the scenarios of econom-ic and population growth specified by the central-ized track. No analysis in this Assessment used thefull template for socioeconomic scenario develop-ment discussed in Appendix 2. The limited use ofsocioeconomic scenarios in this first Assessment haslimited the extent to which impacts can bedescribed or assessed in terms of human relevance– e.g.,in terms of monetary loss or gain, valuation ofnon-market changes,or the incidence of suchextreme events as bankruptcy, property loss orabandonment,or regional economic booms orbusts. Further developing,testing,and applyingsuch methods for constructing scenarios sufficientlyrich and detailed to do impact assessment,but thatstill sufficiently limit complexity and maintainenough consistency to permit aggregation,will be akey methodological and research challenge for sub-sequent assessment of climate impacts.

The approach taken in this Assessment to projectingsocioeconomic futures has obvious limitations. Onthe one hand,it represents a vast simplification ofthe linked climatic,ecological,economic,and socialprocesses that will actually determine climateimpacts and adaptive capacity. On the other hand,itis so complex and difficult to implement,that noanalysis undertaken as part of this first Assessmentwas able to follow the template fully. Still,this gen-eral approach of combining central guidance onover-arching assumptions with structured use ofdecentralized expertise for other assumptions hasallowed us to make a start. It has allowed thisAssessment to take a first look at climate impacts,more detailed and consistent than has hitherto beenconducted,which can be refined and extended assubstantive knowledge and assessment methods areprogressively improved.

MULTIPLE STRESSESHuman society has imposed various stresses on theenvironment,at diverse spatial scales, for centuries.Over the 21st century, climate change will occur inparallel with,and be jointly determined with,manyother environmental stresses and many other formsof change. The same social,economic,and ecologi-cal systems that will bear the stress of climatechange will also often be bearing other simultane-ous stresses. These will include environmentalstresses such as air pollution,acid deposition,coastaland estuarine pollution,loss of habitat and naturalecosystems,and unsustainable exploitation ofmarine resources. They will also include broader

100


socioeconomic stresses such as rapid shifts in tech-nology and world market conditions,potentialincreases in migration and in economic inequality,and overloading of infrastructure in rapidly growingmetropolitan and coastal regions. Other technologi-cal,economic,institutional,or social trends mayhelp to increase systems’adaptability and mitigatethe effects of climate change and other stresses. Asclimate varies and changes,so will these other fac-tors.The aggregate impacts on ecological,economic,and social systems will reflect the joint applicationof multiple environmental and other stresses,as wellas potential interactions between them.

For most US ecosystems,other stresses currentlygreatly exceed those arising from climate. PacificSalmon populations are predominantly stressed byfishing,dams,and watershed alteration. Maple treesin New England are predominantly stressed by pestsand air pollution. Endangered species are predomi-nantly stressed by loss of habitat. Over the comingdecades,some non-climatic stresses are likely todecline while others increase. For example,increas-ingly strict emission controls are likely to reduceacidifying pollution,while larger and wealthier pop-ulations are likely to increase the stresses that devel-opment,land-use conversion,pollution,and recre-ation impose on forests,mountain regions, wetlands,and coastlines. Although non-climatic stressesexceed climatic ones for most systems at present,one cannot assume that this will remain so – partic-ularly for natural ecosystems,which in general aremuch more dependent on climate than socioeco-nomic systems. People may move with little dis-comfort between Alaska and Florida,but speciesadapted to the climate of one of these States couldnot survive in the other:a Martin or an Arctic Terncould not live in the wild in Florida,nor a GreatWhite Heron or a Manatee in Alaska. Moreover, cli-mate variability is already a discernible stress forsome systems. A changing climate,interacting withother environmental and socioeconomic trends,islikely to become an increasingly important stress formany more systems. A system already bearing multi -ple stresses at high levels is likely to be less able,other factors being equal,to adapt to climatechange. This observation is likely to apply not justto natural ecosystems,but also to managed ecosys-tems and communities,such as marginal agriculturallands or resource-dependent communities suf feringjob loss and out-migration.

Although it is likely that interactions among multi-ple stresses will be key dimensions of socioeconom-ic and ecological vulnerability, our current knowl-edge of how stresses interact is very limited. This

first National Assessment has only been able toundertake the most preliminary investigation ofinteractions and multiple stresses. Several specificexamples of multiple-stress effects have been identi-fied as high priority needs for research and analysis,in order to improve our ability to analyze andrespond to multiple stresses in future assessments.

THRESHOLDS,BREAKPOINTS, ANDSURPRISESThe response of many systems to external changesis continuous:if you touch the accelerator, the carspeeds up a little;if you touch the brake,it slowsdown a little. Many of the analyses of climateimpacts discussed in this Assessment assume suchcontinuous responses,so the projected impacts areoften extensions of processes and trends that arealready underway today.

Sometimes,however, systems can respond in highlydiscontinuous or nonlinear ways:if you tighten thepropeller of the rubber-band airplane by one moreturn,it can break into a dozen pieces. While manynatural and social systems are likely to respondgradually to climate change, responses can also besudden if a small additional stress pushes the systemover a threshold or breakpoint.

Such discontinuities or surprises can be seen clearlyafter they happen,and attempting to explain themoften generates important advances in our under-standing,but they are extremely difficult to predict.It is imperative to remember that complex climatic,ecological,and socioeconomic systems might sur-prise us – by sudden or discontinuous response,orby evolving in some other way quite different fromwhat we expect. We have been surprised by envi-ronmental and socioeconomic changes many times.Environmental examples include the failure of rainto “follow the plow”in the 19th Century AmericanWest,the appearance (and cessation) of the 1930sdrought,and the 1980s appearance of the Antarcticozone hole. Several possible surprises and disconti-nuities have been suggested for the Earth’s atmos-phere,oceans,and ecosystems.

Equivalently high-consequence surprises could alsoarise in socioeconomic systems,causing emissions,impacts,or vulnerability to be markedly differentthan we expect. Potential candidates for such sur-prises might include rapid development and deploy-

101


ment of technologies for non-fossil energy or carbonsequestration – which could greatly reduce futurecarbon dioxide emissions – or for coal-based syn-thetic fuels,which would greatly increase emissions.Other candidates include exhaustion of major rein-surance pools from high weather-related casualtylosses,leading to financial destabilization;or climate-related emergence of new epidemic diseases. Stillmore potential for surprise arises from the intrinsicunpredictability of human responses to the chal-lenges posed by climate change.

Even if the probability of any particular surpriseoccurring is low – which is widely assumed,butmay or may not be true – potential surprises are sonumerous and diverse that the likelihood of at leastone occurring is much greater. We do not knowhow far the climate system,or the systems it affects,can be perturbed before they respond in quiteunexpected ways. As with multiple stresses,in thisfirst Assessment we have only been able to identifythis possibility and conduct some preliminary spec-ulation. Potential large-consequence surprises pres-ent some of the more worrisome concerns raised byclimate change,and pose some of the greatest chal-lenges for policy and research.

By their very nature,surprises are unpredictable.But two broad approaches can help us prepare tolive with a changing and uncertain climate, evenconsidering the possibility of surprise. First,someof our assessment ef fort can be devoted to identify-ing and characterizing potential large-impact events,even if we presently judge their probability to bevery small. Second,society can maintain a diverseand advancing portfolio of scientific and technicalknowledge,and conditions that encourage the cre-ation and use of new knowledge and technology.Continually advancing knowledge and technology,and the social,economic,and policy conditions thatsupport them,provide a powerful foundation foradapting to whatever climate changes might come.

INTEGRATED ASSESSMENTThinking About the Future

Multiple climatic and socioeconomic characteristicsjointly determine climate impacts. Further complex-ity arises from the fact that the multiple socioeco-nomic factors likely to determine impacts and adap-tive capacity all influence each other, and are in turninfluenced by patterns of environmental change.Patterns of population growth,technologicalchange,economic growth,and structural change all

century. Barring major wars or other catastrophes,US population growth is likely to continue,thoughat a declining rate,moving toward a stable or nearlystable population in the second half of the 21st cen-tury. The population is also likely to continue togrow older for several decades or more,dependingprincipally on the balance between immigration andincreased life expectancy, and is projected to contin-ue present trends of moving to metropolitan areasand the coast. Income and employment growth areprojected to move with the people to the cities andcoasts,and to continue the long-standing shiftamong sectors away from agriculture, resources,andprimary industry and toward technology, trade,andservices.

The Assessment focused on two target dates,2030and 2100. While both are far in the future,2030 lieswithin the range of some projection models andstrategic-planning tools,while 2100 lies beyond theuseful range of nearly all such tools. Consequently,the Assessment provided socioeconomic projectionswith substantial spatial and sector detail for 2030,but only aggregate national projections of a few keyvariables for 2100.

For 2030,the Assessment provided detailed high,medium,and low scenarios of population and eco-nomic growth. These three scenarios were based onalternative assumed trends in fertility, mortality, andmigration,labor-force participation by age group,and labor productivity, and were implemented usinga commercial regional economic growth model.(Terleckyj,1999a,1999b). This model providedannual projections of population, by sex and by five-year age cohort, for each state,county, and metropol-itan area in the United States. The NAST workinggroup specified the assumed trends in fertility, mor-tality, and migration that determined nationwidepopulation trends,while the model’s demographicmodule calculated the resultant age structure of thepopulation and its economic module distributedpeople around the nation.

In specifying these aggregate demographic trends,the working group used the Census Bureau’sassumptions for future trends in US age-specific fer-tility and mortality (US Bureau of the Census,2000),but applied a wider range of assumptions for futureimmigration. The low scenario followed the CensusBureau’s low immigration assumption,whichreduces net immigration from roughly 750,000 peryear in the mid-1990s (0.3% of population) to300,000 per year in 2000,and holds it numericallyconstant thereafter. The middle and high scenarioseach projected that recent trends of increasing

affect each other, and collectively determine thecharacter and degree of environmental stresses soci-ety imposes,including the emissions that contributeto climate change. Public policies will also con-tribute to this complex mix,both those directedtoward climate change and others. Tax policy caninfluence investments in research. Immigration poli-cy can influence the rates of both population andeconomic growth,and the cultural,educational,andeconomic mix of the population. Reliable predic-tion of such complex and uncertain processes is notpossible.

Since the early 1990s, research groups have soughtto represent linked processes of global environmen-tal change and associated ecological,economic,andsocial processes in “Integrated Assessment”modelsof global climate change. These models are intend-ed to allow consistent examination of the humancontributions to climate change and ways to miti-gate them,with the human consequences of climatechange and ways to adapt to them. They conse-quently allow coherent assessment of possibleresponses including both emissions reduction andadaptation to resultant change,and the tradeoffsbetween them. They also thereby allow consistentcomparison of uncertainties in all domains of theclimate issue – future emissions and means toreduce them,the responses of the climate system atglobal and regional scales,and the resultant impactsand means to adapt to them (Weyant et al.,1996;Parson and Fisher-Vanden,1997;Rotmans andDowlatabadi,1998).

Over the past ten years this work has yielded signifi-cant insights into the economic determinants ofemission trends,potential feedbacks between cli-mate change and managed and unmanaged terrestri-al ecosystems,and the relative contributions ofatmospheric,ecological,and socioeconomic uncer-tainty to uncertainties in future climate impacts andadvantageous responses. While the promise ofimportant further insights from such work remainssubstantial,the characterization of impacts andadaptation remains the weakest element of integrat-ed assessment at present.

APPENDIX 1:Three Scenarios of FutureSocioeconomic Conditions

The three socioeconomic scenarios developed bythe NAST working group all assumed that broad 20th

century trends of population and economic growthare likely to persist,to varying degrees,in the 21st

102


immigration will continue,using growth trendsderived from two different recent periods. The mid-dle scenario took the average trend in the ratio ofannual net immigration to current population thathas prevailed over the past thirty years (1967-1997),and projected this trend forward until 2025.Projected in this way, net immigration reaches 0.46%of the population in 2025,and is held constant atthis fraction of population thereafter. The high sce-nario differed from the middle only in that it calcu-lated the trend in the ratio of immigration to popu-lation over the most recent ten years (1987-1997), aperiod of particularly rapid immigration growth.Projecting this steeper trend forward,net immigra-tion reaches 0.86% of population in 2025,and isheld at that fraction of population thereafter. In allthree scenarios,the aging of the post-war babyboom generation brings sharp increases in the frac-tion of older Americans. The fraction of Americansaged 65 or over begins to surge after 2010 from itspresent value of 12.5%, reaching 20% by 2030. Stillgreater increases are projected in the fraction ofAmericans in the oldest age groups. The fraction ofAmericans aged 85 and over is projected to triple to4.5% by 2050,while those 100 and over are project-ed to increase seven-fold,to 0.2%.1

The three scenarios also provided projections ofemployment and income for thirteen major eco-nomic sectors (including three government sectors),with the same level of spatial detail – by state,met-ropolitan area,and county. As in the case of popula-tion,the NAST-specified assumptions determinednationwide economic trends – in this case,trends innationwide labor productivity and age-specificlabor-force participation rates – while the distribu-tion of employment and income among locationsand economic sectors was calculated internally bythe model.

Rates of labor-force participation were varied onlyfor older workers. For workers under 55,participa-tion rates were held at present levels. For workersaged 55 through 64,all three scenarios projectincreases in participation that extend recent trends,reaching a higher, constant level in 2025. Only thehigh scenario differs,in projecting an increase inparticipation for workers 65 and over, which alsolevels off in 2025. For productivity, the middle sce-nario assumes a continued constant increase of 1.2%annually in real economic output per worker, equalto the average annual increase over the 20th century.The high and low scenarios double and halve this

rate of productivity growth respectively, to 2.4% and0.6% per year.Considering three alternative trends for populationand productivity growth and two for labor-force par-ticipation yields eighteen possible combinations. Ifpracticality dictates using only a few scenarios,onlya small subset of these possible combinations can beconsidered. The recent IPCC scenario exercise faceda similar but much more complex problem of select-ing scenarios from a large set of combinatorial possi-bilities,because their scenarios included diversegrowth trends for multiple world regions. To reducethis complexity, they constructed narrative story-lines that provided broad political and social contextfor particular worldwide patterns of population andeconomic growth (Nakicenovic and Stewart,2000).

In constructing the three scenarios for thisAssessment,high population growth and high eco-nomic growth were combined in the high scenario,while low population and economic growth arecombined in the low scenario. This combination ofhigh population with high economic growth wouldnot be appropriate in constructing scenarios for theworld as a whole,because historical evidence anddemographic theory both suggest that higher ratesof economic growth are associated with lower ratesof population growth. This situation is reversed,however, for projections of American growth in the21st century, because most of the variation in popula-tion growth arises from variation in the assumedlevel of immigration. In contrast with natural popu-lation increase,immigration tends to follow econom-ic opportunity, so the pairing of high populationgrowth with high economic growth is more plausi-ble than the reverse.

For 2100,the more distant target date of theAssessment,a much less detailed set of socioeconom-ic projections was specified. Over this time horizon,the likelihood of fundamental changes in economicstructure,technology, and culture are likely to renderthe incremental methodology of the near-termregional economic model invalid. Indeed,anyattempt to specify century-scale socioeconomictrends with the spatial and sector detail the NPAmodel provides for the near term is likely to be ludi-crous. No detailed assumptions for the regional andsector distribution of population,employment,orincome were specified for the second half of the 21st

century.

Rather, a simple aggregate set of nationwide projec-tions of population,employment,and GDP for the USwere developed. As in the case of the more detailedprojections through 2030,three scenarios were

103


1These figures are for the middle scenario. In the Census Bureau’s

highest scenario – which is not identical to the Assessment high sce-nario because the Census assumes less immigration and consequentlyan older population – centenarians reach 0.75% of the total populationby 2050.

104


Table 1: Scenarios of US Population (Millions)

1997 Growth rate, 2030 Growth rate, 21001995-2030 2030-2100

Present Population 268

Low Scenario 0.39% 305 0.21% 353

Middle Scenario 0.86% 356 0.47% 494

High Scenario 1.21% 398 0.68% 640

Table 2: Scenarios of US GDP (Trillions of 1992 Dollars)

1997 Growth rate, 2030 Growth rate, 21001995-2030 2030-2100

Present GDP $7.2

Low Scenario 1.1% $10.3 0.9% $19.2

Middle Scenario 2.1% $14.4 1.4% $39.2

High Scenario 3.7% $24.1 2.25% $114.7

developed that combined high,medium,and lowpopulation and economic growth. Each of these sce-narios was constructed to track the growth of nation-al population and output in the corresponding moredetailed scenario for the near term,then to convergein growth rates of both population and productivityover the second half of the 21st century. The scenar-ios were constructed using a simple reduced-formintegrated-assessment model (Scott et al,1999),andwere broadly consistent with three of the “marker”scenarios developed for the IPCC Third AssessmentReport (Nakicenovic and Stewart,2000).2

The assumptions specified for these long-run scenar-ios are as follows. Population growth rates in thethree scenarios converge beginning in 2050,untilthey become equal in 2075 and follow a commonpath thereafter, declining from 0.35% per year in2080 to 0.15% in 2100. Aggregate rates of labor-forceparticipation also converge after 2050,but not to fullequality, reaching 75%,77.5%,and 80% in the threescenarios by 2100. Finally, annual growth rates ofeconomic output per worker begin converging anddeclining after 2050, reaching 1.12% per year in 2075and remaining at that level through 2100.

The consequences of these assumptions for US popu-lation and GDP are shown in Tables 1 and 2,andFigures 1 and 2, for both target years 2030 and 2100.

US population is projected to reach 356 million by2030 in the middle scenario (corresponding to anaverage growth rate of 0.86% per year between1995 and 2030),with a range of 305 to 398 millionin the low and high scenarios (0.39% to 1.21% annu-al growth). By 2100,the US population has reached494 million in the middle scenario (average annualgrowth of 0.47% from 2030 to 2100),with a rangefrom 353 to 640 million (0.21% to 0.68% averageannual growth). United States GDP grows from itspresent $7.2 trillion to $14.4 trillion by 2030 (range$10.3 to $24.1 trillion),and to $39.2 trillion by 2100(range $19.2 to $114.7 trillion).In terms of GDP perperson,these scenarios give a range from $33,800to $60,600 in 2030 ($40,500 in the middle sce-nario),and from $54,400 to $179,200 in 2100($79,400 in the middle scenario).3

APPENDIX 2: Template for developing socio-economic scenarios

If a team required more detailed or specific socioe-conomic assumptions to conduct an analysis thanwere provided in the centrally defined scenarios,they were asked to develop and document themusing a common template,as follows.

3All monetary figures are expressed in 1992 dollars.

2The correspondence with IPCC scenarios is not exact,in part

because of differences in time-steps and the definition of regions (theUS is not a complete region in the IPCC scenarios).

105


cussed above. In constructing these ranges,partici-pants were cautioned to draw them wide,seeking tomitigate people’s widely known tendency to be tooconfident in estimating unknown quantities(Kahneman et al.,1982;Morgan and Henrion,1990).

First,each team was to select a few key issues theyjudged would be most important for their region orsector, or would best illustrate important patterns ofimpact. These are the “key issues”discussed in eachof the regional and sector chapters. Second, foreach key issue the team was asked to identify oneor two key socioeconomic factors, such as specificaspects of development patterns,land use,technolo-gies,or market conditions,that they judged likely tohave the most direct influence on climate impacts,capacity for adaptation and vulnerability for thatissue. In choosing their key issues and key socioeco-nomic factors,each team was requested to use what-ever combination of preliminary analysis, expertjudgment,and stakeholder consultation they judgedmost appropriate. They were then to examine theimpacts of specified climate-change scenarios ontheir key issues,under a range of values for theirchosen socioeconomic factors. If they identifiedmore than one key socioeconomic factor, they wereasked to construct a few alternative socioeconomicscenarios by varying the factors jointly betweenhigh and low values. Other than the few key factorsthey chose to vary, any other required socioeconom-ic assumptions were to be fixed at baseline or best-guess values.

The ranges chosen for key socioeconomic factorswere intended to reflect all sources of socioeconom-ic uncertainty except climate change itself and USpolicy responses to climate change. Since the pur-pose of the Assessment was to examine the ef fectsof climate explicitly, these did not need to beembedded in variation of socioeconomic inputassumptions. In contrast,the ranges were to includeclimate-related uncertainty outside the US,if theteam judged such uncertainty to matter for USimpacts. This situation might arise, for example,inestimating demand for US grain exports or immigra-tion to the US,either of which could be influencedby climate-related impacts abroad.

The template also provided some guidance in decid-ing how wide a range of values to assume for thekey socioeconomic factors. In general terms,teamswere asked to make the range wide enough to gen-erate instructive variation in impacts,but to remainwithin their judgment of plausibility. Specifically, therange chosen for any factor should correspond toroughly a 10% chance that the true value would lieabove the upper end of the range,and a 10% chancethat it would lie below the lower end. The NASTworking group followed this same guideline,con-structing ranges to capture the true value with 80%confidence,in developing the three scenarios ofaggregate US population and economic growth dis-

Scenarios of 21st Century Growth in America

Figure 1. The Assessment considered high, medium, and low sce-narios of future US population and economic growth. Future trendsin population, economic growth, and technological change will allshape our contribution to climate change, our vulnerability to it,and our ability to adapt.

0.0

50.0

100.0

150.0

200.0

250.0

300.0

350.0

1990 2000 20102020 2030 2040 2050

2060 2070 2080 2090 2100

0.00

20.00

40.00

60.00

80.00

100.00

120.00

1990 2000 2010 2020 2030 2040 2050 2060 2070 2080 2090 2100

Gross Domestic Product

1990 2000 2010 2020 2030 2040 2050 2060 2070 2080 2090 2100

High

Medium

Low

0.0

100.0

200.0

300.0

400.0

500.0

600.0

700.0

Population

High

Medium

Low

0.0

50.0

100.0

150.0

200.0

250.0

300.0

350.0

400.0

1990 2000 2010 2020 2030 2040 2050 2060 2070 2080 2090 2100

Labor Productivity

Employment

Future employ-ment projectionstrack population,and also includevarying rates ofemployment byolder workers asthe US populationages.

Through the 21st

century, US pop-ulation projec-tions range froma 20% increase toa doubling, prin-cipally due toalternativeassumptions ofimmigrationtrends.

Low, Medium, and High Scenarios

Projections of21st century eco-nomic growthspan a widerange, reflectingthe combinedeffects of trendsin population andproductivity.

The middle sce-nario holds annu-al growth in out-put per worker at1.2%, the averagerate over the 20 th

century. The highand low scenar-ios double andhalve this annualgrowth rate.

Low, Medium, and High Scenarios

High

Medium

Low

High

Medium

Low

High

Medium

Low

High

Medium

Low

LITERATURE CITEDBryson,R.A.,H.H.Lamb,and D. L.Donley, Droughtand the decline of Mycenae, Antiquity, 48,46,1974.

Bureau of Economic Analysis (BEA),Survey of currentbusiness,US Treasury Department, January 2000.

Carter,T. R.,M.L. Parry, H.Harasawa,and S.Nishioka,IPCC Technical Guidelines for Assessing ClimateChange Impacts and Adaptations, National Institutefor Environmental Studies,Tsukuba, Japan,1994.

Changnon,S.A.,R.A.Pielke,Jr., D. Changnon,R.T.Sylves,and R.Pulwarty, Human factors explain theincreased losses from weather and climate extremes,Bulletin of the American Meteorological Society, 81,437-442,2000.

Diamond, J., Guns, Germs, and Steel: The Fates ofHuman Societies, Norton,New York,1997.

Downing,T. E, R.Butterfield,S.Cohen,S.Huq,R.Moss,A.Rahman,Y. Sokona,and L.Stephen,Climate changevulnerability: Toward a framework for understandingadaptability to climate change impacts,Report toUNEP from the Environmental Change Institute,University of Oxford, Oxford,England,May 2000.

Glantz,M.T. (Ed.), Societal Responses to RegionalClimatic Change: Forecasting by Analogy, WestviewPress, Boulder, Colorado,1988.

Huntington,E., Civilization and Climate,YaleUniversity Press,New Haven,Connecticut,1915.

Kahneman, D.A., P. Slovic,and A.Tversky (Eds.),Judgment under Uncertainty:Heuristics and Biases,Cambridge University Press,New York,1982.

Kates,R.W.,The interaction of climate and society, inClimate Impact Assessment, edited by R.W. Kates, J.H.Ausubel,and M.Berberian, John Wiley and Sons,Chichester, UK,SCOPE/ICSU, chap.1,3-36,1985.

Krunkel,K.E.,R.A.Pielke,Jr.,and S.A.Changnon,Temporal fluctuations in weather and climateextremes that cause economic and human healthimpacts: A review, Bulletin of the AmericanMeteorological Society, 80, 1077-1098,1999.

Lambert,L. D.,The role of climate in the economicdevelopment of nations, Land Economics, 47, 339,1975.

Morgan,M. G.,Uncertainty analysis in risk assessment,Human and Ecological Risk Assessment, 4(1),25-39,1998.

The combination of climate scenarios and socioeco-nomic scenarios provides the raw materials forassessing climate impacts. Imposing a specified cli -mate scenario – whether derived from historicalexperience or from a model projection – on a speci-fied socioeconomic scenario,and examining itseffects relative to the baseline climate,provides afirst-order illustration of potential climatic impacts.How this difference varies among alternative cli-mate scenarios illustrates how impacts depend onuncertainty in climate. How the difference variesamong alternative socioeconomic scenarios illus-trates how socioeconomic factors shape vulnerabili-ty to impacts. How it varies with specific hypothe-sized responses (such as changes in management,policy, institutions,or infrastructure),illustrates keydecisions that may shape vulnerability and adaptivecapacity.

For example,in the Pacific Northwest,the projectedclimate scenario had warmer wetter winters anddrier summers. As one of their key issues,theNorthwest team identified the impacts of thischange on seasonal streamflow and freshwater sup-plies. In two separate analyses,they examinedimpacts of climate variability and change on the riskof winter flooding,and on the reliability of summerwater and hydroelectric supplies. For the floodinganalysis,they were able to define flooding purely byhydrological characteristics,without regard tosocioeconomic conditions. However, more detailedassessment that considered not just occurrence offloods but the damages they cause would clearlyhave to consider settlement patterns,constructionstandards,emergency response measures,and othersocioeconomic conditions as well. For the shortageanalysis,they examined impacts of summer short-ages on multiple uses under two alternative sets ofoperating policies to manage flows and allocate sup-ply (Mote et al.,1999).

106


Morgan,M. G.,and M.Henrion, Uncertainty: AGuide to Dealing with Uncertainty inQuantitative Risk and Policy Analysis, CambridgeUniversity Press,New York,1990.

Mote, P.W.,et al.,Impacts of climate variability andchange: Pacific Northwest,A report of the PacificNorthwest Regional Assessment Group for the USGlobal Change Research Program,Climate ImpactsGroup,University of Washington,Seattle,Washington,November 1999.

Myrdal, G., Asian Drama: An Inquiry into thePoverty of Nations, Random House,New York,1972.

Nakicenovic, N.,and R.Stewart (Eds.), EmissionsScenarios:Special Report of the IntergovernmentalPanel on Climate Change, Cambridge UniversityPress,New York,2000.

Parson,E.A.,and K. Fisher-Vanden,Integrated assess-ment models of global climate change, AnnualReview of Energy and the Environment, 22, 589-628,1997.

Patt,A.,B.Tomm,A. Wolfe,and T.Wilbanks, Forecastsof Institutional Change in the United States: ASummary, prepared for the U.S.NationalAssessment of Possible Consequences of ClimateVariability and Change,Oak Ridge NationalLaboratory, Oak Ridge,Tennessee,1998.

Riebsame,W. E.,“Research in climate-society interac-tion,” in Climate Impact Assessment, edited by R.W.Kates, J. H.Ausubel,and M.Berberian,SCOPE/ICSU,Wiley, Chichester, UK, chap.3,69-84,1985.

Rosenberg, N. J. (Ed.), Towards an integrated assess-ment of climate change:The MINK study, ClimaticChange (special issue), 24, 1-173,1993.

Rotmans, J.,and H.Dowlatabadi,IntegratedAssessment Modeling,in Human Choice andClimate Change, edited by S.Rayner and E.Malone,Vol.3, chap.5,291-377,Battelle Press,Columbus,Ohio,1998.

Sachs, J.,T. Panayotou,and A. Peterson, Developingcountries and the control of climate change: A theo-retical perspective and policy implications,CAERIIdiscussion paper no.44,Cambridge,MA,November1999.

Schneider, S.H.,and R.Londer, The Coevolution ofClimate and Life, Sierra Club Books,San Francisco,1984.

Scott,M. J.,R. D. Sands, J. Edmonds,A.M.Liebertrau,and D.W. Engel, . Uncertainty in integrated assess-ment models: Modeling with MiniCAM 1.0, EnergyPolicy, 27,855-879,1999.

Terleckyj, N. E.,Analytic documentation of threealternate socioeconomic projections,1997-2050,NPA Data Services,Inc.,Washington,DC,May 1999a.

Terleckyj, N. E., Development of three alternatenational projection scenarios,1997-2050,NPA DataServices,Washington DC, July 26,1999b.

Weyant, J., O. Davidson,H.Dowlatabadi, J. Edmonds,M.Grubb,E.A. Parson,R.Richels, J. Rotmans, P. R.Shukla,R.S. J.Tol,W. Cline,and S. Fankhauser,Integrated assessment of climate change: Anoverview and comparison on approaches andresults,in Climate Change 1995: Economic andSocial Dimensions of Climate Change, edited by J.P. Bruce,H.Lee,and E. F. Haites,Report of WorkingGroup 3,Intergovernmental Panel on ClimateChange (IPCC), chap.10,367-396,CambridgeUniversity Press,New York,1996.

Wilbanks,T. J.,Forecasts of technological change inthe United States:a summary, prepared for the U.S.National Assessment of Possible Consequences ofClimate Variability and Change,Oak Ridge NationalLaboratory, Oak Ridge,Tennessee,1998.

Wilhite, D.A.(Ed.),Drought: A global assessment inHazards and Disasters,Vol.2.Routledge Publishers,New York,2000.

US Bureau of the Census, Current PopulationReports, Series P23-194,“Population Profile of theUnited States: 1997,”US Government PrintingOffice,Washington,DC,1998.

US Bureau of the Census,“Aging in the UnitedStates,” http://www.census.gov/population/www-socdemo/age.html#elderly, 1999.

US Bureau of the Census, Population Projectionsthrough 2050,Document NP-T1, February 11,2000.

107


CHAPTER 4

POTENTIAL CONSEQUENCES OF CLIMATEVARIABILITY AND CHANGE FOR THENORTHEASTERN UNITED STATESEric Barron1,2


Chapter Summary

Physical Setting and Unique Attributes

Socioeconomic Context

Climate Variability and Change

Ecological Historical Data and Model Outputs

Key Issues

Vulnerability to Changes in Extreme Weather Events

Compounding Stresses on Major Estuaries and Bays

Multiple Stresses on Major Urban Areas

Recreation Shifts

Additional Issues

Forests

Agriculture

Added Insights

Adaptation Strategies


Literature Cited

Acknowledgments

1Pennsylvania State University; 2Coordinating author for the National Assessment Synthesis Team 109

Climate of the Past Century

• The Northeast has been prone to natural disas-ters related to weather and climate,includingfloods,droughts,heat waves,and severe storms.

• Temperature increases of as much as 4°F (2°C)over the last 100 years have occurred along thecoastal margins from the Chesapeake Baythrough Maine.

• Precipitation shows strong increases,with trendsgreater than 20% over the last 100 years occur-ring in much of the region. Precipitationextremes appear to be increasing while theamount of land area experiencing droughtappears to be decreasing.

• The period between the first and last dates withsnow on the ground has decreased by 7 daysover the last 50 years.


• The Northeast has among the lowest rates ofprojected future warming in comparison withthe other regions of the US.

• Winter minimum temperatures are likely to showthe greatest change,with models projectingincreases ranging from 4-5°F (2-3°C) to as muchas 9°F (5°C) by 2100,with the largest increasesin coastal regions. Maximum temperatures willpossibly increase much less,but again the largestchanges are likely to occur in winter.

• For precipitation,model scenarios offer a rangeof potential future changes,from roughly 25%increases by 2100 on average for the entireregion,to little change.

• The variability in precipitation in the coastalareas of the Northeast is likely to increase.

• Models provide contrasting scenarios for changesin the frequency and intensity of winter storms.

The Northeast is characterized by diverse water-ways, extensive shorelines,and a varied landscape inwhich weather and the physical climate are domi-nant variables. The contrasts,from mountain vistasand extensive forests to one of the most denselypopulated corridors in the US,are noteworthy. TheNortheast includes the largest financial market inthe world (New York City),the nation’s most pro-ductive non-irrigated agricultural county (Lancaster,PA),and the largest estuarine region (theChesapeake Bay) in the US. The Northeast is domi -nated by managed vegetation,with much of thelandscape covered by a mosaic of farmland and for-est. The varied physical setting of the Northeast ismatched by its highly diversified economy and bythe character of its human populations. The majori-ty of the population is concentrated in the coastalplain and piedmont regions,and within major urbanareas. The economic activities within the regionrange from agriculture to resource extraction(forestry, fisheries,and mining),to major serviceindustries highly dependent on communication andtravel,to recreation and tourism,to manufacturingand transportation of industrial goods and materials.Assessment of the impacts of climate change isbased on observed climate trends, climate simula-tions,and the importance of past extreme weatherevents.

110


CHAPTER SUMMARYRegional Context

Key Findings

• On the time scale of a century, winter snowfallsand periods of extreme cold are very likely todecrease. In contrast,heavy precipitation eventshave been increasing and warming is very likelyto continue this trend. Potential changes in theintensity and frequency of hurricanes are a majorconcern.

• Climate change is very likely to exacerbate cur-rent stresses on estuaries,bays,and wetlands inthe Northeast with rising sea level and increasingwater temperatures,with significant effects onfish populations,productivity, and human andecosystem health.

• Decreased snowfalls and more moderate wintertemperatures are very likely to lower winterstresses in major northeastern urban areas.However, climate change has greater potential toadd to existing stresses in urban areas due to theimpact of rising sea level and elevated stormsurges on transportation systems,increased heat-related mortality and morbidity associated withtemperature extremes, increased ground-levelozone pollution problems associated with warm-ing,and the impact of precipitation and evapora-tion changes on water supply.

• Typical summer recreational activities involvingbeaches or freshwater reservoirs are likely toexperience extended seasons and the region’sdiverse waterways are very likely to becomehavens for escape from increasing summer heat.In contrast,negative impacts are likely to includeinability of ski areas to maintain snow pack, mut-ing of fall foliage colors,displacement of maple-sugaring,increases in insect populations,acceler-ated beach erosion due to sea-level rise,andworsening ground-level ozone pollution prob-

lems, even in the mountains of New England.• The complex institutional framework of commu-

nities, municipal,county, regional,and statewideformal and informal governing bodies that charac-terize the Northeast is likely to limit the region’sability to deal with extremes in water supply.

• Infectious disease vectors,such as ticks and mos-quitoes,are likely to be altered by warmer andwetter conditions. Increased rainfall and floodinghave a historical association with contaminationof public and private water supplies (e.g.,withCryptosporidium).

• Agriculture in the Northeast is very likely to berelatively robust to climate change although thecrop mix may change. The ability to change croptypes and to take advantage of hybrids limits vul-nerability. Northern cool weather crops are apossible exception.

• The species composition of northeastern forestsis very likely to change under the climate scenar-ios examined in this Assessment,with significantnorthward migration of forest types.

• Projected changes in temperature and precipita-tion are likely to have a direct impact on speciesdistribution mix or an indirect impact associatedwith changing predator-prey relationships orchanges in pests or disease. In many cases,thespecies effected may be truly characteristic of aregion or may be of economic significance (e.g.,lobster, migratory birds,and trout).

• Climate change, resulting in higher temperaturesand poorer air quality, could lead to increases inheat-related mortality and morbidity and respira-tory illness. The elderly, children,those already ill,and lower-income residents are groups most atrisk.

111

Chapter 4 / The Northeastern United States

112


The socioeconomic characteristics of the Northeastvary considerably across the region. In the Mid-Atlantic region,more than 50% of the 35 millionpeople and associated jobs are concentrated in sixurban areas, for the most part within the coastalplain and piedmont regions. The Mid-Atlantic popu-lation increased nearly 20% during the last threedecades,somewhat less than the 33% populationgrowth experienced by the nation as a whole. Theworking age population increased by about 34%,the over-65 population by 72%,and the 0-19 agegroup declined by 16%. Per capita incomeincreased by 82%,with the largest growth in totalincome in the service sector (300%). Farm employ-ment declined by almost one-half, reflecting thenational trend. The economy is diverse and substan-tial. The Mid-Atlantic region alone accounts for 13%of the total US economy, with sizeable export andimport flows. Agriculture, forestry, and mining com-prise about 2% of the region’s economy, while man-ufacturing and service comprise 26% and 20%respectively.

The Northeast is characterized by a megalopolisthat extends from Boston to Washington,D.C. Thegreater New York City area alone includes parts ofthree states,31 counties,and nearly 1,600 cities,towns,and villages with more than 19 millioninhabitants. At the heart of this metropolis is acity with more than 7 million people. This popu-lation places tremendous demands on land andwater resources. Approximately 30% of the 31counties comprising the greater New York Cityland area is fully converted to urban uses,and therate of conversion has accelerated even thoughthe rate of population growth has slowed. As anexample of water demand,the amount suppliedby the New York City water system,serving thecity, most of Westchester, and some additionalcommunities,is approximately 1.3 billion gallonsa day. The area’s development is intimately con-nected with its 1,500 miles of coastline and acomplex transportation infrastructure. The gener-al economy of this metropolitan area is mostlybased on service industries,of which the econom-ic heartbeat is finance,corporate headquarters,and trade centers.

POTENTIAL CONSEQUENCES OF CLIMATEVARIABILITY AND CHANGE FOR THENORTHEASTERN UNITED STATES

PHYSICAL SETTING ANDUNIQUE ATTRIBUTES

The northeastern United States is dominated bydiverse waterways, extensive shorelines and a variedlandscape in which weather and climate are domi-nant variables. The regional contrasts and strengthsare noteworthy. Most mental images of NewEngland focus on an environment of quaint villages,mountain vistas, extensive forests,and brilliant fallcolors,maple sugaring,and skiing through forestedglades. The Northeast also includes the most dense-ly populated corridor in the US and is intimatelyconnected to the North Atlantic coastline andocean-accessible waterways. Of these waterways,the Chesapeake Bay is the largest estuarine region inthe US and is unmatched in terms of its importancefor recreation, fish,and wildlife. The Northeast isalso crossed by a remarkable network of streamsand rivers superimposed on the mountainous ter-rain of the Appalachians. The vegetative landscapehas changed dramatically over the last 100 years,inpart because of significant areas of forest re-growth.

SOCIOECONOMIC CONTEXT

The varied physical setting of the Northeast ismatched by its highly diversified socioeconomiccharacteristics (NPA Data Services,1998;Bureau ofEconomic Analysis,2000; Polsky et al.,2000;Roseet al.,2000). The Northeast includes the largestfinancial market in the world (New York City),aswell as the most productive non-irrigated agricul-tural county (Lancaster, PA) in the US. The eco-nomic activities in the region range from resourceextraction (forestry, fisheries,and mining) and agri-culture,to major service industries highly depend-ent on communication networks and travel,torecreation and tourism,to manufacturing andtransportation of industrial goods and materials.The human populations are largely concentratedin coastal and urban areas. The region containssome of the most densely populated counties inthe United States.

113


The climate of the northeastern US experienced sig-nificant changes in temperature and precipitationduring the 20th century (Hughes et al.,1992;Karl etal.,1996;Karl and Knight,1998). Based on observa-tions derived from the highest quality US observingstations as a part of the US Historical ClimatologyNetwork,temperature increases of as much as 4°F(2°C) over the last 100 years occurred along thecoastal margins from the Chesapeake Bay to Maine.Within the Northeast,the only area with a decreas-ing historical trend in temperature is a small portionof southern Pennsylvania. Analysis of extreme tem-peratures shows little change in the number of daysexceeding 90°F or below freezing.

Precipitation trends also show strong increases overthe last 100 years,with trends greater than 15-20%occurring in Pennsylvania, western New York,andfrom northern New York across the middle of NewEngland (Karl et al.,1996;Karl and Knight,1998).Precipitation extremes show increases of 12% inhighest annual 1-day precipitation total,and 3% inthe number of days per year exceeding 2 inches ofprecipitation over the 20th century. The region ofthe Northeast adjacent to the Great Lakes had thelargest annual increase in precipitation over theperiod 1931-1996 and the largest increases in veryextreme precipitation events (Kunkel et al.,1999a).

For the region as a whole,the period between thefirst and last dates with snow on the ground hasdecreased by 7 days over the last 50 years, resultingin a shorter snow season (Karl et al.,1993;Groismanet al.,2000). The Palmer Drought Severity Index(PDSI),which indicates trends in drought and wetperiods,confirms the precipitation patterns bydemonstrating a tendency toward more wet periodsin Pennsylvania and southern New England,and ten-dencies toward drier periods in northern NewEngland and around the Chesapeake Bay. On aver-age,the amount of land area experiencing droughtappears to have decreased over the last half century.

Projections of future climate change in theNortheast based on the Hadley climate model(Mitchell et al.,1995;Mitchell and Johns,1997; Johnset al.,1997) and Canadian climate model (Boer etal.,1984;McFarlane et al.,1992;Boer et al.,1999;Boer et al.,1999;Flato et al.,1999) suggest lowerthan average temperature increases in comparisonwith other regions of the US. The Hadley modelyields a trend for the 21st century of a 5°F (2.6°C)average temperature increase while the Canadianclimate model indicates a 9°F (5°C) increase. Winterminimum temperatures increase by 4 to 12°F (2 to7°C) in the Canadian model with the greatest

New England is a study in contrasts with a fascinat-ing land use history (Cronon,1983). Most of theregion’s almost 13 million inhabitants live in thedensely populated coastal segments in the southand east. To the north and west,New England isheavily forested and mountainous with more isolat-ed smaller urban concentrations. The New Englandeconomy is dominated by the service sector (indi-vidual state averages range from 27 to 35% of theireconomies), followed by manufacturing of durablegoods, finance,insurance and real estate,and trade.The service sector and the finance,insurance,andreal estate sector, tend to be the fastest growing seg-ments of the economy (ranging from 5 to 10%growth over the decade). New England alsoincludes the state (Connecticut) with the highestper capita personal income in the nation but,withinthe region,only Massachusetts and Connecticut cur-rently have per capita personal incomes that aregrowing at rates above the national average (NPAData Services,1998). Data from the 1979-1989Census placed Connecticut,Massachusetts,and NewHampshire with the highest percentage increases inmedian household income of US states for thedecade of comparison.

The large urban areas of the Northeast include someof the oldest metropolitan centers in the nation,andare often characterized by aging infrastructure and awide variety of stresses. The Boston to Washingtoncorridor is largely an urban landscape,with thedensest population in the nation.

CLIMATE VARIABILITY ANDCHANGEHistorically, the Northeast has been prone to naturaldisasters related to weather and climate. Floods,droughts,heat waves,and severe storms are charac-teristic. For example,seven major tropical stormscrossed the Mid-Atlantic region since 1986 and sixyears of the last 20 have been characterized by sig-nificant drought in some parts of the region. Inaddition,the major cities of the Northeast haveexperienced episodes of increased morbidity andmortality during heat waves (Kalkstein and Greene,1997). The 1990s have been characterized by anumber of significant winter precipitation eventsincluding a number of heavy snowfalls and majorice storms,while winter temperatures have tendedto be mild. Both the weather events and mild win-ters raised the regional consciousness about climateand climate change.

precipitation is projected to increase by as much as25% by 2100. In summer, the greatest increases arein western portion of the region,while in winterthe largest increases are in New England. In con-trast,the Canadian model projects little change inprecipitation or small regional decreases approach-ing 5-10%. The largest decreases are in the Mid-Atlantic region,during both winter and summer.Only small regions have a projected increase in pre-cipitation. The variability in precipitation in theNortheast is projected to increase in both modelprojections (slightly in winter and substantially insummer in the Canadian model and substantially inboth winter and summer in the Hadley model).

Severe storms are a major issue in the Northeastthroughout the year. Given the spatial resolution ofglobal climate models,neither thunderstorm activitynor hurricanes are simulated by the models. Ananalysis of sea-level pressure patterns in the Hadleyand Canadian models,which provides some indica-tion of the path of hurricanes if they form,suggestslittle reason to expect changes in the average trackof hurricanes over the 21st century. Changes in fre-quency and intensity of hurricanes under future cli -mate conditions remains a topic of considerabledebate. Global climate models are capable of resolv-ing and simulating mid-latitude cyclones responsiblefor winter storms,although the simulations areimperfect. For example,Hayden (1999) indicatesthat the Atlantic coast storm track in the Hadleymodel is displaced offshore in simulations of cur-rent conditions. In the climate projections theNortheast continues to be a location of winterstorms,as an analysis of the winter storm variabilityand locations does not shift over the 21st century.However, an analysis of storm counts and intensitiesin the Hadley and Canadian climate scenarios yieldssome differences. The Canadian model producesdecreased counts over much of the easternseaboard,with the exception of small increases inparts of the Mid-Atlantic region. In contrast,theHadley model indicates an increase over the coastalregion with slightly stronger storms. The differencesin these two scenarios reflect the position of the jetstream. The more zonal jet stream in the future sim-ulated by the Canadian model would mean fewercold-air outbreaks in the Northeast. The more north-south jet stream in the Hadley model results in anincrease in east-coast storms. Carnell et al.(1996)and Carnell and Senior (1998) describe results froma storm analysis. Storm tracks in the Atlantic areweakened,but there is a statistically significantincrease in storm counts across the Mid-Atlanticregion. Further, they find a shift toward deeper low-pressure systems,and hence stronger storms. The

increases in the western part of the region and thesmallest increases in southern New England. In theHadley model,winter minimum temperaturesincrease by approximately 7°F (4°C) in the entireregion north of Maryland,with somewhat smallerincreases immediately to the south. Summertimeincreases in minimum temperatures are projected tobe greater than 5°F (3°C) in both models.Maximum temperatures also increase,with thelargest changes in winter (from more than 12°F inWest Virginia to less than 5°F in New England in theCanadian model,and from 3 to 5°F in the Hadleymodel). Climate models differ substantially in theirprojections of the summertime increase in maxi-mum temperatures,with projections ranging from 2-3°F (1.3°C) to as much as 7-11°F (4-6°C),betweenthe Hadley and Canadian models respectively. Bothmodels used in this Assessment generally indicatesmall decreases in the variability of temperature (theexception is an increase in summer temperaturevariance in the Canadian model). The more limitedwarming in the Northeast relative to much of therest of the US may be partially attributed to thecooling effect of sulfate aerosols that are concentrat-ed in the Northeast and the maritime influences inthe coastal regions,offsetting some of the warming.

The Northeast currently has more total precipitationthan all other regions except the Southeast. Themodel simulations offer rather different scenariosfor future changes. In the Hadley model, regional

114


Figure 1. A storm track analysis from the Hadley climate model sce-nario projects a slightly strengthened wintertime storm trackthrough the Northeast, in the 2020s, because the jet stream has amore north-to-south position along the East Coast. This scenarioprojects a slightly stronger winter storm area (dark shaded region).The Canadian climate scenario has a more east-west jet, and ingeneral indicates slightly weaker storminess.

Changes in Storm Tracks

Longitude

strength of the storms is dependent on two factors.First,as the continental regions become warmer inwinter in the future,the decreased temperature con-trast between land and sea in the region tends toproduce weaker storms. However, increased heat-ing associated with higher atmospheric water vaportends to counteract this effect to produce deeperlow-pressure storms.

The Palmer Drought Severity Index derived fromthe model simulations provides two very dif ferentpictures for the Northeast. The Hadley model proj-ects less drought tendencies in the Northeast whilethe Canadian model projects tendencies for severedrought to increase over the 21 st century. Thisresult follows from the precipitation and tempera-ture projections in the two models,with theCanadian model projecting larger temperatureincreases but with smaller precipitation changes.Increased evaporation with the warmer tempera-tures yields a greater drought tendency. In contrast,the Hadley model projects a smaller warming,andwith regional precipitation increases of as much as24% by 2100.

For perspective,the specific conditions projectedby the climate models for the end of the 21st centu-ry can be matched with areas in the US that current-ly experience these specific conditions. Such acomparison provides perspective on the magnitudeand nature of the projected climate changes. Forsummer temperature and precipitation,the Hadleymodel projects that New York State will have sum-mer conditions similar to present day Maryland andsouthern Pennsylvania by the end of the 21st centu-ry. In contrast,the higher temperatures and smallerchange in precipitation found in the Canadianmodel yield a summer climate regime for NY by theend of the 21 st century that is closer to present daycentral Illinois or Missouri.

ECOLOGICAL PERSPECTIVEA mosaic of farmland and forest covers much of thelandscape in the Northeast. The southern part ofthe region (including the Appalachian Mountains ofWest Virginia, Pennsylvania,and Maryland) is charac-terized by higher percentages of corn,cotton,soy,and grasses with forests,than the rest of the region.These forests include oak,oak-hickory-pine,mixedhardwoods,mixed pines,maple-ash-beech,and limit-ed spruce-fir forests. The most northern states ofthe region exhibit northern hardwoods, red spruce-balsam fir,white pine-hemlock,oak,hickory, oak-pine,elm-red maple,and cool mountain spruce-fir

forests. Human activities converted significant frac-tions of forested land to agriculture. However, since1900 there has been extensive abandonment of agri-cultural land,and land covered by forest hasincreased significantly. One exception is in urbanareas where significant conversion of land to humanuse continues. The forests of the Northeast are stillabout 70% of the level of the 1600s if USDA (1998)data are compared with early analyses (see Forestsector foundation chapter). Forest area has been rel-atively stable over much of the region during thelast few decades,although biomass increased withthe maturity of second growth forests (Powell et al.,1996). Estimates of the percentage of forest coverin the region range from 34% in Delaware to 88% inMaine and New Hampshire (Klopatek et al.,1979).

Moderate species richness characterizes the north-ern half of the Northeast region,while the southernhalf exhibits moderately-high to high richness(Ricketts et al.,1999). For example,tree speciesrichness in the Northeast region ranges from about60 species in northern Maine to more than 140species in parts of Pennsylvania, West Virginia,andMaryland (Currie,1991). Reptile species richnessvaries from fewer than 10 species in northern Maineto about 40 in Maryland and Delaware,whileamphibian species richness ranges from fewer than20 to more than 30 species per state.

115


Figure 2. The projected trends in the Palmer Drought SeverityIndex (PDSI) are dependent on the projections of temperature andprecipitation. Large increases in drought tendencies occur in theNortheast in the Canadian model associated with substantialwarming and small changes in precipitation. In contrast, theHadley model yields larger increases in precipitation and a moremodest warming, conditions under which the drought tendencytends to decline. See Color Plate Appendix.


Palmer Drought Severity Index Change

>10

8

6

4

2

0

-2

-4

-6

-8

-10

21st Century

116


The Vegetation/Ecosystem Modeling and AnalysisProject (VEMAP) provides the basis for an assess-ment of changes in vegetation cover (VEMAP mem-bers,1995;Kittel et al.,1997) and primary produc-tivity. The specific forest character is also projectedusing the MAPSS vegetation model (Neilson andDrapek,1998). Climate projections from theCanadian model yield substantial changes in thenature of the forests in the Northeast. The coniferforest of northern New England and much of thenortheast mixed forest of New England,New Yorkand Western Pennsylvania are projected to changeto a temperate deciduous forest similar to southeast-ern Pennsylvania,Maryland,and northern Virginiatoday. The area of southeast mixed forest,todaycharacteristic of the region south of Virginia, wouldbecome compressed into a small area of WestVirginia,southern Pennsylvania,and the coastalplain of Virginia,Delaware,and New Jersey, whilemuch of Virginia would become savanna/woodland.

The changes based on the Hadley model are lessdramatic but still noteworthy. The conifer forest ofnorthern New England is replaced by northeastmixed forest. The area of temperate deciduous for -est in New England and Pennsylvania/West Virginiagrows slightly. The area of southeast mixed forestgrows in Virginia. The differences in vegetation pro-jections are a strong reflection of the differences inmoisture and temperature projections in the two cli-mate model scenarios.

KEY ISSUESRegional perspectives on the potential impacts ofclimate change are naturally influenced by personalexperience related to historical weather and climateand their associated impacts. The key issues alsoreflect perceptions of current stresses and prob-lems. In some cases,the issues are not directly relat-ed to climate change,but are likely to be exacerbat-ed by climate change.

Four key issues are identified that are of majorimportance for the Northeastern US:

• Vulnerability to changes in extreme weatherevents;

• The compounding of climate change with otherstresses for important ecosystems such as theChesapeake Bay and other bays and estuaries;

• The impact of climate change on major urbanenvironments;and

• The potential changes in recreation due to cli-mate change.

Two additional issues are also noteworthy:

• Significant change in the character of forests inthe Northeast;and

• Limited vulnerability in the agricultural sector

1. Vulnerability to Changes inExtreme Weather Events

The fact that the Northeast is prone to naturalweather disasters and weather extremes figuresstrongly in the specific examples identified by stake-holders in each area of the Northeast. The reason-ing is clear. Severe weather presents threats to bothsafety and property. During the period 1950-1989,storms caused more than $12 billion in damages inthe Northeast (Changnon and Changnon,1992).Twenty-two events each caused more than $200 mil-lion in damages,with hurricanes causing the most

Ecosystem Models

Current Ecosystems

Figure 3. The projected changes in vegetation character using out-put of the Canadian (a) and the Hadley (b) models indicates a sub-stantial northward shift in the vegetation types. These changes aresignificantly larger in the Canadian model scenario, which projectsa greater warming trend with little change or a decrease in precipi-tation. Based on the model of Neilson and Drapek, (1998).See Color Plate Appendix.



117


people were without electricity for up to threeweeks. In addition,nearly 17 million acres of ruralforests and urban trees were affected,with five mil-lion acres classified as severely damaged. Hardwoodspecies were most heavily damaged with trees bentand limbs and branches broken under the weight ofthe ice coating. The longer-term ecological impactsof severe tree damage from the storm are not yetclear, especially as the ice storm was followed bythe 1998-1999 drought.

Although the 1998 storm was extreme in terms ofpersistence and extent,other severe events haveoccurred in the 20 th century. Changes in the fre-quency, intensity, or path of ice storms are not evi-dent in the historical record. A primary concern isthe potential for such storms to become more fre-quent. Ice storms occur when warm moist air mass-es are uplifted over cold polar air masses or moveover cold surfaces. Such conditions will possiblybecome more common if the milder winters pro-jected by the climate models increase the frequencyof northern displacement of warm moist air massesas occurred in 1998. However, these effects,if theydo occur, are likely to be transitory. In the Canadianscenario,winter precipitation decreases over muchof the region and minimum winter temperatureseventually increase significantly (by 7 to more than10˚F above present values), reducing the occur-rence of subfreezing temperatures.

Severe flooding. The impacts of severe flooding(such as occurred during tropical storm Agnes in1972 and Floyd in 1999) are amply demonstrated bythe historical record from the Northeast. Records ofsevere floods reveal a diverse set of responsiblemeteorological conditions,including:

damage, followed by thunderstorms,winter storms,and wind. A 1996 blizzard in the Mid-Atlantic andNew England region, followed by flooding,causedan estimated $3 billion in damages and 187 deathsaccording to the National Oceanic and AtmosphericAdministration. Changnon and Changnon (1992)and Agee (1991) note some correlation betweenhigher historical temperatures and increasedcyclonic and anticyclonic activity when five-yearaverages from 1950 to 1989 are analyzed. Thestrongest relationships were with thunderstormactivity and winter storms. The Northeast had rela-tively few “weather disasters”during the cooler1960s, followed by increased numbers of events dur-ing the warming trend that followed. However,changes in severe weather are widely regarded asone of the most uncertain aspects of future climateprojection (USGCRP, 1995;Barron,1995). Further,the nature of these events is likely to change signifi-cantly as climate evolves over the 21st century. Inother words,some of the tendencies for changes insevere weather are likely to be short term or tran-sient features of a changing climate. Northeaststakeholders tend to focus attention on historicalevents with significant impacts,including icestorms,severe flooding,nor’easters,hurricanes andother tropical storms,and severe or persistentdrought.

Ice storms. The ice storm of January 1998 causedsubstantial environmental,economic,and societaldamage. This series of devastating ice storms hitnorthern New York and New England,along withportions of southeastern Canada,causing extensivedamage to forests and energy and transportationinfrastructure,as well as impacting human health.The magnitude of the storm in terms of measure-ments of the number of hours of persistent freezingprecipitation was unprecedented (DeGaetano,2000). While a number of significant ice stormevents have occurred over the 20th centur y, theextensive area of impact (37 counties were declaredFederal disaster areas) was also unusual.

A conservative estimate of the damage approaches 1billion dollars for the US,with insured lossesexceeding 200 million dollars (DeGaetano,2000).Many people across the region were without powerfor up to three weeks in mid-winter. Seventeendeaths occurred,primarily associated with carbonmonoxide poisoning and hypothermia associatedwith the power failure. In Maine,70% of the state’spopulation of 1.2 million people were withoutpower for at least some period of time. Over theentire region (portions of New York,Vermont,NewHampshire,and Maine) approximately 1.5 million

Figure 4. The Northeast is prone to a wide variety of natural weather disas-ters and weather extremes including the 1998 ice storm illustrated.

118


• rapid melting of snow with warming events fol-lowing a major nor’easter;

• spring snow melt following heavy winter snow-fall;

• heavy rainfall (as opposed to snow) as warm airmasses move over a frozen ground that limitspercolation and drainage;

• major summer thunderstorm systems;and • major precipitation events associated with hurri-

canes or tropical depressions.

The frequency and occurrence of future flooding inthe Northeast will depend on how this diverse setof meteorological conditions changes. Several ele-ments of the historical and model-derived future cli-mate projections raise flooding as an increased con-cern. These elements are:

• the historical trends that illustrate increases inextreme precipitation events through the latterhalf of the 20th century (Groisman et al.,2000),

• the Hadley model’s tendency to simulate wetterconditions in summer and winter, and

• the uncertainties associated with projecting howthe intensity and frequency of major hurricanesmay change in the future.

Other elements of model-derived future climate pro-jections suggest that winter and spring floodingincreases are possibly transient in nature,or are like-ly to decline in the future. These elements are:

• the Canadian model’s tendency to simulate drierconditions in summer and winter,

• the model simulations indicating milder wintersand hence the potential for northward move-ment of warm air masses in winter producingrainfall over frozen ground as the climate warms,then with decreased flooding as continuedwarming substantially reduces the length of timethe ground is frozen,and

• the potential for warming events during winterwith associated higher snowfall creating rapidmelting periods as a transient effect as the cli -mate warms, followed by decreased snowmeltevents as the climate continues to warm and pre-cipitation tends to fall increasingly as rain.

Nationally, annual flood damages increased steadilyover the period 1903 to 1997,and flood-relatedfatalities have been high since the 1970s (Kunkel etal.,1999b). Although societal growth is certainly afactor in the increase in flood damages,it is an insuf-ficient explanation. More heavy precipitation events(Karl and Knight,1998) have also been suggested asa factor in this increase. Since 1983, flood damages

in the river-rich Mid-Atlantic states total 4.7 billiondollars (US Army Corp of Engineers,1998). Floodingalso disrupts water supplies and is a significanthealth risk (Solley et al.,1998;Yarnal et al.,1997).Several water-borne diseases present risks even inwealthier countries when flood waters compromisewater systems. These include viruses (e.g., rotovirus),and bacteria-borne (e.g., Salmonella) or protozoan-borne (e.g., Giardia and Cryptosporidium) diseases.

Nor’easters. Major nor’easters produce significantprecipitation accumulations and cause significantcoastal damage,in terms of beach erosion and struc-tural damage,and thus are of major interest to theNortheast. This is particularly true because five statesof the Northeast (New York,Massachusetts,Connecticut,New Jersey, and Maryland,in respectiveorder) represent five of the top six states along theAtlantic and Gulf coasts in terms of value of insuredcoastal property (Insurance Research Council,1995).The climate model projections are divergent withregard to nor’easters. The shift to deeper wintercyclones in the western North Atlantic with strongerwinds for doubled carbon dioxide concentrationsfound by Carnell et al.(1996) indicates the potentialfor increased property damage. In contrast,Stephenson and Held (1993) found little change inthe North Atlantic using the NOAA Geophysical FluidDynamics Laboratory model,and thus future increas-es in storm damages would more reflect develop-ment of coastal property rather than climate change.

The climate models used in this Assessment also pro-vide different scenarios. The Mid-Atlantic region isthe only area of the east coast in which both theCanadian and the Hadley climate models indicateslight increases in the frequency and intensity of win-ter storms with little change in storm track. Theincreases are more significant in the Hadley modelwith the north-south shift of the jet stream underfuture carbon dioxide conditions,while the Canadianmodel suggests decreases in storm counts with theexception of the Mid-Atlantic region.

A significant hazard to coastal areas stems fromchanges in flood levels superimposed on a moregradual rise in sea level. Return periods of coastalflood events will shorten considerably, even in theabsence of any change in storm climatology. Forexample, by 2100,the 100-year flood event (theflood height that occurs on average once every 100years) in New York City is likely to occur much morefrequently (e.g., every 19 years in the Hadley modelscenario) because of the sea-level increase(Rosenzweig and Solecki et al.,2000).

Interestingly, many of the severe winter weatherconditions predicted for the Northeast may seemcounter-intuitive. For example,the Great Lakes arevery likely to experience decreased ice cover or ashorter season of ice cover with climate warming,yet a transient increase in the frequency and intensi-ty of lake effect snows is possible. Specifically, thelack of ice cover allows increased lake ef fect snowsas cold polar air masses move southward. Hence,the lake effect snows in areas such as Buffalo,NYcould break records even though the climatewarms. This result depends on the nature of coldair outbreaks that initially may not be substantiallydifferent from today. The effect is likely to be tem-porary or transient however, because as cli-mate warms substantially, the precipitation increas-ingly falls as rain according to both climate modelscenarios.

Hurricanes. Major tropical storms also remain a sig-nificant concern for several reasons:

• Hurricanes moving inland across the southernstates have produced historically high precipita-tion extremes in the Northeast,and both windand rainfall damage in New England and LongIsland. Hurricanes rank first in terms of severeweather damage in the Northeast (Changnon andChangnon,1992).

• Considerable debate is on-going as to whetherhurricane intensity may increase with warmertropical and extra-tropical temperatures,particu-larly resulting in greater precipitation,higherwinds,or both (see for example Emanuel,1988;Idso et al.,1990;Lighthill et al.,1994;Bengtssonet al.,1996;Knutsen et al.,1998).

• The debate about potential increases in the fre-quency of Atlantic hurricanes is also tied towhether warming will result in an increase ordecrease in the tendency for El Niño-like condi-tions. Historically, during El Niño events,theprobability of US land-falling hurricanes isreduced,while during La Niña events,the proba-bility of US land-falling hurricanes increases(Bove et al.,1998). One modeling study, byTimmermann et al.(1999),suggests more fre-quent El Niño-like conditions and stronger LaNiña events as climate warms.

• East Coast population growth substantiallyincreases the potential health ef fects and proper-ty damage associated with hurricane events.

In 1995,the Insurance Research Council estimatedthat a category 4 hurricane making land-fall inAsbury Park,NJ, New York City, or Long Island hadthe potential to cause insurance losses of $40 to $52

119


billion. Conservatively, total damages can easilyexceed twice the level of insured damages. Inshort,the potential for hurricane damage in theNortheast from a single storm far exceeds theregion’s total damages from hurricanes over the last40 years. Much of the vulnerability stems from aremarkable increase in coastal property values. Acomparison of storm intensity, frequency, and dam-ages during this century (Kunkel et al.,1999b) indi-cate that large increases in damages are associatedwith increasing value of property exposed to weath-er risk. For example,the value of New Yorkinsured property doubled from 1988 to 1993.Unfortunately, hurricanes’spatial scales prevent theirsimulation as features of most global climate mod-els. The Assessment climate models do not indicateany systematic change in the steering forces thatmight govern the path of future hurricanes com-pared to the present. Potential changes in intensityand frequency remain highly uncertain.

Drought. The differences in precipitation and tem-perature projected for the Northeast indicate sub-stantial difficulty in determining how drought ten-dencies may change in the future. There are signifi-cant reasons for concern,but also the potential forlittle change from present conditions. Although theNortheast is on average “water-rich”in comparisonwith precipitation levels for the rest of the nation,drought is a significant concern for three reasons:

• Six of the last 20 years were characterized bydrought in some part of the region and even asingle year drought can result in water restric-tions in many counties.

• The increased warming associated with smallerprecipitation changes in the Canadian model pro-vides a scenario for the Northeast characterizedby a strong tendency toward frequent extremedroughts.

• The lack of water storage in the Northeast is asignificant factor in creating vulnerability. Inaddition,the regions water withdrawals are high-ly dependent on surface flow. For example,inthe Mid-Atlantic region 95% of the water with-drawals are from surface flows (Neff, 2000).Drought is a frequently cited potential impact ofglobal warming because of increased evapora-tion rates associated with rising air temperatures.

In contrast,drought tendencies could change verylittle or decrease given that:

• The Hadley model,with a smaller temperatureincrease and increases in precipitation,yields atendency toward neutral changes in drought

severity or slightly decreased drought tenden-cies,

• Historical analyses indicate that the extent ofarea experiencing drought in the Northeastdeclined somewhat.

The Northeast is currently prone to weather-relatednatural disasters. However, historical analysis andclimate model projections present a range of possi-bilities,including the potential that such weather

120


disasters could increase in both summer and winter.The historical cases of large-scale damages associatedwith these events even under current climate condi-tions add a perspective of significant vulnerability.Many human structures are designed based on histor-ical climate records. If these structures are alreadyvulnerable,then this argues for adaptation strategiesthat focus on “over-designing”critical structures toadd margins of safety and more frequent design-crite-ria review based on updated climate projections.The potential for changes in frequency, path,andintensity of hurricanes,and in the nature of severewinter storms,becomes a key uncertainty in assess-ing climate impacts. Coping with substantial increas-es in severe storms,or even repeat of historicalevents coupled with higher sea level,might necessi-tate relocation of infrastructure away from high-riskzones. Historically, negative economic impacts ofsevere weather on forestry and agriculture resultedin different planting and harvesting methods. Theimpact of changes in severe weather on ecosystemsis a significant unknown.

2. Compounding Stresses on MajorEstuaries and Bays

Major estuaries and bays,such as the Long IslandSound,Delaware Bay, and the Chesapeake Bay, char-acterize the coastal regions of the Northeast. Thesecoastal embayments represent unique ecosystemsand unique resources for fisheries and recreation(Fisher et al.,2000). Importantly, these geographical-ly defined features are unable to “migrate”inresponse to climate change. In addition, growth incoastal populations has added substantial stresses tothese environments. The Chesapeake Bay, thenation’s largest estuary, is a key example.

Chesapeake Bay is characterized by multiple stresseswith significant combined impact on water charac-teristics and ecosystems (Funderbunk et al.,1991;Kearney and Stevenson,1991;Drake et al.,1996;Perry and Deller, 1995;1996; Jones et al.,1997;Ablerand Shortle,2000;Walker et al.,2000). Human land-use practices and upstream industry have a markedimpact on water quality and pollutant levels. Highnutrient and particulate loading from agriculturaland urban runoff and air-borne pollutants reducesoxygen levels,which reduces productivity and organ-ism habitat area. High nutrient loading is very likelyto increase algal blooms,shading deeper water andlimiting submerged aquatic vegetation. Increasedland use and growth in the area covered by impervi-ous surfaces has resulted in “flashier”streams,mean-ing that any rainfall causes a rapid peak flowresponse. In naturally vegetated areas such as

Figure 5. Schematic of the potential changes in severe weather forthe Northeast based on historical data (H), the Hadley model sce-nario (HS), the Canadian model scenario (CS) or an assessment ofpossible transient effects (T). See Color Plate Appendix.

Potential Changes In Severe Weather

Figure 6. The coastal regions of the Northeast are dominated byextensive estuaries and bays.

forests,a significant fraction of the rain percolatesinto the ground. As the soil becomes saturated,morewater flows into the streams. It thus takes sometime for rainfall to cause stream levels to rise. Whenland uses cover much of the ground with streets,buildings,and parking lots,less rainfall can percolateinto the ground;thus it runs off quickly, causing thestreams to “flash.” Increased population pressure hasalso altered the boundary between the land environ-ment and the Bay, providing fixed or “hardened”mar-gins and decreasing the area of wetlands.

Climate change adds an additional stress. The mostimportant influences reflect the potential to changewater temperatures and freshwater inputs (hencesalinity). Warming of the atmosphere is very likely tohave a direct impact on Bay temperatures. Changesin the frequency and intensity of precipitationevents, changes in frequency and strength of hurri-canes,and any change in the strength and frequencyof droughts would substantially influence freshwaterinputs. With “flashier”streams,any increase in highprecipitation events is very likely to have a markedimpact. These changes would then influence salinityin the Bay and stratification of the water mass. Sea-level rise is likely to contribute to these changes.Local rates of sea-level rise in the Chesapeake (closeto 4mm/year, or 0.16 inches per year yielding 16inches in 100 years) are anomalously high for theAtlantic Coast,due to regional subsidence and otherfactors (Gornitz,1999). The global average increasein sea level is closer to 1.2mm/year (0.048 inchesper year) according to Gornitz and Lebedoff (1987).Both temperature and salinity are significant environ-mental controls on organism character, influencingfish populations,productivity, and human and ecosys -tem health. For example, cholera bacteria are pres-ent in the Chesapeake Bay. Increased cholera risk isassociated with rising water temperatures,howeverwater and waste treatment practices should preventUS epidemics (Colwell et al.,1998).

Gibson (1999) and Najjar et al.(2000) used a waterbalance model to project a 24% increase in runoff inthe Susquehanna River Basin under the Hadleymodel scenario and a 4% decrease for the Canadianmodel scenario. Gibson and Najjar (2000) then ana-lyzed changes in Chesapeake salinity as a function ofthese changes in runoff. The increased runoff in theHadley model resulted in a 20% decrease in the sur-face salinity within the northern segment of theChesapeake,with as much as a 4% change penetrat-ing to deeper waters within the southern segment ofthe Bay. The Canadian model scenario results inchanges of 3% or less.The potential for significant changes in temperature

121


and salinity raises several important concerns aboutspecies composition in the Chesapeake and otherbays. Any significant changes in salinity and tempera-ture are very likely to result in the migration or loss ofkey species. The introduction of opportunistic inva-sive species during changing conditions is also verylikely to change predator-prey relationships influenc-ing the character of ecosystems,or result in elimina-tion of key species that may not be vulnerable to thedirect effects of climate change. This type of indirectchange is already evident in Chesapeake waterfowlpopulations in which declines of submerged aquaticvegetation,attributed to excessive nutrients and sedi-mentation,are associated with dramatic declines insome species of waterfowl unable to adapt to chang-ing food sources (Perry and Uhler, 1988). TheChesapeake and Delaware Bays are stopover pointsfor millions of bird species. Numerous waterfowl (theeight most dominant species have been numbered at700,000) winter in the Chesapeake. If affectedspecies have economic significance or if they are con-nected to the uniqueness of a region,the impact isvery likely to be significant.

Sea-level change substantially influences wetlandsthrough inundation,saltwater intrusion into fresh andbrackish marshes,and erosion. Marshes are alreadyestimated to have lost one-third of their original area.

Percent Salinity Change in the Chesapeake Bay

Figure 7. Calculated salinity within the Chesapeake given the run-off calculated from the Hadley (top) and Canadian (bottom) climatescenarios by Gibson and Najjar (2000). The distribution of salinitiesranges from the upper reaches of the Bay (39.66N) to the LowerChesapeake near its Atlantic opening (36.95N). See Color PlateAppendix.

39.66N Latitude 36.95N

39.66N Latitude 36.95N

Urban Areas

The major urban areas of the Northeast are stressedeven without climate change (Rosenzweig andSolecki et al.,2000). In the Northeast,there havebeen major investments in the infrastructures of sys-tem elements such as roads, water supply, communi-cation,energy delivery, and waste disposal. Still, formany large urban areas this infrastructure is charac-terized by aging,problems with under-capacity, over-use,and deferred maintenance. Cities are associatedwith a host of continuing problems,some related toclimate like air quality, and others that are non-cli-mate related such as crime and poverty. The com-plex web of institutional relationships amongcommunities, municipalities, regions,states,thefederal government,and the public,private,andnon-profit sectors often make consideration ofoverarching issues,such as the environment,problematic. Scenarios from climate models suggestthat climate change intersects with a significantnumber of other stresses,all with implications foroverall quality of life.

Sea Level and Storm Surge. One of the more signifi-cant potential climate change impacts in urbancoastal communities is rising sea level (Bloomfield,1999) and elevated storm surge levels. For example,historical events suggest that the metropolitan areasare particularly vulnerable. By using daily tide gaugedata and statistics on extreme events (Ebersole,

Sea-level rise can also submerge protective barrierislands or cause them to retreat landward ontomarsh and lagoonal areas.The hardening of the bayboundaries (through construction of impound-ments, retaining walls,dockage,etc.) adds an addi-tional limitation to the landward movement of wet-lands as sea level rises. Marshes and lagoons are sig-nificant stopover habitats for migratory birds.

Adaptation strategies are governed by three factors.First,estuaries and bays are characterized by a num-ber of compounding stresses. Second,there are lim-ited avenues for protecting critical ecosystems thatare geographically fixed and therefore cannot“migrate.” Third,the uncertainties in projecting thewater balance for the Northeast region under futureclimate conditions results in substantial uncertaintyin determining the future water properties of theBay and other estuaries in the region. One of thefew adaptation strategies available may be to limitthe non-climate stresses on the region in order tominimize any climatic impacts. This may argue forgreater control of land-use at the boundaries of estu-aries and increased concern over nutrient and pollu -tant fluxes into estuaries and bays. Humans are alsogaining some experience in constructing artificialwetlands. However, we know little about the long-term viability of these constructed wetlands.

3. Multiple Stresses on Major

122


Nor’easter of December 1992

The December 11-12,1992 nor’easter produced some of the worst flooding and strongest winds on recordfor the area. It resulted in a near shutdown of the New York metropolitan transportation system and evacua-

tion of many seaside communities in New Jerseyand Long Island. This storm should have provided a“wake-up”call,heralding the vulnerability of thetransportation system to major nor’easters and hur-ricanes. Had flood levels been only 1 to 2 feetabove the actual high water level of 8.5-foot abovemean sea level,massive inundation of rail and sub-way tunnels could have resulted in loss of life. Withrising sea levels, even a weaker storm would pro-duce comparable damage. While hurricanes aremuch less frequent than nor’easters in this area,they can be even more destructive because thegeometry of the New Jersey and Long Island coastsamplifies surge levels toward the New York City har-bor. For a worst-case scenario category 3 hurricane,surge levels could rise 25 feet above mean sea levelat JFK airport and 21 feet at the Lincoln tunnelentrance.

Figure 8. Vulnerable coastal areas for Manhattan, based on a 20foot high flooding zone for the year 2100, derived by Klaus Jacob ofLamont-Doherty Earth Observatory for the Metroeast workshop.See Color Plate Appendix.

1234

1982),the Metro-East Assessment team (Rosenzweigand Solecki et al.,2000) calculated a range of scenar-ios delineating significant coastal vulnerability, par-ticularly in urban areas. Critical points for floodingfor many of the region’s vital transportation systems(including airports,subways,highways,and majorroad and railroad tunnels) are located at elevationsbetween 7 and 20 feet above current sea level andare very likely to be inundated by coastal stormsurges with estimated recurrence periods of about100 years. Taking into account models of sea-levelrise (IPCC,1996) of 9 inches to 3 feet (23 to 96 cm)in the next 100 years,these current recurrence peri-ods are likely to be shortened by factors of 3 to 10before the year 2100.

Water Supply. Water supply systems in the majornortheastern cities also exhibit substantial vulnera-bility to climate change. For example,the New YorkCity water system is large and relatively inflexible interms of demands on the system,and therefore sus-ceptible to large changes in the water balance. Afew key examples illustrate the nature of the prob-lem. New York City’s water supply is derived fromwater collected from a 2,000 square mile area,stored in three upland reservoir systems. This is anecosystem service that the City has secured by mak-ing a capital investment of about $1 billion in thesurrounding communities. The City’s water supplyis sensitive to climate variability and change giventhe demands on the current system,and its depend-ence on annual precipitation levels. Even withincreased winter precipitation, rapid run-off (ratherthan accumulation as snow) has resulted in signifi-cant regional water supply problems in recentdecades (McCabe and Ayers,1989). The climatestresses from current conditions are already evident.The climate vulnerability is compounded by twofactors. First,the upstate communities have experi-enced substantial growth. Second,these upstatecommunities have a legal right to water in times oflow supply. In addition,New York City has a legalobligation to provide water to the Delaware Basinbecause it has access to water in the river head-lands. In the Delaware Basin,additional waterreleases from the reservoir systems might berequired if sea-level rise advances the salt waterfront up the Delaware and Hudson rivers (Alpern,1996). A prolonged drought is likely to force thecity to seek alternative water sources.

Soil Moisture. The combination of future precipita-tion changes and an increase in evaporation associ-ated with higher future temperatures is expected toproduce a decrease in summer soil moisture

(Broccoli,1996). The climate model scenarios usedin this Assessment both project little trend in soilmoisture in the region immediately adjacent to thecity. However, significant climate change is likely toresult in the need for new large-scale investments inorder to replace the current ecosystem service fromthe growing surrounding communities.

Heat-related Illness and Death. Warmer summersare likely to be associated with higher maximumtemperatures and then with increased heat-relatedmorbidity and mortality in major cities of theNortheast (Kalkstein and Greene,1997;Chestnut etal.,1998;Kilbourne et al.,1982). By 2090,increasesin maximum temperatures from 1-2˚F to 5˚F areprojected for the coastal Northeast. An associatedincrease from the current 13 days to a projected 16-32 days above 90˚F might result in a five-foldincrease in heat-related mortality in New York Cityaccording to the Metroeast regional assessment,ifno adaptation occurs. Kalkstein and Greene (1997)and Kalkstein and Swift (1998) utilized three differ-ent climate models to examine winter and summermortality for 2020 and 2050. In all three models,increases in summer mortality exceeded decreasesin winter mortality for Baltimore,Philadelphia,Pittsburgh,and Washington D.C.

Air Pollution. Higher summer temperatures alsoincrease photochemical reaction rates leading to anincrease in ground-level ozone (smog) and otherpollutants. The New York City area already has oneof the nation’s highest rates of respiratory diseaseassociated with airborne pollutants. This is likely tobe exacerbated by increases in the urban heat islandeffect and conditions of persistent elevated summer

123


Figure 9. Water is one of the major ecosystem services provided toNew York City by the surrounding land areas (MetroeastWorkshop). Population growth in the region, and legal rights to thewater by upstate communities, place the adequacy of the watersupply at greater risk under some scenarios of future climatechange.

through stricter controls on pollutants and ozoneprecursors. Construction of dike systems or reloca-tion of critical infrastructure can reduce some of thevulnerability to sea-level rise.

Significant financial investment would be requiredto produce an urban system that is more robustunder conditions of higher storm surges orincreased tendency toward drought. Increasedflooding and a significant increase in severe stormswould argue for establishment of set back zones, re-zoning,buyouts of high risk areas, relocation ofstructures at risk,or altered management. Investingin new reservoirs,limiting growth in water sourceregions,and increasing water costs to promote con-servation are the primary coping strategies forwater-related problems,but major reservoir con-struction projects have associated environmentalconsequences. A more comprehensive regionalwater management strategy with increased emphasison safe water supplies could substantially mitigatemajor water supply and quality problems.

4. Recreation Shifts

Changes in warmth and the seasonal characteristicsof precipitation are likely to have substantial impactson recreation in the Northeast. In particular, this hasbeen identified as a key issue for New England(Rock and Moore,2000). These impacts differ wide-ly with the type of recreation and season. The win-ter ski industry is particularly vulnerable. Increasesin minimum nighttime temperatures,periodic warmspells,and increased occurrence of winter precipita-tion as rain are likely to limit the ability of ski areasto maintain adequate snow pack. Current skiinglocations with marginal climate characteristics arelikely to become untenable.

The recent tendency for mild winters in theNortheast,coupled with climate model projectionsindicating significant increases in winter minimumtemperatures,points to possible benefits. Mild win-ters and extended periods of warmth in fall andspring may encourage new recreational activities inthe forested mountains of the region.

Higher sea level coupled with increased winterstorms is likely to result in loss of beachfront prop-erty and destruction of barrier islands,decreasingthe opportunities for beach recreation during warmmonths. In highly populated areas,prime recreation-al beaches can probably be maintained by more fre-quent episodes of beach nourishment,but costscould increase substantially (Valverde et al.,1999).Wet springs and mild winters are likely to lead to

temperatures.

In summary, the major urban areas of the Northeastare characterized by multiple stresses. Climatechange has the potential to exacerbate many ofthese problems. The most significant climate-relatedstresses in the urban regions of the Northeastare increased heat mortality, greater ozonepollution associated with higher temperatures,infra-structure risk due to higher sea level, water availabil-ity problems associated with significant changes inregional water balance,and increased vulnerabilityto severe storms.

The addition of significant multiple stresses to theurban environment argues for a variety of adapta-tion strategies. Some of the direct health effects ofheat can be mitigated with active warning systems(heat alerts,opening of shelters,spraying water ondark building tops) such as those currently in placein Philadelphia,structural adaptations (constructionthat promotes air-flow, reduction in area of blackroofing,etc.) and “cool community”measures suchas increased planting of trees. Some of the project-ed mortality studies do not reflect potential changesin air conditioning use or other adaptations,whichoffer substantial protection against heat waves.Indirect air pollution problems can be addressed

124


Figure 10. Outdoor recreation is of major economic importance inthe Northeast, and it is tightly coupled to climatic conditions.

increased populations of insects and high pollencounts. Increases in disease-bearing vectors,such asmosquitoes and ticks,add increased health risks.For example,mosquito populations are tightly con-nected to minimum temperature characteristics andwater availability. Increases in pest populations arelikely to adversely impact recreation. The climatemodel projections offer different scenarios;theincreased drought frequency in the Canadian modelyields a different pest population than the warmerand wetter projection from the Hadley model.

The New England region currently experiencessummer air quality and ozone pollution problems,both of which are exacerbated by warming.Ground-level ozone across New England may reachunhealthful levels during summer months,especiallyduring humid periods when maximum tempera-tures exceed 90˚F. The combination of high temper-atures and full sunlight,coupled with nitrogenoxides generated primarily by automobile traffic,and volatile organic compounds from both naturaland human sources, result in elevated levels ofground-level ozone,a form of air pollution oftencalled smog. Exposure to elevated levels of ozonehas a negative impact on both forest health andhuman health. Due to the topographic variabilitytypical of New England and upstate New York,andthe fact that the region is typically downwind frommajor urban centers,high-elevation areas (above3,000 feet) are likely to have unhealthful levels ofozone. The prospect of air quality alerts for hikersin the mountains of New England is likely to have anegative effect on tourism.

In contrast,typical summer recreational activitiesinvolving beaches or freshwater reservoirs are verylikely to experience extended seasons. However,sea-level rise is projected to lead to increased beacherosion and loss due to the impact of both stormsurges and permanent inundation. Due to the largeamount of human development abutting beachareas,establishment of new beaches further inlandis difficult in many cases or could result in signifi-cant costs to land owners or both. Adaptation meas-ures including beach replenishment or hard struc-tures such as sea walls and groins are costly andoften ineffective except in the short-term. Thediverse waterways of the Northeast,including lakes,rivers,beaches,and estuaries,are likely to continueto be havens for escape from the summer heat.

In autumn,a major recreational draw for theNortheast is the display of fall foliage. However,increased autumn warmth is associated with mutingof fall foliage colors. Drought decreases leaf color

and changes the timing of leaf drop. Both factorsdetract from a major tourist attraction in theNortheast. The two climate models used in thisAssessment offer different recreational scenarios asthe Canadian model projects increased droughtwhile the Hadley model projects little change indrought risk.

The key outcome of an analysis of recreational activ-ities is that the extent of potential impacts is highlydependent on the type of activity and on the differ-ences between the model results. Many of the activ-ities are likely to “migrate”out of portions of theNortheast or will move northward (such as the skiindustry). Ski resorts are likely to be required tocontinue current trends toward development ofyear-round attractions. An extended warm weatherseason is very likely to make waterways,mountains,and forests greater attractions. Undoubtedly, humanswill make trade-offs in terms of type and location ofrecreational activities. The uncertainties associatedwith the water balance projections for the Northeastcontribute to uncertainties about the impact of cli-mate change on recreation in the region.

A key uncertainty is whether climate change will125


Figure 11. On warm humid days when temperatures exceed 90ºF,ozone problems are exacerbated across the region. The top figureshows the view on a clear day at the Great Gulf of MountWashington, New Hampshire. The bottom figure shows the sameview when temperatures exceed 90ºF and air quality problems occur.

126


New England Maple Syrup

A successful maple syrup season in New England depends on the proper combination of freezing nights andwarm daytime temperatures,along with prolonged cold temperatures (resulting in a recharge of sugar to thesap) during the months of February and March. When the right combination of these climatic conditionsoccurs,the sugar maple tree produces a sap containing 2-5% sugar. In addition,the first flow of sap in a givenseason generally produces the highest quality maple syrup. A sustained,early flow heralds a good year for themaple industry in an area. If the initial flow occurs too early, before many of the producers have tapped theirtrees,they will miss this profitable opportunity. The maple industry in New England depends to a largeextent on the timing of these critical climate events. Due to changes in both technology (the advent of tub-ing) and climate (very early initial flows and a reduction in freeze/thaw cycles and cold recharge periods),themaple syrup industry is moving from New England into Canada.

In the past,the success of the maple syrup industry in Canada was limited by deep snow cover (limitingaccess to individual trees) and fewer freeze/thaw cycles due to prolonged periods of low nighttime and day-time temperatures. The development of tubing-based sap collection methods,which provide easier access totrees and eliminate the need to make frequent collections,has allowed the Canadians to become more com-petitive in the past several decades. Changes in climate over the past several decades also allowed theCanadians to collect more sap over a longer “sugar season”than in the past. Conditions for sustained sap flownow mark the Canadian season while higher temperatures have led to fewer freeze/thaw cycles and reducedcold recharge periods in New England. This,coupled with earlier and earlier initial flows over the past twodecades,has resulted in a shift in the volume of syrup production from the US to the Gaspe Peninsula ofQuebec. It is interesting to note that in 1928,the major syrup production center in the US was located inGarrett County, Maryland.

If wintertime minimum temperatures continue to increase more rapidly than the maximum temperatures (asindicated in the climate models used in this Assessment) then the current northward shift in maple syrup pro-duction is very likely to continue. In the long term, change in the range of the maple tree is very likely tocompletely dominate the ability to produce maple syrup. The climate model scenarios,when coupled withassessments of the species composition of forests under the Canadian and Hadley model climate projections,project a substantial northward displacement in the distribution of maple trees (see Chapter 17 on Forests).It is likely under these scenarios that maple syrup production will not be possible in many regions of theNortheast because conditions for tree growth are unsuitable.

yield an increased tendency toward drought,whichwill impact fall colors, forest health,and the natureof water-related recreational activities.

ADDITIONAL ISSUES

Two additional issues are likely to be of considerablesignificance for the Northeast:

Forests

The species composition of the forests of theNortheast is very likely to change dramatically underboth the Hadley and Canadian climate scenarios.Two approaches are available for assessing the forestchanges in the Northeast. The MAPSS vegetationanalysis (Neilson and Drapek,1998) provides ananalysis of the distribution changes in large-scale for-

est types (e.g.,the Northeast mixed forest or theconifer forests of New England). The model ofIverson and Prasad (1998) examines changes indominant forest species (e.g.,maple-beech-birch,oak-hickory, elm-ash-cottonwood,and oak-gum-cypress). Although both vegetation models indicatesubstantial changes in forest character in responseto the climate model scenarios,there are also somedifferences between the models as might be expect-ed from two different approaches.

Based on the Canadian model scenario and theMAPSS vegetation analysis,the conifer forest ofnorthern New England and much of the northeastmixed forest of New England,New York,and west-ern Pennsylvania changes to a temperate deciduousforest similar to southeastern Pennsylvania andnorthern Virginia today. The area of southeast mixedforest,today characteristic of the region south ofVirginia,becomes compressed into a small area of

127


Agriculture

Although crop production is tied to climate, agricul-ture is a relatively small and declining fraction ofeconomic output of the Northeast. Further, moststudies indicate that the agriculture in the Northeastis relatively robust to climate change even thoughthe crop mix may change (Abler et al.,1999). Theability to change crop types and to take advantageof hybrids limits vulnerability. The most frequentlycited concern related to agriculture is associatedwith market forces,some of which are likely to beclimate-related, generated by agricultural changesoutside of the Northeast region. Some cool weathercrops and many small family farms are exceptions tothese conclusions. Farmers in the Northeast regionmake a significant contribution to the national sup-ply of dairy products and food crops such as apples,grapes,potatoes, sweet corn,onions,cabbage,andmaple syrup. In addition,small family farmsthroughout the Northeast are vital to the economyof rural areas,and they fill an important marketniche for fresh,high quality, affordable local pro-duce. These farms will be particularly sensitive toclimate change due to the cost of adaptation,espe-cially if compounded by other market forces. Forexample,the Northeast dairy industry is alreadyquite fragile. Milk production by dairy cows is opti-mal at cool temperatures,so an increase in tempera-tures will require substantial increases in air condi-tioning costs. Most research on climate changeimpacts on agriculture has focused on major worldtrade crops such as wheat,soybeans,and corn.

West Virginia,southern Pennsylvania,and the coastalplain of Virginia,Delaware,and New Jersey, whilemuch of Virginia becomes savanna/woodland. TheHadley model predicts less dramatic changes but stillthe conifer forest of northern New England isreplaced by northeast mixed forest. The area of tem-perate deciduous forest in New England andPennsylvania/West Virginia grows slightly. The area ofsoutheast mixed forest grows in Virginia. The north-east mixed forest declines dramatically in total area(72% loss of area in the US according to the ForestSector Foundation report) and specific species, forexample the sugar maple (Acer saccarum),are lostentirely from the US (Watson et al.,1998).

An additional analysis by the Mid-AtlanticAssessment (Fisher et al.,2000) examined the distri-bution of dominant forest types based on the modelof Iverson and Prasad (1998). Oak-hickory forests(46%) and maple-beech-birch (37%) dominate theregion today. In both climate model scenarios,oakand hickory forests replace the maple-beech-birchforests of western and northern Pennsylvania,WestVirginia,and southern New York. These largechanges have significant potential to affect aesthet-ics,tourism, fall colors,and to cause people con-cern. Although in the short-term, forests are regard-ed as only moderately vulnerable to the specific cli-mate changes projected, changes in timber, and non-timber values such as recreation,scenic views,andwildlife habitat may be long-term issues.

Forest issues extend well beyond the species com-position of the forests. These issues include (a) thepotential for alteration of tree resistance to insectsassociated with changes in temperature and wateravailability (Roth et al.,1997),(b) increased fireoccurrence associated with increased tendencytoward drought,(c) reduction in primary productivi-ty and carbon storage despite carbon dioxide fertil-ization if drought becomes as significant as project-ed by the Canadian model,(d) changes in forest dis-turbance as a function of changes in hurricane fre-quency and intensity, or changes in wind,ice,orheavy precipitation events. Climate change can alsocompound the impacts of invasive species,as inva-sive species tend to have high reproductive rates,good dispersal ability, and rapid growth rates(Williamson,1999). Many of these impacts are asso-ciated with aspects of climate model projectionsthat are associated with uncertainty – e.g., changesin the character of extreme weather and the natureof the future water balance in the Northeast underclimate change conditions.

Figure 12. Dominant forest types for the mid-Atlantic region for cur-rent climate, and the potential distribution of these forest types forthe Canadian and Hadley climate scenarios based on the Mid-AtlanticAssessment. Based on the model of Iverson and Prasad (1998).See Color Plate Appendix.

impact on Lyme disease vectors,the complexity ofthe relationships makes changes in the distributionand frequency of the disease under altered climatedifficult to predict (Martens,1999). Changes inmosquito populations and survival are also possiblewith warmer and wetter conditions. Recent exam-ples of outbreaks of West Nile Virus and equineencephalitis in Northeast urban areas have substan-tially raised concerns about vector-borne diseasesand illustrate that improved monitoring and betterunderstanding of these diseases are important tothe region. Increased temperatures in coastal baysare likely to increase algal blooms,which have achance of harboring cholera. Increased rainfall andflooding,if severe,have historically caused contami-nation of public and private water supplies (e.g.,with cryptosporidium) and models project anincrease in very heavy rainfall events and the poten-tial for greater flood risks. However, in large meas-ure,US public health infrastructure and responsecapabilities,if vigorously sustained,are likely tolimit the potential impacts.

Species Changes. Although ecosystem characterwas not selected as one of the major topics for thefirst US National Assessment, regional assessmentteams raised a significant number of issues concern-ing changes in species composition beyond forestcharacter. Changes in temperature and precipita-tion will have direct impacts on species distributionas well as indirect impacts associated with chang-ing predator-prey relationships and/or changes inpests and disease. In many cases,these species maybe truly characteristic of a region or may be of eco-nomic significance. For example,lobster popula-tions are associated with cooler waters and warm-ing may promote northward migration of the popu-lation — a key issue for New England. Coastal pop-ulation pressures combined with sea-level rise arevery likely to limit habitat regions along the AtlanticFlyway for migratory birds. Warming is likely tosubstantially limit trout populations — a key issuefor Pennsylvania. Changes in species mix and intro-duction of climate-driven invasive species are likelyto also have unanticipated feedbacks on ecosys-tems.

Differential Human Impacts. The large differencesin economic status and the aging of the populationin the Northeast is likely to be associated with dif-ferential impacts based on the ability to respond toclimate change. Where impacts are significant, cli-mate change is likely to have greater impact onlower-income residents and the elderly, as well aschildren and the ill (e.g.,those with chronic respira-tory ailments). The key concerns are heat mortality,

Information from these studies has only very limitedapplication to the Northeast where dairy and highvalue horticultural crops dominate the economy.Overall, agriculture in the Northeast is likely to sur-vive a climate change,and may even benefit relativeto some other regions of the US. However, the costsof adaptation could be high,and the vulnerability ofsmall farms is likely to increase.

ADDED INSIGHTS

The results of assessment research in the Northeastyield several additional insights into the impacts ofclimate change on the region. Substantive issuesinclude,but are not limited to the following:

Institutional Complexity. The complex institutionalframework of community, municipal,county, region-al,and statewide formal and informal governing bod-ies that characterize the Northeast have the poten-tial to limit the region’s ability to deal withextremes. For example,the complex array of water-sheds,small management units,and urban depend-encies on broad surrounding regions with dif ferentinstitutional characteristics all have the potential tolimit drought planning and response management.Climate change is likely to cause a change in politi-cal focus and management of critical ecosystemservices like water. Similar issues arise for weatherdisaster social services and energy distribution.There are signs of innovative management evenwith complex institutional structures in theNortheast transportation systems (e.g.,introductionof electronic fare collection systems and EZ Pass).The ability of the Northeast to adapt to extreme sit-uations will depend upon the ability of institutionsto identify and prioritize vulnerable facilities andpopulations. Targeting and flexibility in the use ofresources among the many institutions is needed toadapt more effectively to climate change.

Infectious Diseases. Infectious disease vectors areoften strongly influenced by climate. For example,the primary Lyme disease vector is the deer tick.The tick population is governed by the size of bothmouse and deer populations. Increased acorn mastproduction directly influences rodent populations.Milder winters contribute to a larger survival ratefor deer. Larger deer populations increase thehuman contact with deer and deer ticks,increasingthe possibility of Lyme disease infections. Milderwinters are projected by virtually all climate models(the two primary models used in this Assessmentboth indicate large increases in winter minimumtemperatures). Although climate has a strong

128


susceptibility to disease,and changes in air quali-ty factors such as ozone. For example,lowerincome residents,often including the elderly, mayhave additional burdens because of lack of airconditioning,poor housing conditions,and unsafeneighborhoods leading to unwillingness to openwindows or seek relief out of doors or away fromthe city.

ADAPTATION STRATEGIES The most important elements of adaptationstrategies proposed for the Northeast include:

• relocating structures at risk from severeweather (e.g.,hurricanes) and flooding (bothin coastal regions and in river systems proneto high water levels);

• strengthening design criteria for critical infra-structure (e.g.,power supply) to ensurerobust operation under possible changes inweather and climate extremes;

• increasing reservoir construction and improv-ing management of water supplies to increaserobustness of the water systems under condi-tions of flooding or drought, recognizing thepotential for other negative consequences;

• greater emphasis on water quality and airquality controls to minimize the compoundingof climate impacts,including stricter adher-ence to existing regulations and potentiallystricter controls on pollutants and their pre-cursors;

• incorporating active warning systems (e.g.,Philadelphia’s heat wave warning system) andstructural adaptations (e.g.,construction thatpromotes air-flow, and reduction in area ofblack roofing) to limit the potential forincreased heat mortality and morbidity;

• limiting the non-climate stresses in order tominimize impacts on critical regions that aregeographically fixed and therefore can’t“migrate”such as the Chesapeake Bay;

• limiting agricultural and forestry economicimpacts by introducing different planting andharvesting methods based on historicalresponses to weather and climate;and

• limiting coastal development through existingregulatory frameworks to protect coastalregions,increasing the focus on “smart”landuse to reduce vulnerability to floods andstorm damage,and limiting pollution deliveredto coastal regions by water run-off from theadjacent land area.

The social and economic framework of the future islikely to be substantially more advanced than at pres-ent. As a result,the adaptation strategies maybecome more numerous,more effective,and moreeasily implemented.

CRUCIAL UNKNOWNS ANDRESEARCH NEEDSSeven critical factors limit our ability to assess thepotential importance of climate change on theNortheast:

• For the Northeast, changes in extreme weatherare viewed as a critical issue in assessing potentialimpacts. A key issue is an ability to assess thepotential changes in severe storms,including hur-ricane intensity, path,and frequency and the char-acter and frequency of nor’easters. Changes inthe frequency and intensity of extreme weatherevents with climate change remain one of themost uncertain aspects of future climate.

• Improved projections of the spatial and temporaldistribution of future temperature and precipita -tion would enable a more robust assessment. Thedifference in the projections of changes in thePalmer Drought Severity Index between theCanadian and Hadley models presents significant-ly different scenarios,with dramatically differentimplications for ecosystems, water, recreation,andagriculture.

• The ability to combine multiple stresses and tosimulate the resultant effects on the environmentis severely limited in many specific environments(e.g.,the Chesapeake Bay). Fully integrated obser-vational networks for multiple variables and com-prehensive models will be required to addressmultiple stresses.

• Changes in population and land use patterns willhave dramatic effects on ecosystems and on thenature and magnitude of climate impacts, yet theability to project these changes is limited.Continued growth in coastal populations in con-junction with sea-level rise introduces substantialadditional risk due to severe weather and greaterconcerns about habitat loss.

• Understanding of the potential change in speciescomposition and character in response to climatechange is in its infancy. This includes changes ineconomically important species,pests,invasivespecies,and predator-prey relations. Validation ofbiological response models is a major problem.

• Understanding of how people and societies willadapt,due to uncertainties in overall changes in

129


Bove,M.C., J. B.Elsner, C.W. Landsea,X.Niu,and J. J.O’Brien, Effect of El Ni ñ o on U.S.land falling hurri-canes, revisited, Bulletin of the AmericanMeteorological Society, 79, 2477-2482,1998.

Broccoli,A. J.,The greenhouse effect: The sciencebase,in The Baked Apple? Metropolitan New York inthe Greenhouse, edited by D. Hill,Annals of the NewYork Academy of Sciences, vol.790,23,1996.

Bureau of Economic Analysis,Survey of current busi-ness, June 2000, U.S.Government Printing Office,2000. Available online athttp://www.bea.doc.gov/bea/regional/data.htm.

Carnell,R.E.,and C.A.Senior, Changes in mid-latitudevariability due to increasing greenhouse gases andsulphate aerosols, Climate Dynamics, 14,368-383,1998.

Carnell,R.E.,C.A.Senior, and J. F. B.Mitchell,Anassessment of measures of storminess:simulatedchanges in northern hemisphere winter due toincreasing CO2, Climate Dynamics, 12,467-476,1996.

Changnon,S.A.,and J. M.Changnon,Temporal fluctua-tions in weather disasters:1950-1989, ClimaticChange, 22,191-208,1992.

Chestnut,L. G.,W. S. Freffle, J. B.Smith,and L.S.Kalkstein,Analysis of dif ferences in hot-weather-relat-ed mortality across 44 US metropolitan areas,Environmental Sciences Policy, 1,59-70,1998.

Colwell,R. P. Epstein, D. Gubler, M.Hall, P. Reiter, J.Shukla,W. Sprigg,E.Takafuji,and J.Trtanj,Global cli-mate change and infectious diseases, EmergingInfectious Diseases, 4,451-452,1998.

Cronon,W., Changes in the Land, Hill and Wang,NewYork,241 pp.,1983.

Currie, D. J.,Energy and large-scale patterns of animal-and plant-species richness, The American Naturalist,137, 27-49,1991.

DeGaetano,A. T.,Climate perspective and impacts ofthe 1988 northern New York and New England icestorm, Bulletin of the American MeteorologicalSociety, 81,237-254,2000.

Drake,B. G., G. Peresta,E.Beugeling,and R.Matamala,Long-term elevated CO2 exposure in a ChesapeakeBay wetland: Ecosystem gas exchange,primary pro-duction,and tissue nitrogen,in Carbon Dioxide and

the economy, technology, and societal values,is amajor area of uncertainty.

• Uncertainties in estimating sea-level rise substan-tially limit the ability to assess the height ofstorm surges and the recurrence interval offlooding events with substantial economicimpacts.

LITERATURE CITED

Agee,E.M.,Trends in cyclone and anticyclone frequen-cy and comparison with periods of warming and cool-ing over North America, Journal of Climate, 4,263-267,1991.

Abler, D. G. and J. S.Shortle,Climate change and agricul-ture in the Mid-Atlantic region, Climate Research, 14,185-194,2000.

Alpern,R.,Impact of global warming on waterresources: Implications for New York City and theNew York metropolitan region,in The Baked Apple?Metropolitan New York in the Greenhouse,Annals ofthe New York Academy of Sciences, vol.790,edited byD. Hill,86,1996.

Barron,E. J.,Climate models: How reliable are theirpredictions?, Consequences, 1,16-27,1995.

Bengtsson,L.,M.Botzet,and M.Esch,Will greenhouse-gas-induced warming over the next 50 years lead tohigher frequency and greater intensity of hurricanes?,Tellus, 48, 57-73,1996.

Bloomfield, J.,Hot nights in the city: global warming,sea-level rise and the New York metropolitan region,Environmental Defense Fund,New York,1999.

Boer, G. J., N.A.McFarlane,R.Laprise, J. D. Henderson,and J. P. Blanchet, The Canadian Climate Centre spec-tral atmospheric general circulation model,Atmosphere-Ocean, 22(4),397-429,1984.

Boer, G. J., G. M.Flato,M.C.Reader, and D. Ramsden,Atransient climate change simulation with historical andprojected greenhouse gas and aerosol forcing: experi-mental design and comparison with the instrumentalrecord for the 20th century, Climate Dynamics, 16,405-426,1999.

Boer, G. J., G. M.Flato,and D. Ramsden,A transient cli -mate change simulation with historical and projectedgreenhouse gas and aerosol forcing:projected climatefor the 21st century, Climate Dynamics, 16,427-450,1999.130


Terrestrial Ecosystems, edited by G.W. Koch and H.A.Mooney,Academic Press,San Diego,California,197-214,1996.

Ebersole,B.A.,Atlantic coast water-level climate,WISReport 7,US Army Corps of Engineers, Washington,DC,1982.

Emanuel,K.A.,The maximum intensity of hurricanes,Journal of the Atmospheric Sciences, 45,1143-1154,1988.

Fisher,A.,et al.,Preparing for a changing climate: Thepotential consequences of climate variability andchange,Mid-Atlantic foundations,US NationalAssessment, http://www.essc.psu.edu/mara/results/foun-dations_report, in review, 2000.

Flato, G. M., G. J. Boer,W. G. Lee, N.A.McFarlane, D.Ramsden,M.C.Reader,A. J.Weaver,The Canadian Centrefor Climate Modelling and Analysis Global CoupledModel and its climate, Climate Dynamics, 16,451-468,1999.

Funderbunk,S.L.,S. J. Jordan, J.A.Mihursky, and D. Riley(Eds.), Habitat Requirements for Chesapeake BayLiving Resources, 2nd edition,Chesapeake ResearchConsortium,Inc.,Solomons,Maryland,1991.

Gibson, J. R.,and R. G. Najjar, Modeling Chesapeake Baysalinity under climate change, Limnology andOceanography, in review, 2000.

Gibson, J. R.,Modeling Chesapeake Bay salinity and phy-toplankton dynamics in response to varying climate,M.S.thesis, Penn State University, University Park, PA,1999.

Gornitz,V.,Regional sea-level variations in eastern NorthAmerica: A geological perspective, Transactions.American Geophysical Union, 80,(17) S85,1999.

Gornitz,V.,and S.Lebedeff, Global sea-level changes dur-ing the past centur y, in Sea-Level Rise and CoastalSubsidence, edited by D. Nummedal, O. Pilkey, and J.Howard, Kluwer Academic Publishers,Dordrecht,Netherlands,357-364,1987.

Groisman, P.Ya ,R.W. Knight,T.R.Karl,and H.F. Lins,Heavy precipitation and streamflow in the UnitedStates: Trends in the 20 th century, Bulletin of theAmerican Meteorological Society, in press,2000.

Hayden,B. P.,Climate change and extratropical stormi-ness in the United States: An assessment, Journal of theAmerican Water Resources Association, 35,1387-1398,1999.

Hughes, P.Y.,E.H.Mason,T. R.Karl,and W.A.Brower,United States historical climatology network daily tem-perature and precipitation data,EnvironmentalSciences Division Publication 3778,Carbon DioxideInformation and Analysis Center, Oak Ridge NationalLaboratory, Oak Ridge,Tennessee,55 pp.,1992.

Idso,S.B.,R.C.Balling,Jr.,and R.S.Cerveny, Carbondioxide and hurricanes:implications of northern hemi-spheric warming for Atlantic/Caribbean storms,Meteorology and Atmospheric Physics, 42,259-263,1990.

Insurance Research Council,Coastal exposure andcommunity protection,Insurance Research Council,Wheaton,Illinois,45 pp.,1995.

Iverson,L.R.,and A.M.Prasad,Predicting abundance of80 tree species following climate change in the easter nUnited States, Ecological Monographs,68(4),465-485,1998.

IPCC (Intergovernmental Panel on Climate Change),Climate Change 1995: The IPCC Second ScientificAssessment, Cambridge University Press,Cambridge,UK,1996. Johns,T. C.,R.E.Carnell, J. F. Crossley, J. M.Gregory, J. F. B.Mitchell,C.A.Senior, S. F. B.Tett,and R.A.Wood,The second Hadley Centre coupled ocean-atmos-phere GCM: Model description,spinup,and validation,Climate Dynamics, 13,103-134,1997.

Jones,K.B.,et al., An Ecological Assessment of theUnited States Mid-Atlantic Region:A Landscape Atlas,US EPA,Of fice of Research and Development:Washington,DC,1997 (Document #EPA/600/R-97/130).

Kalkstein,L.S.,and J. S.Greene,An evaluation of cli-mate/mortality relationships in large U.S.cities and thepossible impacts of climate change, EnvironmentalHealth Perspectives, 105,84-93,1997.

Kalkstein,L.S.,and J. J. Swift,An evaluation of cli-mate/mortality relationships in the Mid-Atlantic regionand the possible impacts of climate change,Preparedfor inclusion in the Mid-Atlantic Regional Assessment,1998.

Karl,T. R., P.Ya.Groisman,R.W. Knight,and R.R.Heim,Jr.,Recent variations of snow cover and snowfall inNorth America and their relation to precipitation andtemperature variations, Journal of Climate, 6,1327-1344,1993.

131


McCabe, G. J.,Jr.,and M.A.Ayers,Hydrologic effects ofclimate change in the Delaware River Basin, WaterResources Bulletin, 25(6),1989.McFarlane, N.A., G. J. Boer, J. P. Blanchet,and M.Lazare,The Canadian Climate Centre second-generation gener-al circulation model and its equilibrium climate,Journal of Climate , 5,1013-1044,1992.

Mitchell J. F. B.,T. C. Johns, J. M.Gregory, and S.Tett,Climate response to increasing levels of greenhousegases and sulphate aerosols, Nature, 376,501-504,1995.

Mitchell J. F. B.,and T. C. Johns,On modification of glob-al warming by sulfate aerosols, Journal of Climate, 10,245-267,1997.

Najjar, R. G.,et al., The potential impacts of climatechange on the mid-Atlantic coastal region, ClimateResearch, 14,219-233,2000.

Neff, R.,H.Chang,C. G. Knight,R. G. Najjar, B.Yarnal,and H.A. Walker, Impact of climate variation andchange on mid-Atlantic hydrology and water resources,Climate Research, 14,207-218,2000.

Neilson,R. P.,and R. J. Drapek, Potentially complex bios-phere responses to transient global warming, GlobalChange Biology, 4,505-521,1998.

NPA Data Services,Inc.,Regional Economic ProjectionSeries,Washington,DC,1998.

Perry, M.C.,and A.S.Deller, Review of factors af fectingthe distribution and abundance of waterfowl in shal-low-water habitats of Chesapeake Bay, Estuaries, 19,272-278,1996.

Perry, M.C.,and A.S.Deller,Waterfowl populationtrends in the Chesapeake Bay area,in Toward aSustainable Coastal Watershed: The ChesapeakeExperiments. Proceedings of a Conference,Chesapeake Research Consortium Publication No.149,edited by P. Hill and S.Nelson,Chesapeake ResearchConsortium,Inc.,Edgewater, Maryland,1995.

Perry, M.C.,(Ed.), F. M.Uhler, Food habits and distribu-tion of wintering canvasbacks,Aythya valisinerina,onChesapeake Bay, Estuaries, 11,57-67,1988.

Polsky, C., J.Allard, N. Currit,R. G. Crane,and B.Yarnal,The mid-Atlantic region and its climate:past,present,and future, Climate Research, 14, 161-173.

Karl,T. R.,and R.W. Knight, Secular trends in precipita-tion amount,frequency, and intensity in the UnitedStates, Bulletin of the American MeteorologicalSociety, 79,231-241,1998.

Karl,T. R.,R.W. Knight, D. R.Easterling,and R. G. Quayle,Trends in U.S. climate during the Twentieth century,Consequences, 1,2-12,1996.

Kearney, M.S.,and J. C.Stevenson,Island land loss andmarsh vertical accretion rate evidence for historicalsea-level changes in Chesapeake Bay, Journal ofCoastal Research, 7,403-415,1991.

Kilbourne,E.M.,K.Cho,T. S. Jones,S.B.Thacker, and F. I.Team,Risk factors for heat stroke: A case central study,JAMA, 247,3332-3336,1982.

Kittel,T. G. F., J.A.Royle,C.Daly, N.A.Rosenbloom,W. P.Gibson,H.H. Fisher, D. S.Schimel,L.M.Berliner, andVEMAP2 participants,A gridded historical (1895-1993)bioclimate dataset for the conterminous United States,in Proceedings of the 10th Conference on AppliedClimatology, October 20-24,1997,Reno,Nevada,American Meteorological Society, Boston,Massachusetts,219-222,1997.

Klopatek, J. M.,R. J. Olson,C. J. Emerson,and J. L. Jones,Land-use conflicts with natural vegetation in theUnited States, Environmental Conservation, 6,191-200,1979.

Knutson,T. R.,R.E.Tuleya,and Y. Kurihara,Simulatedincrease of hurricane intensities in a CO2-warmed cli-mate, Science, 279,1018-1020,1998.

Kunkel,K.E.,K.Andsager, and D. R.Easterling,Long-Term trends in extreme precipitation events over theconterminous United States, Journal of Climate, 12,2515-2527,1999.

Kunkel,K.E.,R.A.Pielke Jr.,and S.A.Changnon,Temporal fluctuations in weather and extremes thatcause economic and human health impacts, Bulletin ofthe American Meteorological Society, 80, 1077-1098,1999.

Lighthill, J., G. Holland,W. Gray, C.Landsea, G. Craig, J.Evans,Y. Kurihara,and C.Guard,Global climate changeand tropical cyclones, Bulletin of the AmericanMeteorological Society, 75,2147-2157,1994.

Martens, P.,How will climate change affect humanhealth?, American Scientist, 87, 534-541,1999.

132


Powell, D. S., J. L. Faulkner, D. R.Darr, Z.Zhu,and D.W.MacCleery, Forest Resources of the United States,1992, USDA Forest Service,133 pp.(General TechReport,RM-243),1996.

Ricketts,T. H.,E.Dinerstein, D. M.Olson,and C.Loucks,Who’s where in North America? Patterns of speciesrichness and the utility of indicator taxa for conserva-tion. BioScience, 49, 369-381,1999.

Rock,B.,and B.Moore III et al.,The New Englandregional assessment of the potential consequences ofclimate variability and change,US National Assessment,in prep.2000.

Roth,S.,E. P. McDonald,and R.L.Lindroth,AtmosphericCO2 and soil water availability consequences for tree-insect interactions, Canadian Journal of ForestResearch, 27,1281-1290,1997.

Rose,A.,Y. Cao,and G. Oladosu,Simulating the econom-ic impacts of climate change in the Mid-Atlantic region,Climate Research, 14,175-183,2000.

Rosenzweig,C.,and W. Solecki,et al.,Climate changeand a global city:an assessment of the metropolitanEast Coast region,US National Assessment,in review,2000.

Solley,W. B.,R.R.Pierce,and H.A. Perlman,Estimatedwater use in the United States in 1995, US GeologicalSurvey Circular 1200 ,Washington,DC,1998.

Stephenson, D. B.,and I.M.Held,GCM response ofnorthern winter stationary waves and storm tracks toincreasing amounts of carbon dioxide, Journal ofClimate, 6,1859-1870,1993.

Timmermann,A., J. Oberhuber,A.Bacher, M.Esch,M.Latif, and E.Roeckner. 1999. Increased El Niño fre-quency in a climate model forced by future greenhousewarming. Nature, 398,694-696.

US Army Corps of Engineers, Annual Flood DamageReport to Congress for Fiscal Year 1997, 1998.Available online at http:/www.usace.army.mil/inet/functions/cw/cecwe/flood.htm.

USDA Forest Service, Report of the Forest Service.Fiscal Year 1998,Washington,DC,1998.

US Global Change Research Program Forum on GlobalChange Modeling.Washington, D.C.USGCRP Report 95-02,Washington,DC,1995.

US Historical Climatology Network: Available on line athttp://www.ncdc.noaa.gov/ol/climate/research/ushcn/daily.html

Valverde,H.R.,A.C.Trembanis,and O. H.Pilkey,Summary of beach nourishment episodes on the U.S.East Coast barrier islands, Journal of Coastal Research,15,1100-1118,1999.

VEMAP members,Vegetation/Ecosystem Modeling andAnalysis Project (VEMAP):Comparing biogeographyand biogeochemistry models in a continental-scalestudy of terrestrial ecosystem responses to climatechange and CO2 doubling, Global BiogeochemicalCycles, 9,407-437,1995.

Walker, H.A., J. S.Latimer, and E.H.Dettman,Assessingthe effects of natural and anthropogenic stressors inthe Potomac Estuary: Implications for long-term moni-toring, Environmental Monitoring and Assessment, inpress,2000.

Watson,R.T.,M.C.Zinyowera,and R.H.Moss (eds).The Regional Impacts of Climate Change: AnAssessment of Vulnerability. Cambridge UniversityPress,New York,1998.

Williamson,M.,Invasions, Ecography, 22,5-12,1999.

Yarnal,B., D. L. Johnson,B. Frakes, G. I.Bowles,and P.Pascale,The flood of ’96 in the Susquehanna Riverbasin, Journal of the American Water ResourcesAssociation, 33,1299-1312,1997.

133


Mid-Atlantic Workshop and Assessment TeamsAnn Fisher*, Pennsylvania State UniversityDavid Abler, Pennsylvania State UniversityEric J. Barron, Pennsylvania State UniversityRichard Bord, Pennsylvania State UniversityRobert Crane, Pennsylvania State UniversityDavid DeWalle, Pennsylvania State UniversityC.Gregory Knight, Pennsylvania State UniversityRay Najjar, Pennsylvania State UniversityEgide Nizeyimana, Pennsylvania State UniversityRobert O’Connor, Pennsylvania State UniversityAdam Rose, Pennsylvania State UniversityJames Shortle, Pennsylvania State UniversityBrent Yarnal, Pennsylvania State University

New England and Upstate New York Workshop andAssessment Teams

Barry Rock*,University of New HampshireBerrien Moore III*,University of New HampshireDavid Bartlett,University of New HampshirePaul Epstein,Harvard School of Public HealthSteve Hale,University of New HampshireGeorge Hurtt,University of New HampshireLloyd Irland,Irland Group,MaineBarry Keim,New Hampshire State climatologistClara Kustra,University of New HampshireGreg Norris,Sylvatica Inc.,MaineBen Sherman,University of New HampshireShannon Spencer, University of New HampshireHal Walker, EPA,Atlantic Ecology Division,Rhode

Island



Metropolitan East Coast Workshop and AssessmentTeams

Cynthia Rosenzweig*,National Aeronautics andSpace Administration,Goddard Institute forSpace Studies,and Columbia University

William Solecki*,Montclair State UniversityCarli Paine,Columbia UniversityPeter Eisenberger, Columbia University Earth

InstituteLewis Gilbert,Columbia University Earth InstituteVivien Gornitz,Columbia University Center for

Climate Systems ResearchEllen K.Hartig,Columbia University Center for

Climate Systems ResearchDouglas Hill,State University of New York,Stony

BrookKlaus Jacob,Lamont-Doherty Earth Observatory of

Columbia UniversityPatrick Kinney,Columbia University Joseph A.

Mailman School of Public HealthDavid Major, Columbia University Center for Climate

Systems ResearchRoberta Balstad Miller, Center for International Earth

Science Information Network (CIESIN)Rae Zimmerman,New York University Institute for

Civil Infrastructure Systems, Wagner School

134


137

CHAPTER 5

POTENTIAL CONSEQUENCES OF CLIMATEVARIABILITY AND CHANGE FOR THESOUTHEASTERN UNITED STATESVirginia Burkett1,2, Ronald Ritschard3, Steven McNulty4, J. J. O’Brien5, Robert Abt6,James Jones7, Upton Hatch8, Brian Murray9, Shrikant Jagtap7, and Jim Cruise3


Chapter Summary

Physical Setting


Ecological Context


Key Issues

Weather-related Stresses on Human Populations

Agricultural Crop Yields and Economic Impacts

Forest Productivity Shifts

Water Quality Stresses

Threats to Coastal Areas

Additional Issues

Climate Model Limitations

Water Resources

Impacts on Coastal Ecosystems and Services

Health Issues Related to Water Quality

Socioeconomic and Insurance Issues

Urban Issues


Literature Cited

Acknowledgments

1US Geological Survey; 2Coordinating author for the National Assessment Synthesis Team; 3University of Alabama inHuntsville; 4USDA Forest Service; 5Florida State University; 6North Carolina State University; 7University of Florida;8Auburn University; 9Research Triangle Institute


• Southeastern temperature trends varied betweendecades over the past 100 years,with a warmperiod during the 1920s-1940s followed by adownward trend through the 1960s. Since the1970s,temperatures have been increasing,with1990’s temperatures reaching peaks as high asthose of the 1920s-1940s.

• Annual rainfall trends show very strong increasesof 20-30% or more over the past 100 years acrossMississippi,Arkansas,South Carolina,Tennessee,Alabama, and parts of Louisiana,with mixedchanges across most of the remaining area. Thepercentage of the Southeast landscape experi-encing severe wetness increased approximately10% between 1910 and 1997.


• Climate model simulations provide plausible sce-narios for both temperature and precipitationover the 21 st century in the Southeast. Both ofthe principal climate models used in the NationalAssessment suggest warming in the Southeast bythe 2090s,but at different rates.

• The Canadian model scenario shows theSoutheast experiencing a high degree of warm-ing,which translates into lower soil moisture ashigher temperatures increase evaporation. TheHadley model scenario simulates less warmingand a significant increase in precipitation by2100 (about 20% on average,but with some dif-ferences within the region).

• Both climate model simulations indicate that theheat index (a measure of comfort based on tem-perature and humidity) will increase more in theSoutheast than in other regions. Heat index inthe Southeast is projected to rise by as much as8 to 15ºF (4 to 8ºC) in the Hadley model and byover 15ºF (8ºC) across the entire region in theCanadian model simulation by 2100.


The Southeast “sunbelt”is a rapidly growing regionwith a population increase of 32% between 1970and 1990. Much of this growth occurred in coastalcounties,which are projected to grow another 41%between 2000 and 2025. The number of farms inthe region decreased 80% between 1930 and 1997as the urban population expanded,but theSoutheast still produces roughly one quarter of USagricultural crops. The Southeast has becomeAmerica’s “woodbasket,”producing about half ofAmerica’s timber supplies. The region also producesa large portion of the nation’s fish,poultry, tobacco,oil,coal,and natural gas. Prior to European settle-ment,the landscape was primarily forests, grass-lands,and wetlands. Most of the native forests wereconverted to managed forests and agricultural landsby 1920. Although much of the landscape has beenaltered,a wide range of ecosystem types exists andoverall species diversity is high.

138


Key Findings

• The increase in the southeastern summer heatindex simulated by both the Hadley andCanadian climate models would likely af fecthuman activity and possibly demographics in theSoutheast during the 21st century.

• Agriculture could possibly benefit from increasedCO2 and modest warming (up to 3 to 4ºF, or 2ºC)as long as rainfall does not decline,but there aredifferences in individual crop responses.Management adaptations could possibly offsetpotential losses in individual crop productivitydue to increased evapotranspiration.

• Biological productivity of pine and hardwoodforests will likely move northward as tempera-tures increase across the eastern US. Hardwoodsare more likely to benefit from increases in CO2and modest increases in temperature than pines.Physiological forest productivity and ecosystemmodels suggest that,without management adap-tations,pine productivity is likely to increase by11% by 2040 and 8% by 2100 across theSoutheast compared to 1990 productivity. Thesemodels suggest that hardwood forest productivitywill likely increase across the region, by 25% by2090 compared to 1990 regional hardwood pro-ductivity.

• Under the Hadley model scenario,the region’sland use change in the next century is likely to

be dominated by non-climate factors such ascommodity prices and demographic forces.Urbanization will likely continue to convert for-est and agricultural land,while continued move-ment of land from agriculture to forest is alsoexpected. Under more extreme climate scenar-ios,land reallocation could possibly be more dra-matic as the productivity of land-based activitiessuch as forestry and agriculture is more pro-foundly influenced by climate.

• During the 21st century, the IPCC projects thatsea-level rise will likely accelerate 2- to 5 -foldcompared to the global average rate during the1900s (which was 4-8 inches). This would verylikely have dramatic effects on population cen-ters,infrastructure,and natural ecosystems in thelow-lying Gulf and South Atlantic coastal zone.

• Water and air quality are concerns given thechanges in temperature and precipitation thatare simulated by climate models.

• Changes in minimum temperature, rainfall,andCO2 will likely alter ecosystem structure,butinteractions are difficult to model or predict,par-ticularly relative to disturbance patterns.

• Changes in fresh water and tidal inflows intocoastal estuaries will likely alter the ecologicalstructure and function of these highly productiveand valuable ecosystems.

139

Chapter 5 / Southeastern United States

140


about 25% (NPA,1999). However, warming at high-er latitudes combined with increased heat stress inthe southeastern US may serve to decrease popula-tion migration towards the Southeast.

Based on the 1990 Census, about 61% of theSoutheast’s population is considered urban. Thenumber of farms in the region decreased 80%between 1930 and 1997 (USDA,1999). The 20thcentury was one of dramatic transition from anagrarian economy to one based on a combination ofnatural resources,manufacturing and trade,technol-ogy, and tourism. Roughly one half of 1990 employ-ment fell in the categories of manufacturing andwholesale/retail trade,compared with an average ofless than 5% in agriculture (US Bureau of theCensus,1994). Prior to 1950,corn and cotton werethe most important crops. Today agriculture is morevaried;soybean and hay outweigh cotton in acreageharvested and rice has become increasingly impor-tant. The Southeast still produces roughly one quar-ter of the nation’s agricultural crops,but timber har-vests are more valuable in terms of annual econom-ic impact in most states. Forest products industrieswere among the top four manufacturing employersin Mississippi,Alabama,North Carolina and Georgiain 1997. The Southeast has become America’s“woodbasket,” producing about half of America’stimber supplies. The region is also responsible for alarge portion of the nation’s fisheries,poultry, tobac-co,oil,natural gas,bauxite,coal,and sulfur produc-tion.

According to the 1990 census,18% of the popula-tion of the Southeast region lives below the povertylevel. The most poverty-prone areas include theLower Mississippi River Valley and parts ofAppalachia. While certain measurements,such asper capita income,have moved in the direction ofthe national averages,poverty rates in some areasare as much as two and a half times the nationalaverage. Levels of education of the population insome areas also lag behind national standards. Someof the smaller, more remote and geographically iso-lated areas of the region suffer from a lack of eco-nomic opportunities,have significant dependentpopulations,and lack the public institutions needed

POTENTIAL CONSEQUENCES OF CLIMATEVARIABILITY AND CHANGE FOR THESOUTHEASTERN UNITED STATES

PHYSICAL SETTING

The Southeast region (Fig.1) represents 15.4% ofthe land area of the US and 22.6% of its citizenry(US Bureau of the Census,1994). Although theSoutheast has considerable variation in landforms,itis possible to divide it into five fairly distinct regionsbased on physical geography. The lower third of thelow, flat Florida peninsula is a sub-tropical provincewith unique features such as the Everglades and theFlorida Keys. The Coastal Plain,which dominatesthe region,is a broad band of territory parallelingthe Gulf and South Atlantic seacoast from Virginia toTexas,with a deep extension up the lowerMississippi River valley. The Coastal Plain is relative-ly flat,with broad,slow-moving streams and sandyor alluvial soils. The Piedmont is a slightly elevatedplateau that begins at the “fall line,”where rivers cas-cade off the eastern edge of the plateau onto theCoastal Plain,and ends at the AppalachianMountains. This land is rolling to hilly, with manystreams. The Highlands comprise the inland moun-tain regions and include the southernmostAppalachian Mountains in the east and the Ozarkand Ouachita Mountains in the west. The InteriorPlains stretch into the north-central portion of theregion,including parts of Tennessee and Kentucky.The southern portions of the State of Virginia areincluded in aspects of this regional analysis. Thenorthern portions are included in the Northeastchapter with a discussion of the Chesapeake Baywatershed.

SOCIOECONOMIC CONTEXT The 544,000 square mile southeastern “sunbelt”isone of the fastest growing regions in the US. Thesoutheastern population increased by 21%,morethan double the national rate,between 1970 and1980 and another 11% between 1980 and 1990.Much of the historical population growth fromTexas to North Carolina occurred in the 151 coun-ties within the southeastern coastal zone,which areprojected to grow another 41% between 2000 and2025,compared to the projected national average of

141


southern boundaries while expanding west intoparts of eastern Oklahoma and Texas,and north intoparts of Missouri,West Virginia, Kentucky, andVirginia. Water stress and increased fire disturbancerestricts the Southeast forest under the Canadian cli-mate scenario,and large areas of the Southeast areconverted to savanna (grasslands with scatteredtrees and shrubs) and the Southeast forest movesinto the northcentral part of the US.

One of the biogeochemistry models (TEM) used inthe National Assessment projects large differences incarbon storage for the Southeast depending uponthe climate scenario used. Under the Canadian cli-mate model scenario, vegetation is projected to loseup to 20% of its carbon mass by 2030. However,under the Hadley climate scenario the same biogeo-chemistry model indicates that vegetation will addbetween 5 and 10% to its carbon mass over the next30 years. These differences in carbon storage reflectthe differences in climate scenario projections fortemperature and precipitation that are greatest inthe southeastern part of the US (Felzer and Heard,1999).

Ecological models used in the National Assessmentdo not simulate species-level response,nor do theysimulate land use changes,invasive species impacts,or other influences on ecosystems that cannot beeffectively modeled based on historical or empiricalevidence,unless the ecological models are linked

to support progressive development (Glasmeier,1998). These distressed counties present a profoundchallenge to policy makers concerned about climatechange mitigation strategies and issues,particularlyin the Appalachian coal-producing and Gulf Coastoil-producing regions.

ECOLOGICAL CONTEXTPrior to European settlement,the Southeast wasdominated by upland forests, grasslands,and wet-lands. Nearly one-third of the region may have beenwetland (Dahl,1990),but by 1990, wetlands hadbeen reduced to about 16% of the southeasternlandscape (Hefner, et.al.,1994). A wide range ofecosystem types is presently found in the region,ranging from coastal marshes to high-elevationspruce fir forests. Diversity of both plant and animalspecies is high compared to other regions consider-ing the extent of landscape alteration that hasoccurred. On an area basis,the Southeast has rela-tively high overall species richness indices (Ricketts,et.al.,1999). Vascular plant diversity is second onlyto Puerto Rico. Tree species richness is greatest inthe Southeast,with approximately 180 tree speciesfound in parts of South Carolina and more than 140tree species identified in most of the remainder ofthe region.

Forests still dominate parts of the Southeast;theshare of forestland in each state averages about 30%.About 20% of the present forests exist as pine plan-tations. Native longleaf pine was the predominantspecies in the Coastal Plain in the late 1800s butless than 3 million of the 60 million acres of south-eastern longleaf pine remain today (Boyer, 1979).More than 60 species of mammals occur in a rela-tively small area of the southern Appalachian moun-tains,while 40 or fewer mammal species are foundin the Coastal Plain (Currie,1991). The region hasvery high diversity of amphibians and reptiles.Roughly half of the remaining wetlands in the USare located in the Southeast,and more than three-quarters of the nation’s annual wetland losses overthe past 50 years occurred in the region.

Two types of ecosystem models (biogeochemicaland biogeographical) show a wide range of poten -tial changes in vegetation in the Southeast duringthe 21st century, depending upon the climate sce-nario selected. One of the biogeographical models(MAPSS) projects significant shifts in major biomesunder the Canadian climate scenario,but not underthe Hadley climate scenario. Under the Hadley cli -mate scenario,the Southeast mixed forest retains its

Figure 1. The Southeast region includes all of nine states(Alabama, Florida, Georgia, Kentucky, Louisiana, North Carolina,Mississippi, South Carolina, and Tennessee), the southern portionof Virginia, and 50 counties in east Texas. Four subregional work-shops were conducted in the Southeast region. See Color PlateAppendix

Land Cover Map of the Southeast Region

in assuming what the net effects of CO2 enrichmentmight be across a region or biome.

CLIMATE VARIABILITY ANDCHANGEPast Century

The Southeast has some of the warmest conditionsin the US. However, it is the only region to showwidespread but discontinuous cooling periods of 2to 3.5ºF (1 to 2ºC) over almost the entire area dur-ing the past 100 years (Figure 2). Peninsular Florida,North Louisiana,and a few small areas in theAppalachian Mountains have shown a modest warm-ing of around 2°F (1ºC) since 1900. The reason theSoutheast temperature record shows a net coolingtrend over the past 100 years is that there was awarm period between the 1920s and 1940s,then asignificant downward trend through the late 1960s.The mid-1900s cooling trend may have been due tonatural variation. Human-caused sulfate aerosolemissions during this period may have also playedsome role. Sulfate aerosols reflect some sunlightback into space,thereby cooling the atmosphere(Kiehl,1999). Since 1970,the average annual tem-perature increased,with the most significantincreases occurring during the 1990s. Trends intemperature extremes over the past one hundredyears exhibit a decrease of about 5 days in the num-ber of days per year exceeding 90ºF (32.2ºC),and anincrease of 6 days in the number of days belowfreezing over the entire region. However, over thepast fifty years the average annual length of thesnow season decreased by 4 days.

The Southeast receives more rainfall than any otherregion. Annual precipitation trends show increasesof 20-30% or more over the past 100 years acrossMississippi,Arkansas,South Carolina,Tennessee,Alabama,and parts of Louisiana,with mixed changesacross most of the remaining area. The southernmountains of North Carolina,the southern tip ofTexas,and a couple of other small areas have slight-ly decreasing trends in annual precipitation. Muchof the increase in precipitation was associated withmore intense events (rainfall greater than 2 inchesor 5 cm per day). A small percentage of theincreased precipitation was associated with moder-ate rainfall events,which are generally beneficial toagriculture and water supply. Analysis of streamflow trends during 1944-1993 showed little changein annual maximum daily discharge,but significantincreases in annual median and minimum flows in

with other process models,both biological and eco-nomic. For example,Harcombe and others (1998)observed that Chinese tallow, a freeze-intolerantnon-native tree species,increased dramatically insoutheastern Texas over the past few decades.Chinese tallow increased by a factor of 30 between1981 and 1995,often out-competing native specieswhen canopy gaps form in mesic (medium mois-ture) and wet sites (Harcombe,et.al.,1998). Thesekinds of interactions and changes in forest dynamicsare difficult to simulate.

Mixed responses among species to fertilizationeffects of elevated atmospheric CO2 further con-found our ability to model ecosystem structure andproductivity. Several studies showed that elevatedCO2 increased photosynthesis rates and improvedwater use efficiency in many forest species and agri-cultural crops (Acock,et.al.,1985;Allen,et.al.,1989;Nijs,et.al.,1988). Two reviews of CO2 exposurestudies with deciduous and coniferous speciesfound that increases in growth rates varied widely,but that generally tree growth was stimulated byincreases in CO2 (Eamus and Jarvis,1989;NCASI,1995). However, limits on the availability of soilnutrients and water in many natural or semi-naturalecosystems can severely constrain the potentialimprovement in water use ef ficiency due to sup-pressed transpiration induced by enhanced CO2 lev-els,thereby offsetting potential gains in productivity(Lockwood,1999). Temperature,plant pests,air pol-lution,and light availability could also limit thepotential enhancement of growth by elevated CO2(NCASI 1995). Hence,one should be very cautious

142


Figure 2. Decadal average temperatures in the Southeast.Source D. Eastering, NOAA National Climatic Data Center. SeeColor Plate Appendix

Southeast US Annual Mean Temperature

the lower Mississippi Valley, and decreases in thesecategories in parts of Georgia and North Carolina(Linns and Slack,1999). Increased precipitationintensity in extreme events during the next centur yis suggested by climate models under doubled CO2

for the US (Mearns,et.al.,1995) and there is evi-dence that moisture in the atmosphere is increasingover the Caribbean region (Trenberth,1999).Heavy rains are less efficient (more water runs offinto the sea) and are more likely to cause flooding,which is a serious problem in the region.

Trends in wet and dry spells during the 20th centu-ry, as indicated by the Palmer Drought SeverityIndex (PDSI),are spatially consistent with theregion’s annual precipitation trends,showing astrong tendency to more wet spells in the GulfCoast states,and a moderate tendency in mostother areas. The percentage of the southeasternlandscape experiencing “severe wetness”(periodsin which the PDSI averages more than +3)increased approximately 10% between 1910 and1997.

Effects of El Niño on Climate in the SoutheastThe El Niño/Southern Oscillation (ENSO) phenome-non contributes to variations in temperature andprecipitation that complicate longer-term climatechange analysis in certain parts of the country, par-ticularly the Southeast. ENSO is an oscillationbetween warm and cold phases of sea-surface-tem-perature (SST) in the tropical Pacific Ocean with acycle period of 3 to 7 years. US climate anomalies(departures from the norm) associated with ENSOextremes vary both in magnitude and spatial distri-bution. El Niño events (the warm phase of theENSO phenomenon) are characterized by 2 to 4°F(about 1 to 2ºC) cooler average wintertime air tem-peratures in the Southeast. During the spring andearly summer months,the region returns to near-normal temperatures. Precipitation anomaly pat-terns following warm events indicate that GulfCoast states encounter wetter than normal winters(by about 1 to 2 inches per month). By the spring,the entire eastern seaboard shows increased precip-itation. In summer, climate impacts of warm eventsare more localized; for example,drier conditions arefound in eastern coastal regions,and from northTexas to northern Alabama. El Niño events also cre-ate upper atmospheric conditions that tend toinhibit Atlantic tropical storm development, result-ing in fewer hurricanes,while La Niña events havethe opposite effect, resulting in more hurricanes.Figure 3 depicts US Gulf of Mexico hurricane land-fall trends and the probability of hurricane landfallduring El Niño and La Niña years.

143


Figure 3(a). US Hurricane Landfall Trends in the Gulf of Mexico.This figure shows the number of US hurricanes making landfall inthe Gulf of Mexico by decade for the past 100 years. There werepeaks in activity during the 1910s and 20s, as well as a lower peakin the 1960s. The past 30 years have shown a decrease in the num-ber and intensity of Gulf hurricanes making landfall. See ColorPlate Appendix

Gulf Landfalling Hurricanes by Decade

Figure 3(b). Effect of ENSO Phase on Hurricane LandfallThis figure shows the probability of the number of hurricane land-falls on the US in a given hurricane season and ENSO phase (ElNiño, Neutral, La Niña). Based on the past 100-year record, theprobability of at least 1 hurricane landfall is similar for all threephases, with probabilities ranging from 78% for El Niño to 90% forLa Niña. For multiple landfalls, however, the differences caused byENSO phase become apparent. The probability of at least 2 land-falls during El Niño is 28%, but is 48% in neutral years, and 66%during La Niña. The probability of at least 3 landfalling hurricanesis near 0% for El Niño, 20% for neutral years, and 50% for La Niña.It is clear that El Niño years have few multiple hurricane strikes onthe US, while neutral years and La Niña years often see multiplehurricane strikes on the US coast. (Source: Florida StateUniversity, Center for Ocean-Atmosphere Prediction). See ColorPlate Appendix

US Hurricanes

144


During La Niña events (the cold phase of ENSO),theanomalies are sometimes reversed from those asso-ciated with warm events,but not everywhere.Above-average wintertime temperatures are presentEast of the Mississippi. By spring,the warmer anom-alies in the east are focused in the Ohio Valley andnorthern Florida,Georgia,and South Carolina.Wintertime precipitation patterns associated withcold events show increases (1-2 inches or 2.5-5 cmper month) in the band stretching from northernMississippi to southwestern Pennsylvania. In thespring,Gulf Coast areas have increased precipita-tion. In summer, the extreme southern US is colderthan normal and greater precipitation is evident inthe Southeast. Dry to very dry conditions are foundin parts of Texas and Louisiana. Thus,as suggestedby these results,the climate anomalies associatedwith opposite phases of ENSO are not in directopposition. For example, climate anomalies inFlorida are nearly opposite (for cold and warmevents) while those in many of the midwesternstates are of the same sign for both precipitationtotals and temperatures during both warm and coldevents. Further evidence demonstrates that climateanomalies associated with strong warm events arenot amplifications of normal warm events(Rosenberg,et al.1997).

Scenarios for the Future

The Hadley Centre climate model projects that by2030,maximum summer temperatures in the

Southeast will increase by about 2.3°F (1.3ºC) whilemaximum winter temperatures will increase by1.1°F (0.6ºC). The projected increase in mean annu-al temperature of 1.8ºF (1.0ºC) by 2030 and 4.1ºF(2.3ºC) by 2100 represents a smaller degree of pro-jected warming than for any other region. Thesmaller simulated warming rate is possibly due tothe buffering affects of the oceans,large amounts ofsurface water for evaporative cooling,and the sul-fate aerosol emissions that are prevalent throughoutthe eastern US. Sulfate aerosols may help explainthe mid-20th century cooling trend in parts of theUS;however, over the past two decades,sulfateemissions decreased,and the future cooling affect ofsulfate aerosols is not expected to be as importantdue to Clean Air Act restrictions. Although theincrease in temperature under the Hadley model issmall compared to other regions,the resultingincrease in the summer heat index by 8 to15ºF (4 to8ºC) (calculated from monthly maximum tempera-ture and relative humidity) would likely af fecthuman activity and,possibly,demographics in the21st century.

The Canadian Centre climate model projects highertemperature scenarios and higher southeastern heatindices by the end of the 21st century than does theHadley model. The Canadian model simulates anincrease in mean annual southeastern temperatureof about 3ºF (1.7ºC) by 2030 and 10ºF (5.5ºC) by2100. In the Canadian model,increases in maxi-mum summer temperature are the highest in theNation for both 2030 (5ºF or 2.8ºC) and 2100 (12°For 6.5ºC). The Canadian climate model simulates anincrease in average summer heat index above 15ºF(8ºC) across the entire region by 2100 (Figure 4).

Another important difference between the twomodels for the Southeast lies with the simulatedchanges in rainfall;the Canadian climate model simu-lates reduced average annual precipitation (10% lessthan present by 2090) while the Hadley model simu-lates more precipitation than present (20% more by2090). These differences have important implica-tions for hydrologic impacts on the Southeast,because the Canadian model simulates decreasedsoil moisture,while the Hadley model simulatesincreased soil moisture (Felzer and Heard,1999).Differences between the two models are illustratedin Figure 5,which depicts the simulated summer cli-mate in Georgia in 2030 and 2090. According to theHadley climate model scenario,the Southeast willremain the wettest region of the US for the next 100years. The precipitation changes projected by theHadley model by 2100 are consistent with otherparts of the eastern and midwestern US.

July Heat Index Change - 21st Century

Canadian Model Hadley Model+25ºF

+20ºF

+15ºF

+10º

+5ºF

0º

+25ºF

+20ºF

+15ºF

+10º

+5ºF

0º

Figure 4. The changes in the simulated heat index for theSoutheast are the most dramatic in the nation with the Hadleymodel suggesting increases of 8 to 15ºF for the southern-moststates, while the Canadian model projects increases above 20ºF formuch of the region. Heat indices simulated for the Southeast by2100. (source, NOAA National Climatic Data Center). See ColorPlate Appendix

145


The Max Planck Institute climate model (ECHAM4/OPYC3),one of a few models with suf ficient resolu-tion in the tropics to adequately simulate narrowequatorial upwelling and low frequency waves,sim-ulates more frequent El Niño-like conditionsand stronger La Niñas under a doubling of CO2,which is consistent with the Hadley model pro-jections with a doubling of CO2. The Max-Planck model also suggests that the mean cli-mate in the tropical Pacific region will shifttoward a state corresponding to present-day ElNiño conditions (Timmermann,et.al.,1999).McGowan,Cayan,and Dorman (1998) showedthat the frequency of warm sea surface eventsoff the western coast of North Americaincreased since 1977,but relationships betweenthis trend and reduced hurricane landfall in theGulf Coast region have not been established.

Table 1: Billion-Dollar Southeast Weather Disasters, 1980-1999

Disaster Year Estimated Damages/ EstimatedCosts* Deaths**

AR/TN Tornadoes 1999 $ 1.3 billion 17TX Flooding 1998 $ 1.0 billion 31Hurricane Georges 1998 $ 5.9 billion 16Hurricane Bonnie 1998 $ 1.0 billion 3Southern Drought/Heat Wave 1998 $ 6.0-9.0 billion 200El Niño/Tornadoes and floods 1998 $ 1.0 billion 132MS/OH Valley Floods/Tornadoes 1997 $ 1.0 billion 67Hurricane Fran 1996 $ 5.0 billion 37Hurricane Opal 1995 $ 3.0 billion 27TX/OK/LA/MS Severe Weather 1995 $ 5.0-6.0 billion 32TX Flooding 1994 $ 1.0 billion 19Tropical Storm Alberto 1994 $ 1.0 billion 32Southeast Ice Storm 1994 $ 3.0 billion 9Summer Drought/Heat Wave 1993 $ 1.0 billion ***Hurricane Andrew 1992 $ 27.0 billion 58Hurricane Bob 1991 $ 1.5 billion 18TX/OK/LA/AR Flooding 1990 $ 1.0 billion 13Hurricane Hugo 1989 $ 9.0 billion 57Hurricane Juan 1985 $ 1.5 billion 63Hurricane Elena 1985 $ 1.3 billion 4Florida Freeze 1985 $ 1.2 billion 0Florida Freeze 1983 $ 2.0 billion 0Hurricane Alicia 1999 1983 $ 3.0 billion 21Total $ 83.7-87.7 billion 856

* not adjusted for inflation,** US only, *** undetermined (Source:National Climatic Data Center, 1999)

Figure 5. Illustration of how the summer and winter climates inGeorgia would shift under the Hadley climate scenario (HADCM2).For example, the summer climate in Georgia in the 2030s would bemore like the current climate of the Florida panhandle. Source:NOAA, National Climatic Data Center. See Color Plate Appendix

Summer Climate Changes from Hadley Centre Scenario

KEY ISSUES1. Weather-related Stresses on Human Populations 2. Agricultural Crop Yields and Economic Impacts3. Forest Productivity Shifts4. Water Quality Stresses5. Threats to Coastal Areas

Changes in average climate and weather extremeshave important economic implications in theSoutheast. There are several reasons why thisregion is of relatively high interest and concern.First,there is a strong ENSO signal,primarily inthe Gulf Coast states,that results in seasonal andyear-to-year variations in temperature and precipi-tation. Understanding potential future climatechange in the context of current natural variabili-ty can provide an important contribution to ongo-ing discussions of mitigation options. A secondconsideration is that the Southeast experiencesmany extreme climate events such as hurricanes,heat waves,tornadoes,ice storms, floods,andlightning storms that cause significant economiclosses to industry and local communities. The agri-culture and forestry sectors make substantial contri-butions to the regional and national economy andthese sectors are quite vulnerable to climate vari-ability. Water resources,air quality, coastalresources,and land use are other important regionalissues that may be strongly influenced by climatictrends and variability.

1. Weather-related Stresses onHuman Populations

The US experienced 42 weather-related disastersover the past 20 years that resulted indamages/costs in excess of $1 billion each;23 ofthese disasters occurred in Southeast states, result-ing in total damages/costs of about $85 billion.Most of the property damages were associated withfloods and hurricanes. Low-lying Gulf and SouthAtlantic coastal counties are particularly vulnerableto storm surge. Between 1978 and 1998,56% of theNational Flood Insurance Program (NFIP) policies inforce and 74% of total NFIP claim paymentsoccurred in southeastern coastal counties (HeinzCenter, 1999).

In addition to the projected shift towards more fre-quent El Niño-like conditions in the Southeast,someclimate models suggest that rainfall associated withEl Niño events will increase as atmospheric CO2increases. Increased flooding in low-lying coastalcounties from the Carolinas to Texas could possibly146


have adverse effects on human health. Floods arethe leading cause of death from natural disasters inthe Southeast and nationwide. Flooding,however, isnot the only problem stemming from unusual mete-orological events. The southern heat wave anddrought of 1998 resulted in damages in excess of $6 billion.

El Niño and the 1998 Florida WildfiresFlorida consistently ranks among the top five statesin terms of wildfire frequency and acreage affected,due largely to frequent thunderstorms and a warmmoist climate that promotes lush growth of volatileunderstory plants. To limit the accumulation offuels that promote disastrous wildfires,landownersin Florida routinely treat close to 2 million acres ayear with prescription fire. The unseasonably warmweather and copious rainfall brought about by ElNiño conditions during the winter of 1997-98 result-ed in even higher plant growth than usual and highsoil moisture that limited the acreage that could betreated effectively with dormant-season prescriptionfire. As El Niño conditions began to subside inMarch,1998, record breaking rainfall changed torecord-setting drought. Many prescription burnswere postponed further because of the increasingprobability of crown fires and root damage to trees.Lightning activity picked up as usual in May, butwith lower than average rainfall. By June 1,droughtindices reached record heights. During the ensuingsix weeks,more than 2,500 fires burned roughly500,000 acres in Florida,destroying valuable timberand damaging roughly 350 homes and businesses.Predicting El Niño conditions holds obvious benefitsfor fire preparedness and prevention. Changes in ElNiño patterns would have both ecological and eco-nomic implications.

Also of concern in the Southeast are the effects thatelevated surface temperatures have on humanhealth as a result of prolonged or persistent periodsof excessive summertime heat events coupled withdroughty conditions. For example,it is known thaturban surface temperatures in cities in the Southeastcan be elevated as much as 5 to 10ºF (approximate-ly 3 to 5ºC) over non-urbanized areas (Lo,et.al.,1997;Quattrochi and Luvall,1999). These elevatedurban surface temperatures are a heat stress tohumans and can significantly contribute to increas-ing both the duration and magnitude of photochem-ical smog,particularly ozone concentrations(Southern Oxidants Study, 1995;Quattrochi,et.al.,1998). Increases in maximum summer temperaturesare of particular concern among lower incomehouseholds that lack sufficient resources to improveinsulation and install and operate air conditioningsystems.

Adaptation OptionsUnderstanding the risks and vulnerability of commu-nities to weather-related hazards (considering hid-den and reported costs and the actual frequencywith which these disasters occur in the Southeast)is important to the quality of adaptation strategies.Across the region,intense precipitation hasincreased over the past 100 years and some modelssuggest that this trend will continue as the atmos-phere warms. Traditional approaches to mitigationsuch as flood proofing,elevated structures,andbuilding codes,are no longer adequate in them-selves,particularly in the coastal zone. Even ifstorms do not increase in frequency or intensity, sea-level rise alone will increase the propensity forstorm surge flooding in virtually all southeasterncoastal areas.

The National Oceanic and AtmosphericAdministration (NOAA) and the Federal EmergencyManagement Agency (FEMA) commissioned a studyon the true costs and mitigation of coastal hazardsin 1996. The report of this study calls for a strategicshift in hazard mitigation and focuses on modelstate programs developed in Florida and othersparts of the country to foster more disaster-resilientcommunities. Recommendations include improve-ments in disaster cost accounting and risk assess-ment,insurance/mitigation policy linkages,integrat-ed approaches to coastal management/develop-ment,and community-based mitigation planning(Heinz Center, 1999). Changes in climate and sea-level rise should be an integral consideration asSoutheast coastal communities develop strategiesfor hazard preparedness and mitigation. Severalstates have implemented permanent “set backs”or“rolling easements”to prevent further developmentin areas that will become more flood-prone as sealevel rises (see Coastal and Marine Resources chap-ter).

Health advisory systems,community-wide heat147


emergency plans,improved weather predictioncapabilities,and other adaptations that would likelyreduce urban heat stress and air pollution-relatedhealth effects are presented in the Northeast andHealth Chapters.

2. Agricultural Crop Yields andEconomic Impacts

Current ConditionsGreat agricultural changes have taken place in theSoutheast over the past 150 years. In 1849,theSouth produced more corn than the Midwest;Southeast acreage in corn was higher than in cot-ton. Cotton production expanded greatly after theCivil War, and by the late 1920s,cotton was moredominant in the South than it was a century before.There was complete mechanization of crop produc-tion in the Southeast after World War II and millionsof sharecroppers moved to the big cities in theNorth. This had an important impact,not only onlabor requirements,but on the whole economicstructure of agriculture. There has also been a shift-ing cropland base in this region. For example, overthe last 50 years soybeans changed from a minorforage crop to an agricultural staple second only tocorn in value of production. As soybeans and ricereplaced corn and cotton, farmers chose soils mostsuitable for the new crops. Drained wetland soils inArkansas were more productive in soybeans thanthe old Piedmont soils abandoned by cotton farm-ers.

In terms of agricultural potential,one of theSoutheast’s most important assets is its potential toexpand the acreage devoted to crops beyond thecurrent level. The land from which new croplandcan be drawn is currently about evenly dividedbetween pasture and forestland. Although theSoutheast could substantially increase acreage devot-ed to agriculture,it fares poorly in terms of native

Table 2: Principal Crops in the Southeast(103 acres)

(source,USDA,Census of Agriculture,1996).

1929 1949 1969 1987 1996

Corn 23,940 20,417 7,896 4,309 5,005

Cotton 23,228 13,031 4,711 3,345 5,931

Peanuts 2,207 2,348 1,046 971 927

Rice 598 1,011 1,194 1,654 2,156

Soybeans 1,321 2,599 13,894 25,645 12,303

148


soil fertility. In addition to having low fertility, mil-lions of acres of soils in the Southeast are moder-ately to severely eroded,the result of decades ofcontinuous corn and cotton production under poorsoil management. However, another of the region’sagricultural assets is its latitude and proximity tothe warm Gulf and Caribbean maritime influence.Overall,the Southeast has a consistent 30- to 90-daylonger growing season than the Corn Belt and theGreat Plains. The Southeast has enormous suppliesof fresh water in the form of rainfall,surface waterflowing through streams and creeks,and groundwa-ter. Water availability gives the Southeast some sub-stantial advantages,including irrigation possibilitiesthat have barely been exploited.

The Southeast’s mild climate and frequent rainfallpredispose the region to an array of agriculturalpest problems more serious than anywhere else in

the nation. Agricultural pests can reduce crop yieldsand raise production costs. Another consequence ofthe region’s pest problems has been relatively highuse of pesticides. Although the Southeast accountsfor only 14% of the nation’s cultivated cropland,itconsumes 43% of the insecticides and 22% of theherbicides used by farmers (USDA,Census ofAgriculture 1994).

Potential Impact of Climate Change on Crop Yieldand Water Use at the Field ScaleTwo families of mechanistic crop simulation modelsCROPGRO (for soybean and peanut) and CERES (forcorn,wheat, rice,and sorghum) in DSSAT V 3.5(Tsuji et al.,1998; Jones et al.,1999) were used tosimulate potential dryland and irrigated crop pro-duction using state- and crop-specific managementpractices throughout the Southeast. The HadleyCentre model was chosen for this analysis by theSoutheast working group because this model wasgridded properly for the region at the time theanalysis was begun,and because the model per-formed best,among those tested,at hindcastingsoutheastern ENSO-related climate anomalies.

Crop yield changes simulated under dryland (non-irrigated) conditions suggest that yields were sensi-tive to the Hadley climate change and CO2 fertiliza-tion and that the response varied by crops and loca-tions (Figure 6). Dryland crop yields generallydecreased along the Gulf Coast by 1 to 10% due towater stress. Furthermore,increased demand forwater by other sectors under higher temperatures islikely to amplify these impacts. Increases in atmos-pheric CO2 concentrations may reduce crop wateruse to some extent,due to increases in leaf stomataresistance,but cannot fully compensate for increas-es in crop water demand due to the higher tempera-tures in climate change scenarios.

In the crop simulation model,dryland corn yieldincreased by 1 to 30% in the Coastal Plain anddecreased up to 10% in Louisiana and large parts ofMississippi,Arkansas,and Kentucky. The shortergrowing cycle reduced yield while increased CO2

and rainfall boosted yield. Due to lower water useefficiency, the model suggests that it could possiblybecome uneconomical to ir rigate corn,promptingfarmers to increase dryland corn production.Sorghum yields increased in the model results by 1to 30% in parts of Alabama,Georgia,South Carolina,and North Carolina where seasonal rainfallincreased by 5 to 15%. Simulated yields from irrigat-ed sorghum were 4 to 7% lower almost everywhere,even where higher yields were predicted under dry-land conditions,largely due to shorter growing sea-sons.

Figure 6. Dryland crop yield changes in 2030 (a) and 2090 (b) with-out adaptation for various climate sensitivity scenarios (source:Auburn University, Global Hydrology and Climate Center; Universityof Florida, Agricultural and Biological Engineering Department).See Color Plate Appendix

Dryland Crop Yield Changes

Simulated soybean and peanut yields increased by 1to 30% mostly within the Coastal Plain and mid-southand more than 30% in parts of North and SouthCarolina. Yields also increased,in parts of Arkansasand upper Mississippi,where dryland corn andsorghum yields decreased. Irrigated yields increased1 to 10% over almost all of the region includingwhere losses had been predicted under dryland con-ditions. The models simulated decreases of up to30% in dryland peanut yields in the lower Delta andalong the Gulf Coast,but when irrigation was addedin the same areas,yield increased by over 30%. If themodel-simulated changes were to occur slowly overnext 25 to 45 years, farmers would be likely to slow-ly increase irrigation as the marginal value of irriga-tion increased. The spatial variation in simulatedyield induced by climate change suggests that manyfarmers in the lower Delta and Gulf Coast may dropdryland production of peanuts while production ofthese crops may expand in other parts of the region.

Simulated winter-wheat yield increased in all regionsexcept in the lower Delta and parts of Arkansas.Simulated irrigated yields increased following a simi-lar trend as the dryland yield. Demand for irrigationincreased by 20 to 50% in the Delta where rainfalldecreased,and evaporation increased due to highertemperatures. Over the same period,irrigated yieldsdeclined. In parts of Arkansas and Louisiana,whereirrigated rice dominates,the model simulated 1 to10% yield losses by the 2030s and 3 to 39% increasesby the 2090s. One of the major threats to rice pro-duction is increasing competition for and increasingcosts of irrigation water.

Sensitivity Analysis of Crop Response to Climatic Change To avoid the narrow range of the climate conditionssimulated by the Hadley model, we conducted a sen-sitivity analysis in 10 agriculturally-dominant south-eastern areas using 25 combinations of anticipated

temperature and rainfall at 445 and 680 ppm CO2

levels superimposed on the current climate. Thesensitivity analysis identified the climate conditionsthat would be particularly damaging to the drylandproduction (see Table 3) and allowed us to considerto what extent yields can be maintained with cur-rent management practices. Results indicate thatyield would likely decrease compared to the currentvalues if temperature exceeds the current valuemore than the corresponding value for change inrainfall amount.

Figure 7 shows that under simulated 2030s CO2levels,+1ºC (1.8ºF) change in temperature increaseddryland production of soybean,peanuts,and wheatunder current rainfall,while yields of corn andsorghum declined. A 2ºC (3.6ºF) increase in temper-ature in 2030 resulted in further yield losses for allcrops simulated. In contrast,the effects of 2ºC(3.6ºF) temperature increase in 2090 with nochange in rainfall suggested a generally positiveeffect on crop yields. However, decreases in rainfallby 20% accompanied by temperature increases of 1to 2ºC (1.8-3.6ºF) almost doubled yield losses for alldryland crops studied. Under conditions of loweror the same growing season rainfall amounts,yieldsof all crops increased more due to increased CO2

levels than due to higher seasonal rainfall. For allcrops and combinations of temperature and CO2

changes,decreases in precipitation resulted in differ-ential decreases in crop yields. Changes in yieldswith 20% lower rainfall were of similar magnitude atall other temperatures and CO2 levels simulated.Furthermore, results showed that crop yields weremuch less sensitive to changes in temperature com-pared to changes in precipitation. An increase ingrowing season rainfall of about 20% almost com-pletely offset the negative ef fect of temperatureincrease. Irrigated yields were simulated assumingcurrent rainfall by varying only the temperatures

149


Table 3: Temperature Tolerance Limits for Various Cropsto the projected rainfall changes in 2030s and 2090s.

445 ppmv CO2 as in 2030s 680 ppmv CO2 as in 2090s Change in Current Rainfall Change in Current Rainfall

-20% 0% 20% -20% 0% 20%

Temperature Tolerance

Soybean +0 +1 +3 +1 +3 +4 Peanuts +0 +2 +3 +2 +3 +4 Corn +0 +0 +1 +2 +3 +3 Sorghum +0 +0 +1 +1 +1 +1 Wheat +0 +1 +2 +3 +3 +3

variety and planting date adaptations either by shift-ing from a decreased yield without managementchanges to an increased yield when managementwas changed. For most dryland crops,adaptationdid not completely eliminate yield loss under a 20%less rainfall condition. For corn,peanuts,sorghum,and wheat,there may be little need to change cur-rently adapted varieties under all combinations oftemperature and rainfall changes,in part due tostrong CO2 fertilization effects on crop yields.

Sub-regional Impacts on Productivity andProfitability The largest threat to crop production in theSoutheast appears to occur where decreases in pre-cipitation coincide with higher temperatures. Thiscombination of climate change would likelyincrease evapotranspiration demands in a climatewith lower water availability, which would increasewater stress and reduce crop growth and yields. Inmany areas of the Southeast,however, projectedtemperature rises of no more than 3 or 4ºF (2ºC orless) together with declining needs for irrigation,should enhance crop production. The Hadleymodel scenario for 2030 indicates improved condi-tions for water availability in the Tennessee Valley,coastal North and South Carolina,and the lowerMississippi Valley. Water supplies are projected tobe much worse in the Mississippi Delta and slightlyworse in Louisiana,Southern Alabama,the Floridapanhandle,and the Coastal Plain of Georgia. By2090 they also become much worse in NorthernMississippi,Southern Alabama,and SouthwesternSouth Carolina.

A farm management model was used to simulatechanges in crop mix, water use,and farm incomeassociated with the climate-induced yield changesdescribed above (Hatch et al.,1999). Of the majorcrop growing areas of the Southeast,the southernMississippi Delta and Gulf Coast areas are more neg-atively impacted,while the northern Atlantic CoastalPlain is more positively impacted. Analyses indicatethat farmers could possibly mitigate most of thenegative effects and possibly benefit from changesin CO2 and moisture that benefit crop growth. Thediscussion that follows is organized around the twoprincipal row crop growing areas of the Southeast,the Mississippi Delta and Atlantic Coastal Plain.

The southern portion of the Mississippi Delta isexpected to endure the severest negative impactswith the northern portion relatively less impacted.In both 2030 and 2090,simulated crop yield, wateruse,and income all are relatively worse off in thesouthern area of the Delta,particularly Louisiana,

from +1 to +5ºC (1.8 to 9ºF). If future rainfall ishigher, then irrigation requirements will likelydecrease,and if it is lower, irrigation needs will like-ly increase. Most irrigated crops were sensitive tosimulated climate changes and CO2 fertilization.

Implications for Field Scale AdaptationFuture climate change may strongly affect agricul-ture in the Southeast. In cropping systems,a widerange of low cost adaptations to climate change mayexist to maintain or even increase crop yields underfuture climates. Farmers may be able to respond tochanges in environmental conditions by choosingthe most favorable crops,cultivars,and croppingsystems. Our findings indicate that changes in plant-ing dates and maturity groups could possiblyincrease yield under the climate change conditionsstudied. In the simulations,all crops benefited from150


Figure 7. Dryland yields changes in 2030(a) and 2090(b) withoutadaption for various climate sensitivity scenarios. (source: AuburnUniversity, Global Hydrology and Climate Center; University ofFlorida, Agricultural and Biological Engineering Department). SeeColor Plate Appendix

Simulated Changes in Dryland Yields for Southeastern Crops based on the Hadley

(HADCM2) Scenario

Mississippi,and Arkansas. This picture contrastsrather sharply with the largely beneficial impacts inmuch of the Coastal Plain,especially the northerntier. Southern Alabama,the Panhandle of Florida,and southwest Georgia,the crop growing areas inproximity to the Gulf coast,are the areas of theCoastal Plain that are negatively impacted. The restof this important crop growing area,that stretchesfrom central Georgia to North Carolina,is expectedto see beneficial impacts from the climate-inducedyield changes and the resultant changes in farmmanagement.

Simulated changes in water use for ir rigation of rowcrops show a distinct north-south pattern. That is,the southern tier of the Southeast is expected toincrease its needs for irrigation water whereas thenorthern tier is expected to decrease its relativeneed for irrigation water. This pattern is somewhatevident in the 2030 simulation and very pro-nounced in the 2090 crop and management simula-tions.

Economic sensitivity to increased temperatures wasalso investigated. Two sensitivity scenarios wereanalyzed to provide an indication of sensitivity toincreased temperature without any changes in pre-cipitation. The sensitivity scenarios were “hot”and“very hot;”the former was an increase of 1ºC in2030 and 2ºC in 2090,and the very hot scenarioincreased temperature by 2ºC in 2030,and 5ºC in2090. These temperature changes were selectedbecause they roughly reflect the temperaturechanges associated with the Hadley and Canadianmodels, respectively, without the simulated changesin precipitation. The “hot”scenario had a slightlymore negative impact than the Hadley scenario inmany areas of the Southeast because the hot sce-nario did have the Hadley’s accompanying increasein moisture. The “very hot”scenario producedrather dramatic negative effects, again because thesewere not mitigated by additional moisture.

The heterogeneous growing conditions of theSoutheast and the great diversity of crops and man-agement systems used in the region make broadgeneralizations about regional climate ef fects onagriculture very difficult. The Southeast is oneregion of the nation that is very likely to experiencechanges in the mix of crops that can be profitablygrown. As a result,the Southeast is a region that willgain from improved information on climate effectsand on improved dissemination of this informationto farm managers. Improvements in understandingclimate and forecasting weather will improve theability of managers to deal effectively with these

and future changes, for example by providing themwith forecasts based on ENSO phase (Legler, et.al.,1999).

Additional Adaptation OptionsExpected changes in productivity and profitabilitywill very likely stimulate adjustments in manage-ment strategies. As yields change,commodity priceswill also change. Producers have several options bywhich they can adapt to changes in yield and priceexpectations. As previously pointed out,they canchange to alternative crops. They can also grow thesame crops,but adjust cultural practices, varyingplanting dates,seeding rates, row spacing,patternsof water usage,crop rotations,and the amounts,tim-ing,and application methods for crop nutrients andpesticides.

Technology can also be expected to respond withnew products and methods to optimize productionunder changing climatic conditions. Plant breederswill very likely respond by developing new varietiesto accommodate climatic changes. Combinations oftechnological advances and adaptive managementpractices could very likely minimize the potentialadverse effects and amplify the potential positiveeffects of climate change on agricultural productivi-ty in the Southeast.

3. Forest Productivity Shifts

Current ConditionsMost of the Southeast’s native forests were convert-ed to farmland by 1920,with a large percentage ofthis conversion occurring prior to the Civil War. By1860, about 43% of the total land area in theSoutheast was reported as farmland,but a substan-tial part of the farm holdings remained in forest,which was often used as a place for grazing live-stock (USFS,1988). With continued expansion ofsettlements,timberland continued to decline untilthe early 1920s. Significant changes in agriculturetook place after 1920 that caused abandonment oflarge areas of crop and pasturelands. These includedthe boll weevil infestation,which made cottongrowing unprofitable in many parts of theSoutheast. Some of the abandoned land was plantedwith trees,but the majority reverted naturally to for-est leading to increases in timberland acreage (USFS1988).

By the late 1950s and early 1960s,the decline intimberland began again in the Southeast,caused pri-marily by the clearing of forest for soybean andother crop production. Much of this timberlandreduction occurred in the bottomland hardwood 151


and their relationships with climate variability andchange remains an integral part of the economicwell being of the Southeast.

Potential Impacts of Climate ChangeAs part of this Assessment,PnET-II,a physiologically-based forest process model,combines climate,soil,and vegetation data to simulate annual soil waterstress,drainage and biological productivity in south-ern US forests (Aber, et.al.,1993;McNulty, et.al.,1994). Using Forest Inventory and Analysis datafrom the USDA Forest Service,predictions of forestproductivity per unit area were projected intoregional growth using current volume for southernpine forests (USDA,1988). Model projections offuture forest productivity included the influence ofdoubled CO2. The Hadley climate scenario (the wet-ter of the principal climate models used in theNational Assessment) was used for reasons cited ear-lier. Total changes in standing volumes were calcu-lated by multiplying growth per unit area by thetotal area of pine forest across the region.

The PnET-II model indicated a 12% increase in over-all southeastern forest productivity by 2100,butthere were important differences between hard-woods and pines. In model simulations using theHadley scenario,southern pine plantations experi-enced an 11% increase in productivity by 2040 andan 8% increase by 2100,while hardwood and mixedpine hardwood forest (which represent 64% of thetotal forest area) experienced a 22% increase in pro-ductivity by 2040 and a 25% increase by 2100 com-pared to 1990 productivity estimates. This differ-ence would likely be accentuated under the warmertemperatures simulated by the Canadian model.This is significant because pines (used for pulp andpaper),presently account for almost two-thirds ofthe region’s forest industry land and about half ofthe nation’s softwood inventory. Climate modelsused as input in both the forest and VEMAP modelssuggest a northward shift in forest productivity overthe next century, but they do not consider changesin management that could potentially ameliorateadverse effects. At least two ecosystem models runwith the Canadian climate scenario suggest thatthere will be a 25 to 50% increase in fires and thatpart of the southeastern pine forest will be replacedby pine savannas and grasslands due to increasedmoisture stress (see Vegetation and BiogeochemicalScenarios Chapter).

Figure 8 shows PnET-II model outputs for southernforest net primary productivity in the Southeast atpresent (baseline),and at 2040 and 2090 decadalaverages under the Hadley model climate scenario.

forest areas of the Mississippi Delta. Forest reduc-tions were further fueled by growth in urban areas,highways,power lines,and related development.Throughout the 1970s,timberland was cleared foragricultural use and for an expanding export mar-ket. In the decade 1982-92,the National ResourcesInventory reports roughly a half million-acre loss(less than 1%) in forestland in the Southeast.

Land use changes in the region are sensitive to anyprojected changes in the value of agricultural andforest lands. Expansion of urban and built-up areasin the Southeast also represents a significantdemand for land,but one that will continue to besmall relative to the total land base. For example,although developed land increased about 27% in thedecade 1982-92,the total land use in this categoryrepresents only 8% of the total land use in theSoutheast. Future land use changes are likely tohave major impacts on things which do not havemarket prices:wildlife and habitat,topsoil,aesthet-ics,pollution of groundwater by agricultural chemi-cals,soil erosion,sedimentation,and loss of wet-lands. The management of these natural resources

152


Hadley Model Southern Har dwoods 2000 Hadley Model Southern Har dwoods 2100

(10 year average)

300 - 800800 - 12001200 - 15001500 - 18001800 - 21002100 - 2500

grams/meter square/year300 - 800800 - 12001200 - 15001500 - 18001800 - 21002100 - 2500

grams/meter square/year

Figure 8. Potential net primary productivity (NPP) of loblolly pineand southern hardwoods simulated by the PnET model with theHadley climate scenario (HADCM2). (Source: USDA Forest Service,Southern Global Change Program). See Color Plate Appendix

Hadley Model Southern Pines 2000 Hadley Model Southern Pines 2100

(10 year average)

Potential Southern Pines and Hardwoods NetPrimary Productivity (NPP)

300 - 800800 - 12001200 - 15001500 - 18001800 - 21002100 - 2500

grams/meter square/year300 - 800800 - 12001200 - 15001500 - 18001800 - 21002100 - 2500

grams/meter square/year

The principal factor influencing the lower increasein pine relative to hardwood productivity by 2090 isthe fact that pines have greater water demands thando hardwoods on a year-round basis. Even with theincreased water use efficiencies associated withincreased atmospheric CO2, the southern pines arelimited by water as evapotranspiration rates increasewith air temperature.

The impact of climate change on the distribution andimpact of forest pests and diseases remains uncer-tain. Southern pine beetles caused over $900 millionin damage to southern US pine forests between 1960and 1990. Higher winter air temperatures willincrease over-wintering beetle larva survival rate,andhigher annual air temperatures will allow the beetlesto produce more generations per year. Both of thesefactors could increase beetle populations. On theone hand, field research has demonstrated that mod-erate drought stress can increase pine resin produc-tion and therefore reduce the colonization successrate of the beetle. However, severe drought stressreduces resin production and greatly increases thesusceptibility of trees to beetle infestation.Insufficient evidence currently exists to predictwhich of these factors will control future beetle pop-ulations and impacts (McNulty, et.al.,1998).

Potential Effects on Timber MarketsThe Sub-Regional Timber Supply (SRTS) model (Abt,et.al.,2000) was developed to link forest inventorymodels with timber market models. The model usesestimated relationships between prices,harvest,andinventory to model market impacts of shifts in sup-ply or demand. The SRTS model uses the spatiallyexplicit and species specific growth changes fromPnET-II to modify inventory accumulation. Thecumulative nature of inventory tends to dampen themarket impacts of the variability found in annualgrowth rates.

This analysis of the future of the forest sector comesat an important turning point in historical trends.Since the turn of the century, southern inventorieshave been increasing due to recovery from exploita-tion in the 1920s and the emergence of industrialforestry in the 1950s. During the last decade,removals of both hardwoods and softwoods haveincreased rapidly and are approximately equal togrowth. This implies that even subtle climate changeimpacts may influence the direction of future inven-tory changes. Overall,the SRTS model (using theHadley climate scenario) indicated that climatechange would more likely favor the Mid-Atlantic overthe East Gulf, and hardwoods over softwoods,andthat growth over the 2000-2020 period would be sig-

nificantly lower than over the 2020-2050 time peri-od. This,along with currently favorable growth/removal ratios in the Mid-Atlantic region,led toshifts northward in pine and hardwood harvest inall model runs.

Beyond the spatial and market adjustments to cli-mate change within the forest sector, land-use feed-backs from the agricultural sector, discussed below,also tend to move inventory and harvest to areaswith comparative advantages. Sensitivity analysis tohigher temperatures (Hadley +2ºC,or 3.6ºF) indicat-ed that the northern shift in inventories and marketsbecame more pronounced and regional pricesincreased as the mid-Gulf region experienced signifi-cant growth declines.

Potential Effects on Forestland AreaAlthough forests and agriculture dominate theSoutheastern landscape,the effect of a changing cli-mate on the relative productivity of these activitiesis just one of many factors that will determine howthe region’s land will be used in the 21st century.Urban and other developed uses,while currently arelatively small part of the regional land base,haveexpanded substantially in the last two decades andare likely to continue to do so in the future. Inrecent years, much of the forest area lost to develop-ment in the region was about equally offset by gainsfrom forest establishment on previously agriculturalland due to the decline in agricultural returns. Thishas tended to stabilize net forest area trends whileexacerbating losses in agricultural land.

Without accounting for climate change, forest area isprojected to remain fairly stable in aggregate to2040. But within the region,there are expected tobe areas with substantial land use change (Figure 9).Urbanized areas in the North Carolina and Georgiapiedmont and southern Florida are projected tocontinue the conversion of forestland and agricul-tural land to developed use,but on a regional basis,these losses are expected to be offset by movementof land from agriculture to forest in other areas,such as the Mississippi delta.

Relative changes in forest and agricultural returnsbrought on by climate change could possiblychange the pattern of stable forest areas in thefuture if, as some scenarios suggest, agriculture canadapt to climate change in some parts of the regionbetter than forests can. Under the Hadley base cli-mate scenario,our model simulations suggest rela-tively little change in the way that land is allocatedbetween forests and other uses between now and2040,though some northern migration of forest area

153


Adaptation OptionsIn general,the biological productivity of southeast-ern forests will likely be enhanced by atmosphericcarbon enrichment,as long as precipitation doesnot decline or air temperature increase soil moisturestress to a level that would offset potential CO2 ben-efits on productivity. The modeling systememployed to analyze regional impacts of climatechange captures some adaptation through its model-ing of economic behavioral responses to changes inthe biophysical conditions in forestry and agricul-ture. For instance,the northward shift in forest pro-ductivity is projected to lead to relative increases inthe proportion of regional timber harvests thatcome from the northern reaches of the region. Thiswill tend to compensate for the biophysical effectsof climate by reducing harvest pressures in themore negatively affected southern parts and increas-ing pressure in the more positively af fected north-ern parts. In addition,landowners are projected toshift land between forests and agriculture in placesand at times where the change in relative productiv-ity warrants it.

Other potential adaptation strategies not modeled inthis Assessment include genetic and silvicultural sys-tem improvements that increase water use efficien-cy or water availability. The knowledge of the roleof fire,hurricanes,droughts,and other natural distur-bances will be important in developing forest man-agement regimes and increasing stand productivityin ways that are sustainable over the long term.Under a hotter, drier climate,an aggressive fire man-agement strategy could prove to be very importantin this region.

Timber productivity associated with increased tem-perature, growing season length,and carbon enrich-ment may be further enhanced by improved geneticselection,bio-engineering,use of marginal agricul-tural land for tree production,and more intensiveforest management. Reduction of air pollutants(e.g.,ozone,nitrogen oxides) could also be animportant strategy for increasing forest productivity.

4. Water Quality Stresses

Current ConditionsThe Southeast has abundant surface waterresources,most of which are intensively managed.Almost all major river systems in the region havebeen dammed and there are few minor streams thathave not been af fected by landscape alteration,channelization,surface or ground water with-drawals,or other human activities. Based on 50-yearor longer streamflow records at 395 stations,Lins154


Timberland Acreage Shift

Figure 9 (a). Changes in land use based on timberland acreageshift for 1993-2040: baseline without climate change. Forestlandlosses are projected in the more urbanized areas of the Southeast,from northern Virginia through the Georgia piedmont and southernFlorida. The movement of land from agriculture to forest is project-ed in many parts of the mid-South. See Color Plate Appendix

Figure 9(b). Timberland acreage shifts by 2040 due to Hadley cli-mate change. In 2040, forestland is projected to be slightly higherwith Hadley base climate change than without climate change insome of the northern reaches of the Southeast, but slightly lowerunder climate change in parts of the deep South. Year 2040 landallocation effects in most of the region are fairly neutral. (Source:North Carolina State University, Department of Forestry; ResearchTriangle Institute, Center for Economics Research). See Color PlateAppendix

is expected (Figure 9). However, sensitivity analysisreveals that substantial variation from the Hadleybase scenario (e.g.,+2 or 4ºC,3.6ºF or 7.2ºF) wouldlikely have more dramatic effects on land allocationand lead to larger net losses in forest area. Theseprojections should be evaluated with caution,how-ever, as the land use model has employed more lim-ited information on the sensitivity of agriculturaleconomic returns to climate change than on on eco-nomic returns to forestry.

5% - 25% Decline<5% Change5% - 25% Increase

>25% Decline5% - 25% Decline<5% Change5% - 25% Increase

and Slack (1999) found that the conterminous US isgetting wetter. Exceptions were noted in parts ofthe Southeast and the Pacific Northwest,wheresome stream gauges showed a decrease in minimumdaily discharge. Parts of Florida and Georgia appearto be experiencing a trend towards decreased mini-mum flows,while the Lower Mississippi River Valleystations showed an increase in both annual mediandaily and annual minimum daily discharge (Lins andSlack,1999). In Louisiana,when Keim and others(1995) modeled historical streamflow based on pre-cipitaion,streamflow per unit drainage areaincreased significantly since 1900.

Potential Impacts of Climate Change Changes in climate that result in decreased runoffduring early summer generally reduce water qualityin the Southeast (Mulholland,et.al.,1997). Summerlow flows occur when water quality (particularlydissolved oxygen) of many southeastern streamsand rivers is at its lowest (Meyer, 1992). Reduceddissolved oxygen during summer months can resultin massive fish kills and harmful algal blooms inboth coastal and inland waters.

An assessment of southeastern water quality associ-ated with changes in climate was conducted usingEPA’s GIS-based BASINS model to evaluate currentand future water quality conditions under bothmean and extreme hydrologic conditions in theSoutheast. Analyses were conducted for each of theUS Geological Survey’s eight-digit Hydrologic UnitCodes (HUCs). Water quality indices included dis-solved oxygen,nitrogen and nitrates,and pH. Theassessment included three steps:identification ofbasins with current and potential water qualityproblems,prediction of general change in streamflow conditions under scenarios of future climate,and re-evaluation of affected basins using the Hadleyclimate model scenarios for 2030 and 2100.

While water quality problems across the Southeastare not critical under current conditions,qualityattainment status is not met in several cases duringthe majority of the year, and can become criticalunder extreme low flow conditions during someportions of the year. Stresses on the water quality ofthe Southeast appear to be associated with intensiveagricultural practices,urban development,coastalprocesses,and possibly mining activities. As mightbe expected,the impacts of these stresses appear tobe more frequent during extreme conditions,proba-bly associated with dry weather. Analysis of the cur-rent status (based on 1990-97 observations) of thewatersheds in the Southeast for dissolved oxygenrevealed few problems under average conditions.

However, it must be recognized that because theBASINS database is indexed to major USGS hydro-logic units,it necessarily consists of observations ofconditions on the larger streams in the regions.Thus,smaller tributaries may exhibit water qualitydegradation that was not apparent at the larger scaleof this analysis.

The analysis suggested that only scattered HUCwatersheds in a few states currently exhibit dis-solved oxygen (DO) conditions below, or nearlybelow the recommended 5 mg/l during averageconditions. However, dissolved oxygen problemsunder extreme low flows do arise at a few locationsin most,if not all,states. Nitrate levels in streams inthe Southeast are used as an indication of nutrientcontent in these HUCs. While some nutrients areessential for ecosystem health, excessive levels canresult in harmful conditions such as algal blooms,which can negatively impact DO levels. The currentstatus (1990-97) based on observations for totalaverage nitrate nitrogen content for streams of theSoutheast reveals that many exhibit levels above 0.5mg/l and in some cases above 4 to 5 mg/l,which is3- to 4-times higher than levels common in mostsoutheastern streams.

One interesting observation is that streams that cur-rently exhibit low dissolved oxygen levels do notcorrespond to those basins where nutrient levelsappear to be high. However, in both cases,theproblems are most prevalent in watersheds withintensive forestry or agricultural operations.

Climate scenarios for the southeastern US providecontrasting results in terms of temperature and pre-cipitation estimates over the region,so that in somecases conditions may improve while in others theymay degrade. The Canadian model results show lit-tle overall change until 2030, followed by drierweather in most of the region over the next seventyyears. On the other hand,the Hadley Centre modelpredicts a slight decrease in overall precipitationover the region during the next 30 years,afterwhich precipitation increases significantly. TheHadley model results also show significantlydecreased precipitation during the first six monthsof the year with rainfall returning to normal,or nearnormal, for the last six months,particularly by theend of the centur y. These results are particularlystriking for the immediate Gulf Coast region andindicate that this area may be exceptionally vulnera-ble to degraded conditions. Intensive agriculturalactivity including disking and planting in the earlyspring, fertilizer application in the late spring/earlysummer, and harvesting in the fall may significantly

155


5. Threats to Coastal Areas

Current ConditionsFew regions have the combination of special charac -teristics and vulnerabilities found in southeasterncoastal areas. The interaction of sea-level rise,storms,beach erosion,subsidence,salt water intru-sion,urban development,and human populationgrowth,and shifts in the transition zone where landmeets ocean,creates conditions for potential adverseeffects on the largest segments of the southeasternpopulation. Large cities located in the coastal zone(such as Houston,Charleston,and New Orleans)already suffer frequent and severe flood damages.

Potential Impacts of Climate ChangeSea-level rise is regarded as one of the more certainconsequences of increased global temperature,andsea level has been rising gradually over the past15,000 years. Globally, average sea level rose 4 to 8inches (10 to 20 cm) during the past 100 years andthis average rate is projected to accelerate 2 to 5-foldover the next 100 years (IPCC,1998). Parts of theCity of New Orleans that are presently 7 feet belowmean sea level may be 10 or more feet below sealevel by 2100,due to a combination of rising sealevel and subsidence of the land surface.

Low-lying marshes and barrier islands of the south-eastern coastal margin are considered particularlyvulnerable to sea-level rise,but all are not equallyvulnerable. Cahoon and others (1998) found thatsome Gulf coastal marshes and one mangrove site insouth Florida are being gradually submerged becausethey do not accumulate sediment quickly enough tokeep up with present rates of sea-level rise. Incoastal Louisiana,landforms created by MississippiRiver sediment deposition over the past 8,000 yearsare naturally de-watering and compacting. As sealevel rises,inundation or displacement of coastalwetlands and barrier islands is occurring. Theimpacts of subsidence (the lowering of the land rela-tive to sea level) are aggravated by human activitiessuch as levee construction along the MississippiRiver, ground water withdrawals,and canal dredgingthough marshes,passes,and barrier islands. Changesin tidal amplitude have caused salt-water intrusioninto many formerly fresh and brackish water habi-tats. Roughly one million acres of south Louisianawetlands have been converted to open water since1940,and Louisiana’s barrier islands have eroded totwo-thirds of the size they were in 1900.

exacerbate water quality conditions during this period.

Preliminary hydrologic analyses based on the Hadleyscenario suggest that streamflow in the Southeast(particularly along the Gulf Coast) may decline by asmuch as 10% during the early summer months overthe next 30 years (Cruise,et.al.,1999;Ritschard,et.al.,1999). These results lead to the conclusions thatwater quality conditions may become critical duringmore frequent periods of extreme low flow.Correlation of the hydrologic analyses with the landuse in basins where water quality problems alreadyexist suggests that the problems may be most acutein areas of intensive agricultural activity, in coastalareas,or near coastal streams (Cruise et al.,1999).

Many of the basins with high nitrate levels form theboundary between two states. The Chattahoocheeboundary between Georgia and Alabama and theTombigbee boundary between portions of Alabamaand Mississippi are two outstanding examples. TheHadley scenario suggests decreased water availabilitythroughout much of this region over the next 50years. As streamflow and soil moisture decrease,intensity of fertilizer application may increase andirrigation needs may become critical. These issueswould likely lead to intense competition for scarcewater resources and conflicts between these statesover runoff treatment and water quality.

Water quality is also a concern in nearshore marineenvironments. Both the Canadian and Hadley cli-mate models suggest an increase in rainfall in theUpper Mississippi Valley (see Great Plains andMidwest chapters). A large (8,000 to 18,000 km2

during 1985-1997) zone of oxygen-depleted (hypox-ic) coastal waters is found in the north-central Gulfof Mexico and is influenced in its timing,duration,and extent by Mississippi River discharge and nutri-ent flux (Rabalais,et.al.,1999; Justic et al.,1997).Nitrate delivered from the Mississippi Basin to theGulf of Mexico,principally from non-point agricul-tural sources,is now about three times larger than itwas 30 years ago as a result of increases in nutrientloading per unit discharge (Goolsby, et.al.,1999).Hypoxia,which is most prevalent in the lower watercolumn,can adversely affect marine life and is agrowing concern to those who harvest and manageGulf fisheries. An increase in Upper MississippiBasin streamflow, where the majority of the nitrogenand phosphorus loading occurs,portends anincrease in the hypoxic zone offshore.

156


If sea-level rise accelerates during the 21st centuryas predicted by the Intergovernmental Panel onClimate Change (IPCC),many other southeasterncoastal areas will experience shoreline retreat andcoastal land loss. Under the IPCC’s “best estimate”of average global sea-level rise over the next 100years,the Big Bend area of the Florida Gulf coastwill likely undergo extensive losses of salt marshand coastal forest (Doyle,1998, Figure 11). Since1980,losses of coastal forests in parts of Florida,South Carolina,and Louisiana have been attributedto salt water intrusion and/or subsidence. Since1991,landowners and public land managers inFlorida have observed extensive die-offs of Sabalpalm along a 40-mile stretch of coast between CedarKey and Homosassa Springs. Williams and others(1999) attribute the forest decline to salt waterintrusion associated with sea-level rise. Since 1852,when the first topographic charts were prepared ofthis region,high tidal flood elevations haveincreased approximately 12 inches.

Rising sea levels due to climate warming may alsoaffect estuaries and aquatic plant communities. Sea-level rise reduces the amount of light reaching sea-grass beds (light penetration decreases exponential -ly with water depth),thereby reducing growthrates. Some marine grass beds may be eliminatedbecause their shoreward migration is impeded byshoreline construction and armoring (Short andNeckles,1999). Increased tidal range associatedwith sea-level rise may have deleterious effects onestuarine and fresh water submerged aquatic plantsby altering both salinity and water depth. (SeeCompounding Stresses on Major Estuaries and Baysin the Northeast chapter for a discussion of sea-levelrise effects on salinity in the Chesapeake Bay.)

Coastal Wetland Vulnerability Thresholds at which sea-level rise results in coastalwetland loss vary among sites due to differences inrates of vertical accretion (sediment build up) andlocal subsidence or uplift processes. Cahoon andothers (1998) estimated the potential for submer-gence of 10 southeastern wetlands by simultaneous-ly measuring surficial sediment accretion and soilsurface elevation changes,and then comparingthese rates to observed sea-level change from tidegauges. Three of the 10 sites are experiencing a netelevation deficit relative to sea-level rise. The othersites are presently accumulating enough sediment tokeep pace with sea-level rise. If sea-level rise wereto accelerate 4-fold,the Oyster Bayou site would besubmerged by about the year 2045. The OysterBayou site would not be submerged if sea-level riseincreases 3-fold or less,unless the site is impactedby a hurricane or other disturbance. Both long-termprocesses (e.g.,accretion,compaction,and decom-position) and episodic events (e.g.,hurricanes)affect the threshold at which coastal wetlands aresubmerged by sea-level rise.(See Figure 10)

157


Figure 10. Relationship of vertical accretion and marsh surface elevation change withlocal relative sea level-rise for sites located in a) low salt marsh, b) high salt marsh,c) mangrove forest and d) brackish marsh. The diagonal line indicates parity betweenaccretion or elevation change and sea-level rise. (source: Cahoon et al., 1998) SeeColor Plate Appendix

Sea Level Rise and Marsh Changes

number of deaths attributed to hurricanes declineddramatically since the 1950s (Pielke and Pielke,1997).

Adaptation OptionsThere are few practical options for protectingcoastal communities and ecosystems as a wholefrom increased temperature, changes in precipita-tion,or rapidly rising sea level. Still,a variety ofmanagement measures could be applied on a site-by-site basis to increase the resiliency of specificcommunities and ecosystems or to reduce or partial-ly compensate for impacts. Many of these measurescould be justified based solely upon non-climatethreats to coastal regions. For example,increasedprotection for existing coastal wetlands and removalof other stresses (such as dredge-and-fill activitiesand water pollution) may not only reduce the sensi-tivity of coastal communities, wetlands,and barrierislands to small changes in average sea level but alsoachieve broader conservation goals (Burkett andKusler, 2000).

Other no-risk measures for achieving broader objec-tives and reducing climate change impacts include:limiting construction in areas where coastal wet-lands may be displaced as sea level rises;installingsediment diversions for dams;linking presently frag-mented wetlands and waters to provide the corri-dors needed for plant and animal migration;usingwater control structures for some wetlands toenhance particular functions and address decreasedprecipitation and/or increased evaporation;increas-ing management programs for invasive species con-trol;and implementing various coastal restorationmeasures (Burkett and Kusler, 2000).

ADDITIONAL ISSUESSix additional climate-related issues for theSoutheast region are:

• Climate Model Limitations: Existing general cir-culation models cannot adequately resolve somecomponents of climate or certain geographic ortopographic features that are important becauseof their interaction with regional climate fea-tures. For coastal regions, much uncertaintyexists about the effects of global climate changeand variability on tropical storms,the mostimportant natural hazard affecting regional vul-nerability. Effects of climate change on areas ofhurricane origination and threshold for hurri-cane formation,intensity, frequency, and tracksare poorly understood.

Storm surge is intensified as sea level rises and natu-ral coastal defenses deteriorate. It is important tonote that even if there is no significant climate-change-driven increase in the frequency or intensityof Atlantic hurricanes,these storms will be moredamaging when making landfall on coastal regionsas sea level rises and coastal landforms erode. SouthAtlantic and Gulf coastal populations increased 15%between 1980 and 1993. During this period,thevalue of insured coastal property in the USincreased by 179%(Insurance Institute for PropertyLoss Reduction and Insurance Research Council,1995). Florida topped the list of coastal states withpotential hurricane damages,with $872 billion ininsured properties. Insurance Research Councildemographic projections suggest that the number ofpersons living in the most hurricane prone countieswill have increased to 73 million by 2010,a dou-bling since 1995 (Insurance Institute for PropertyLoss Reduction and Insurance Research Council,1995).

During the past 30 years,the Gulf of Mexico hasseen a decrease in the number of hurricanes makinglandfall. Hurricanes Andrew, Hugo,and Brett are theonly Category 4 storms to make landfall since 1969,but since 1900,35% of all hurricanes hit Florida andalong the middle Gulf Coast,and 50% or more of allhurricanes making landfall are Category 3 or higher(Heinz Center, 1999). Property losses due to hurri-canes increased from less than $5 billion per decadebetween 1900 and 1940 to about $15 billion perdecade during the 1960s to 1980s;however, the158


Figure 11. Changes in Florida’s Big Bend region forest, marshes, and open water underIPCC (1998) sea-level rise scenarios. (Source: Doyle, 1998). See Color Plate Appendix

Changes in Florida’s Big Bend

38 in (95 cm) - High Estimate

20 in (50 cm) - Mid Estimate

6 in (15 cm) - Low Estimate

159


issues,traffic patterns,and evacuation/shelter

infrastructure are all important areas for potential

mitigation and future research. The potential

impacts of climate change on oil,natural gas,and

navigation infrastructure is also of concern in the

region.

CRUCIAL UNKNOWNS ANDRESEARCH NEEDS

Precipitation Uncertainties

The Southeast is the only region for which currentclimate models simulate large and opposing changesin precipitation patterns over the next 100 years.The range of differences is so great that it is difficultto state with any degree of confidence that precipi-tation will increase or decrease in the Southeastover the next 30-100 years as atmospheric CO2increases. Until climate models are improved (oruntil there is a way to validate and compare theaccuracy of the existing models),people in thisregion must consider a wide range of potentialfuture changes in soil moisture and runoff, as wehave in this Assessment.

Human Health

There are serious human health concerns related tothe plausible increases in maximum temperature forthis region,particularly among lower income house-holds. Similarly, the flood prone nature of coastalcounties from the Carolinas to Texas could have sig-nificant human health implications in addition to theeconomic losses discussed earlier. Approaches formodeling human health effects should include theseaspects of the population in addition to the climaticscience. Water quality degradation could possiblybecome a more serious human health problem in theregion;improvements in our capability to modelstreamflow and water quality in both inland andcoastal waters are needed.

Agriculture and Forestry

Our current understanding of the potential conse-quences of climate change on agriculture in theSoutheast is focused on a few key row crops. In thefuture,it will be important to include the affects ofclimate variations on other high value crops (e.g.,cit-rus and vegetables) and on animal management prac-tices. Furthermore,the role of climate on pests andpest management systems needs to be included infuture assessments. It will also be necessary to

• Water Resources: Fresh water plays an importantrole in many sectors including coastal resources,health, agriculture,estuarine fisheries,andforestry. Competing demands from urban devel-opment, agriculture,and recreation for alreadystressed ground water systems would likely beexacerbated by changes in precipitation and saltwater intrusion due to sea-level rise.

• Impacts on Coastal Ecosystems and Services:Sea-level rise, changes in fresh water delivery tocoastal estuaries,and increased atmospheric tem-perature and CO2 all portend changes in thestructure and function of coastal and estuarinesystems. Losses of coastal marshes and sub-merged aquatic vegetation will have impacts athigher trophic levels. Gulf coastal states current-ly produce most of the Nation’s shrimp, oysters,and crabs,and each of these estuarine fisheries isdependent upon the primary productivity ofcoastal ecosystems.

• Health Issues related to Water Quality: Theeffects on surface waters of changes in precipita-tion have important health implications in theregion. Increased precipitation promotes thetransportation of bacteria as well as otherpathogens and contaminants by surface watersthroughout the region. Health consequencesmay range from shellfish infections transmittedto humans to ground water contamination asso-ciated with saltwater intrusion.

• Socioeconomic and Insurance Issues: In the ulti-mate analysis,the issue of climate change and theneed for an assessment of potential conse-quences will be relevant to the degree that it isplaced on a human scale. To this end,the poten-tial societal impacts of climate change must beidentified and understood. Insurance exposureand/or the insurability of coastal and island facili-ties are issues that should be examined in thecontext of climate change and variability.

• Urban Issues: A distinctive characteristic of

southeastern coastal regions is their current high

level of urbanization and rate of population

growth,which could possibly be affected by

changes in climate that are presented in this

Assessment. The urban environment will have its

own responses to the impacts of climate change

and variability. While responses will be driven by

stakeholder and policymaker decisions,they

need to be evaluated and understood within the

framework of different potential regional scenar-

ios of climate change consequences. Design and

construction factors,building code issues,infra-

structure and lifelines,energy use,structural vul-

nerability to natural hazards,land use and zoning

Aquatic and Coastal Resources

If precipitation patterns continue to change on ascale similar to that observed over the past 100years,many southeastern aquatic ecosystems,includ-ing estuaries,will be affected by changes in stream-flow. There are several additional unknown variablescorresponding to future conditions that might affectthe quality of southeastern water resources. They fallinto two categories:future pollutant loadings (naturaland anthropogenic) and biophysical reactions. Thepollutant loadings,both point and non-point,will bedirectly related to changes in land use activity includ-ing the presence of confined animal systems andgrowing population centers. Also,atmospheric depo-sition of nitrogen will be tied to continued emissionsof nitrogen oxides (NOx). The biophysical reactionson the land surface that might serve to uptake nitro-gen and other constituents will be associated withland cover conversion and vegetation. Futureresearch programs that include these three criticalunknowns seem crucial to gaining a clearer under-standing of the relationship between variations in cli-mate and water quality in the Southeast.

Quantitative data describing the response of nativesoutheastern plant communities to atmospheric car-bon enrichment, water quality (e.g.,salinity),andchanges in temperature and soil moisture are limitedto a few key species. Moreover, very few studieshave addressed the potential interactions among cli-mate variables and between plant species,and evenfewer studies deal with climate effects at secondaryand higher levels in the food web. Several recentstudies suggest that a number of invasive species willbe favored by climatic change in the Southeast,suchas the freeze-intolerant Chinese tallow, which is nowa serious invader in the Gulf coastal plain. Modelsthat integrate environmental change,species respons-es,and interactions among species are needed todescribe pathways that are likely to alter plant andanimal community structure. Research is also need-ed to determine secondary and higher-order effectson ecosystem goods and services. Carbon sequestra-tion is one ecological function that has been poorlydescribed in this region of abundant wetlands andforests that play potentially significant roles as car-bon sinks.

Research and demonstration projects are needed toidentify and prioritize methods that may be imple-mented to minimize the adverse ecological effects ofclimatic change on native southeastern flora andfauna. Monitoring is needed to evaluate long-termtrends in the abundance and distribution of native

develop, validate,and evaluate new technologycapabilities such as new genetic varieties that willhelp farmers cope with any changes in future cli-mate. The effects of biotechnology (e.g.,transgeniccrops) on future agricultural productivity in thisregion,including the benefits of using less fertilizer,pesticides,or water, need to be evaluated in light ofplausible climate scenarios. Because the incorpora-tion of new climate-related technological capabili-ties into agriculture is relatively new and yetunproven,future pilot studies should explore thecommunication of such information to the agricul-ture community.

The potential effects of climate change on agricul-tural prices are an equally complex interaction ofphysical effects and managerial responses world-wide. Spatial equilibrium economic models thatwould address these market issues require that theinformation from all regions be reasonably similar.In the case of climate change,detailed informationfrom one region set against very general informationfrom many important competing crop growingareas would not provide a consistent framework forunderstanding worldwide response in the agricul-ture sector. Thus,it would be very useful to investi-gate in greater detail climate-induced productioneffects in major international crop areas to integratesuch farm management results from importantgrowing areas worldwide to address potential cli-mate-induced price effects.

Although extensive laboratory and field research hasbeen completed on the individual impacts of chang-ing air temperature,precipitation,ozone,carbondioxide,and nitrogen availability on forest produc-tivity, water use,and carbon sequestration,there isstill little understanding of the synergistic impacts ofenvironmental change on southern forests. Fieldexperiments with multiple treatment factors (e.g.,variables) are quite costly, and there are scalingproblems associated with laboratory experiments.Therefore,improved development,testing,and vali-dation of integrated stress impacts through comput-er modeling are crucial future research needs.Models can provide a mechanism for examiningchanging atmospheric and socioeconomic impactson forest structure and function. However, beforeany confidence can be given to such model projec-tions,priority needs to be given to testing,modelverification,and analysis.

160


and non-native species, focusing on species andgroups that are considered highly sensitive to therange of predicted climatic changes in the region(e.g.,amphibians in pine flatwood ponds,benthicinvertebrates in coastal estuaries,and salt-tolerantinvasive aquatic plants that could out-competenative plant species that are important wildlife foodsources).

Extreme Events/DisturbancePatterns

Changes in disturbance patterns (e.g.,hurricanes,floods,droughts) are possibly more significant interms of potential economic losses than longer-termchanges in precipitation or temperature.Ecosystems are also impacted by climate-related dis-turbance. Disturbance is a natural process that,inmany cases,not only structures ecosystems but sus-tains them as well. Our limited understanding ofthe role of disturbance in natural ecosystems andour inability to predict climate extremes is problem-atic for those interested in mitigating the potentialadverse impacts of climate change. Research shouldbe undertaken to examine potential changes in dis-turbance regimes that may be expected as the cli-mate warms and precipitation patterns change.Disturbance topics should not be limited to weatherevents,however. Fire,harmful algal blooms,andinsect outbreaks are ecological disturbances thatmay be heavily influenced by climatic conditions.The ecological effects of these types of disturbancesare difficult to model or predict because they areoften poorly understood. Basic information is need-ed to identify ecological changes that are likely tooccur as the type,frequency, and spatial patterns ofdisturbance are altered as the climate changes.

LITERATURE CITEDAber, J. D.,C.Driscoll,C.A. Federer, R.Lathrop, G. Lovett,J. M.Melillo, P. Steudler, , and J.Vogelmann, A strategyfor the regional analysis of the effects of physical andchemical climate change on biogeochemical cycles innortheastern (U.S.) forests, Ecological Modeling, 67, 37-47,1993.

Abt,R.C., F.W. Cubbage,and G. Pacheco, Forestresource assessment using the subregional timber sup-ply model (SRTS), Forest Products Journal, 50(4),25-33,2000.

Acock,B.,V. R.Reddy, H. F. Hodges, D. N. Baker, and J. M.McKinion,Photosynthetic response of soybeancanopies to full-season carbon enrichment, AgronomyJournal, 77, 942-47,1985.

Allen,L.H.,K. J. Boote, J.W. Jones,A. J. Rowland-Bamford,G. Bowes, D.A.Graetz,K.R.Reddy,Temperature andCO2 effects on rice:1988. US Department ofAgriculture,Carbon Dioxide Research Division,Washington,DC,1989.

Boyer,W. D.,Regeneration of natural longleaf pine for-est, Journal of Forestry, 77, 572-575,1979.

Burkett,V.,and J. Kusler, Climate change: Potentialimpacts and interactions in wetlands of the UnitedStates, Journal of American Water ResourcesAssociation, 36(2),313-320,2000.

Cahoon, D. R., J.W. Day, D. J. Reed,and R.S.Young,Global climate change and sea-level rise: estimating thepotential for submergence of coastal wetlands,inVulnerability of Coastal Wetlands in the SoutheasternUnited States: Climate Change Research Results , USGeological Survey, Biological science reportUSGS/BRD/BSR 1998-0002, pp.21-34,1998.

Cruise, J. F.,A.S.Limaye,and N.Al-Abed,Assessment ofthe impacts of climate change on water quality in thesoutheastern United States, Journal of American WaterResources Association, 35(6),1539-1550,1999.

Currie, D. J.,Energy and large-scale patterns of animal -and plant-species richness, American Naturalist,137(1),27-49,1991.

Dahl,T. E.,Wetland losses in the United States:1780’s to1980’s,US Department of Interior, Fish and WildlifeService,Washington DC, 21 pp.,1990.

161


Insurance Institute for Property Loss Reduction andInsurance Research Council, Community exposure andcommunity protection: Hurricane Andrew’s legacy(IIPLR and IRC).Wheaton,IL (IILPR) and Boston,MA(IRC),1995.

Intergovernmental Panel on Climate Change, TheRegional Impacts of Climate Change: An Assessmentof Vulnerability, Cambridge University Press,UnitedKingdom,514 pp.,1998.

Jones, J.W.,S.S. Jagtap,and K. J. Boote,Climate change:implications of soybean yield and management in theUSA,paper presented at the World Soybean ResearchConference VI,Chicago,Illinois,August 7,1999.

Justic´, D., N. N. Rabalais,and R.E.Turner, Impacts of cli -mate change on net productivity of coastal waters:Implications for carbon budget and hypoxia, ClimateResearch, 8, 225-237,1997.

Keim,B. D., G. E. Faiers, R.A.Muller, J. M.Grymes,andR.V. Rohli,Long-term trends of precipitation and runoffin Louisiana,USA, International Journal ofClimatology, 15, 531-541,1995.

Kiehl, J.T.,Climate change enhanced: Solving theaerosol puzzle,Science, 283(5406),1273-1275,1999.

Legler, D. M.,K. J. Bryant,and J. J. O’Brien,Impact ofENSO-related climate anomalies on crop yields in theUS, Climatic Change, 42(2),351-375,1999.

Lins,H. F.,and J. R.Slack,Streamflow trends in theUnited States, Geophysical Research Letters, 26(2),227-230,1999.

Lo,C. P., D.A.Quattrochi,and J. C.Luvall,Application ofhigh-resolution thermal infrared remote sensing andGIS to assess the urban heat island effect,International Journal of Remote Sensing , 18, 287-304,1997.

Lockwood, J. G.,Is potential evapotranspiration and itsrelationship with actual evapotranspiration sensitive toelevated CO2 levels?, Climatic Change, 41(2),193-212,1999.

McGowan, J.A., D. R.Cayan,and L.M.Dorman,Climate-ocean variability and ecosystem response in the north-east Pacific, Science, 281, 210-217,1998.

Doyle,T.W.,Modeling global change effects on coastalforests,in Vulnerability of Coastal Wetlands in theSoutheastern United States: Climate ChangeResearch Results, US Geological Survey, BiologicalScience Report USGS/BRD/BSR 1998-0002,Lafayette,Louisiana,pp.67-80,1998.

Eamus, D.,and P.G. Jarvis,The direct effects of increasesin the global atmospheric CO2 concentration on natu-ral and commercial temperate trees and forests,Advances in Ecological Research , 19, 1-55,1989.

Felzer, B.,and P. Heard,Hydrological implications ofGCM results for the US national assessment, Journal ofAmerican Water Resources Association, 35(6),1327-1339,1999.

Glasmeier,A.,The history of Central and SouthernAppalachia: A socioeconomic analysis,paper presentedat the Central and Southern Workshop on ClimateVariability and Change,Morgantown,West Virginia,May27-28,1998.

Goolsby, D.A.,W.A.Battaglin, G. B.Lawrence,R.S.Artz,B.T.Aulenbach,R. P. Hooper, D. R. Keeney, and G. J.Stensland, Flux and source of nutrients in theMississippi-Atchafalaya River Basin, In Review,WhiteHouse Office of Science and Technology Policy,Committee on Environment and Natural Resources,Hypoxia Working Group,May 1999, pp.13-37,2000.

Harcombe, P.A.,R.B.W. Hall, J. S.Glitzenstein,E.S.Cook,P. Krusic,and M.Fulton, Sensitivity of Gulf Coastforests to climate change,in Vulnerability of CoastalWetlands in the Southeastern United States: ClimateChange Research Results, US Geological Survey,Biological Science Report,USGS/BRD/BSR 1998-0002,Lafayette,Louisiana,pp.47-67,1998.

Hatch, U.,S. Jagtap, J. Jones,and M.Lamb, Potentialeffects of climate change on agricultural water use,Journal of American Water Resources Association,35(6),1551-1562,1999.

Hefner, J. M.,Southeast wetlands:Status and trends,mid-1970’s to mid-1980’s,US Department of Interior, Fishand Wildlife Service,Atlanta,Georgia,32 pp.,1994.

Heinz Center, The Hidden Costs Of Coastal Hazards:Implications For Risk Assessment And Mitigation,Island Press,Washington,DC, 220 pp.,1999.

162


McNulty, S. G., P. L.Lorio,Jr.,M. P.Ayres, and J. D. Reeve,Predictions of Southern Pine Beetle populations usinga forest ecosystem model,in The Productivity andSustainability of Southern Forest Ecosystems in aChanging Environment, edited by R.Mickler and S.Fox, Springer Publishing,pp.617-634,1998.

McNulty, S. G., J. M.Vose,W.T. Swank, J. D.Aber, and C.A.Federer, Landscape scale forest modeling: Data basedevelopment,model predictions,and validation using aGIS, Climate Research , 4, 223-231,1994.

Mearns,L. O., F. Giorgio,L.McDaniel,and C.Shields,Analysis of daily variability of precipitation in a nestedregional climate model: Comparison with observationsand doubled CO2 results, Global Planetary Change,10, 55-78,1995.

Meyer, J. L.,Seasonal patterns of water quality in black-water rivers of the coastal plain,southeastern UnitedStates,in Water Quality in North American RiverSystems, edited by C. D. Becker and D.A.Neitzel,Battelle Press,Columbus,Ohio,pp.249-276,1992.

Mulholland, P. J., G. R.Best,C.C.Coutant, G. M.Hornberger, J. L.Meyer, P. J. Robinson, J. R.Stenberg,R.E.Turner, F.Vera-Herra,and R. G.Wetzel,Effects of climatechange on freshwater ecosystems of the southeasternUnited States and the Gulf Coast of Mexico,Hydrological Processes, 11, 949-970,1997.

National Climatic Data Center, Billion dollar US weatherdisasters,1980-1999, PDF version,www.ncdc.noaa.gov/ol/reports/billions.html.July 20,1999.

National Council of the Paper Industry for Air andStream Improvement (NCASI),Global change and for-est responses: theoretical basis,projections,and uncer-tainties,NCASI,New York,Technical Bulletin No.690,pp.11-18,1995.

Nijs,I.,I.Impens,and T. Behacghe,Leaf and canopyresponses of Lolium perenne to long-term elevatedatmospheric carbon-dioxide concentration, Planta177, 312-320,1988.

Pielke,Jr.,R.A.,and R.A.Pielke,Sr.,Vulnerability to hur-ricanes along the US Atlantic and Gulf coasts:Considerations of the use of long-term forecasts,inHurricanes: Climate and Socioeconomic Impacts , edit-ed by H. F. Diaz and R.S.Pulwarty, Springer Publishing,New York,147-184,1997.

Quattrochi, D.A., J. C.Luvall,M. G. Estes,Jr.,C. P. Lo,S. Q.Kidder, J. Hafner, H.Taha,R. D. Bornstein,R.R.Gillies,and K. P. Gallo,Project ATLANTA (ATlanta Land useANalysis:Temperature and Air quality): A study of howthe urban landscape af fects meteorology and air qualitythrough time,Preprints, American MeteorologicalSociety, 104-107,1998.

Quattrochi,D.A.,and J. C.Luvall, Urban sprawl andurban pall: Assessing the impacts of Atlanta’s growthon meteorology and air quality using remote sensingand GIS, Geographical Information Systems, 9, 26-33,1999.

Rabalais, N. N.,R.E.Turner, D. Justic´, Q. Dortch,and W.J.Wiseman,Characterization of hypoxia, Topic #1, Gulfof Mexico Hypoxia Assessment, NOAA CoastalProgram Decision Analysis Series, National Oceanicand Atmospheric Administration,Coastal Ocean Office,Silver Spring,Maryland,144 pp.plus 3 appendices,1999.

Ricketts,T. H.,E.Dinerstein, D. M.Olson,and C.Loucks,Who’s where in North America? Patterns of speciesrichness and the utility of indicator taxa for conserva-tion, BioScience, 49(5),369-381,1999.

Ritschard,R.L., J. F. Cruise,and L. U. Hatch,Spatial andtemporal analysis of agricultural water requirements inthe Gulf Coast of the United States, Journal ofAmerican Water Resources Association, 35(6),1585-1596,1999.

Rosenberg, N. J.,R.C.Izaurralde,M.Tiscareño-Lopéz, D.Legler, R.Srinivasan,R.A.Brown,and R. D. Sands,Sensitivity of North American agriculture to ENSO-based climate scenarios and their socio-economic con-sequences: Modeling in an integrated assessmentframework, Pacific Northwest National Laboratory,Seattle,Washington,146 pp.,1997.

Short, F.T.,and H.A.Neckles,The effects of global cli-mate change on seagrasses, Aquatic Botany, 63, 169-196,1999.

Southern Oxidants Study Team. 1995. The state of theSouthern Oxidants Study: Policy-relevantFindings inozone pollution research,1988-1994. SouthernOxidants Study, Raleigh,North Carolina,94 pp.

Timmermann,A., J. Oberhuber,A.Bacher, M.Esch,M.Latif, and E.Roeckner, Increased El Niño frequency in aclimate model forced by future greenhouse warming,Nature, 398, 694-697,1999.

163


ACKNOWLEDGMENTSDedicationThis chapter is dedicated to Dr. Ronald Ritschard,a good friend, colleague, and scientist, withoutwhose efforts this Assessment would not have beenpossible. Ron will be deeply missed.

Many of the materials for this chapterare based on contributions from participants on and

those working with the

Central and Southern Appalachians Workshop TeamWilliam T. Peterjohn (PI),West Virginia UniversityRichard Birdsey, USDA Forest ServiceAmy Glasmeier, Pennsylvania State UniversitySteven McNulty, USDA Forest ServiceTrina Karolchik Wafle,West Virginia University

Gulf Coast Workshop and Assessment TeamZhu Hua Ning*,Southern University and A & M

CollegeKamran Abdollahi*,Southern University and A & M

CollegeVirginia Burkett,USGS National Wetlands Research

CenterJames Chambers,Louisiana State UniversityDavid Sailor,Tulane UniversityJay Grymes,Southern Regional Climate CenterPaul Epstein,Harvard UniversityMichael Slimak,US Environmental Protection Agency

Southeast Workshop and Assessment TeamsRon Ritschard*,University of Alabama – HuntsvilleJames O’Brien*,Florida State UniversityJames Cruise*,University of Alabama – HuntsvilleRobert Abt,North Carolina State UniversityUpton Hatch,Auburn UniversityShrikant Jagtap,University of FloridaJames Jones,University of FloridaSteven McNulty, USDA Forest ServiceBrian Murray, Research Triangle Institute

* signifies Assessment team chairs/co-chairs

Trenberth,K.E.,Conceptual framework for changes ofextreme of the hydrological cycle with climate change,Climatic Change, 42, 327-339,1999.

Tsuji, G., G. Hoogenboom, P.Thornton,(Eds.),Understanding Options for Agricultural Production,Kluwer Academic Publishers,399 pp.,1998.

US Bureau of the Census,County and City Data Book,US Government Printing Office,Washington,DC,GPO#003-024-08753-7,1994.

US Department of Agriculture,Census of Agriculture,National Agricultural Statistics Service,Washington,DC,1994.

US Department of Agriculture,Census of Agriculture,National Agricultural Statistics Service,Washington,DC,1996.

US Department of Agriculture, Census of Agriculture,National Agricultural Statistics Service,Montgomery,Alabama,1999.

US Forest Service,The South’s fourth forest:alternativesfor the future, US Department of Agriculture, ForestService,Washington DC, Forest Resource Report No.24, 512 pp.,1998.

Williams,K.,K.C.Ewel,R. P. Stumpf, F. E.Putz,and T.W.Workman,Sea-level rise and coastal forest retreat onthe west coast of Florida, Ecology, 80(6), 2045-2063,1999.

164


167

CHAPTER 6

POTENTIAL CONSEQUENCES OF CLIMATEVARIABILITY AND CHANGE FOR THEMIDWESTERN UNITED STATESDavid R. Easterling1,2 and Thomas R. Karl1,2


Chapter Summary



Ecological Context


Key Issues

Water Resources

Agricultural Ecosystems

Natural and Semi-natural Ecosystems

Quality of Life

Additional Issues



Literature Cited

Acknowledgments

1NOAA National Climatic Data Center; 2Coordinating author for the National Assessment Synthesis Team


Over the 20th century, the northern portion of theMidwest,including the upper Great Lakes,haswarmed by almost 4°F, while the southern portionalong the Ohio River valley has cooled by about 1°F.

Annual precipitation has increased,up to 20% insome areas,with much of this coming from moreheavy precipitation events.


During the 21 st century, it is highly likely that tem-peratures will increase throughout the region,likelyat a rate faster than that observed in the 20 th centu-ry, with models projecting a warming trend of 5 to10°F over 100 years.

Precipitation is likely to continue its upward trend,with 10 to 30% increases across much of the region.Increases in the frequency and intensity of heavyprecipitation events are likely to continue in the 21st

century.

Despite the increase in precipitation, rising air tem-peratures and other meteorological factors are likelyto lead to a substantial increase in evaporation,caus-ing a soil moisture deficit, reduction in lake andriver levels,and more drought-like conditions inmany areas.


The Midwest is characterized by farming,manufac-turing,and forestry. The Great Lakes form theworld’s largest freshwater lake system,providing amajor recreation area as well as a regional watertransportation system and access to the AtlanticOcean via the St.Lawrence Seaway. The regionencompasses the headwaters and upper basin of theMississippi River and most of the length of the OhioRiver, both critical water sources and means ofindustrial transportation providing an outlet to theGulf of Mexico. The Midwest contains some of therichest farmland in the world and produces most ofthe nation’s corn and soybeans. It also has impor -tant metropolitan centers,including Chicago andDetroit. The largest urban areas in the region arefound along the Great Lakes and major rivers. The“North Woods”are a large source of forestry prod-ucts and have the advantage of being situated nearthe Great Lakes,providing for easy transportation.

168


Key Findings

• A reduction in lake and river levels is likely tooccur as higher temperatures drive increasedevaporation,with implications for transportation,power generation,and water supply.

• Agriculture as a whole in this region is likely tobe able to adapt and increase yields with thehelp of biotechnology and other developments.

• For both humans and other animals,a reductionin extremely low temperatures is likely to reducecold-weather stress and mortality due to expo-sure to extreme cold,while an increase inextremely high temperatures is likely to decreasecomfort and increase the likelihood of heat stressand mortality in summer.

• Preventative measures such as adequate storm-water discharge capacity and water treatmentcould help offset the likely increased incidenceof water-borne diseases due to an increase inheavy precipitation events. As temperaturesincrease there is also a chance of more pest-borne diseases,such as St.Louis encephalitis.

• Changes in seasonal recreational opportunitiesare likely, with an expansion of warm weatheractivities during spring and fall,and a reductionduring summer due to excessively hot days.

Cold weather activities are likely to decline aswarmer weather encroaches on the winterseason.

• Boreal forest acreage is likely to be reducedunder projected changes in climate. There is alsoa chance that the remaining forestlands would bemore susceptible to pests,diseases,and forestfires.

• Major changes in freshwater ecosystems are like-ly, such as a shift in fish composition from coldwater species,such as trout, to warm waterspecies,such as bass and catfish.

• Higher water temperatures are likely to create anenvironment more susceptible to invasions bynon-native species.

• The current extent of wetlands is likely todecrease due to declining lake levels.

• Changes in bird populations and other nativewildlife have already been linked to increases intemperature and more changes are likely in thefuture

• Eutrophication of lakes is likely to increase asrunoff of excess nutrients due to heavy precipita-tion events increases and warmer lake tempera-tures stimulate algae growth.

169

Chapter 6 / The Midwestern United States


PHYSICAL SETTING ANDUNIQUE ATTRIBUTESThe Midwest region is dominated by the GreatLakes, two major river systems,and large tracts ofboth forests and agricultural lands. The landscape ofmost of this region was scoured by a series of conti-nental glaciers resulting in thousands of lakes, wet-lands,and huge expanses of relatively flat land. Thelast of these glaciers retreated about 10,000 yearsago and glacial meltwater carved many importantriver valleys. As the glaciers retreated,the lands thatwere once covered by ice up to two miles deepdeveloped some of the richest and most productiveagricultural soils in the world. In addition to theGreat Lakes,they also formed tens of thousands ofsmall- to moderate-size lakes,which have become a

characteristic of the region,the “land of 10,000lakes,” as Minnesota license plates proclaim (Lew,1998).

The Great Lakes are one of the world’s largestinland lake systems,containing 20% of the world’sfreshwater reserve (Botts and Krushelnicki,1987).They also provide a regional water transportationsystem and access to the Atlantic Ocean via the St.Lawrence Seaway. This is a vital competitive advan-tage for transporting manufactured goods and agri-cultural products produced in this region. Thisregion also contains the headwaters and upperbasin of the Mississippi River and most of the lengthof the Ohio River, both critical water sources andmeans of industrial transportation providing an out-let to the Gulf of Mexico.

The Midwest region is comprised of the area northof the Ohio River, including the GreatLakes region,and west to the states adja-cent to the Mississippi River (Figure 1).It contains 12% of the US land area and22% of its population. This area covers awide range of ecosystems with someextensive urban/suburban development.The region can be divided into threesubregions:the rolling forested land-scapes of southern Missouri,Illinois,Indiana,and Ohio;the relatively flatfarmland in the northern portions ofthese states as well as in Iowa;and theheavy evergreen and deciduous forestsof Wisconsin,Minnesota,and Michigan.Native vegetation ranges from mixeddeciduous forests in the south to the so-called Prairie Peninsula,(a region of tallgrass prairie in southwestern Minnesota,Iowa,and large portions of Missouri,Illinois,and southern Wisconsin), givingway to the deciduous and coniferousforests in the northern Great Lakes. Theflora of the region are accustomed toperiodic extreme droughts, flooding,late-spring or early-autumn frost,lowminimum temperature,high maximumtemperature,and severe storms.

POTENTIAL CONSEQUENCES OF CLIMATEVARIABILITY AND CHANGE FOR THE MIDWESTERNUNITED STATES

Figure 1. Map of Midwest region.170

171


ent from predominantly agriculture to predominant-ly forested. Third,the region is at the southern mar-gin of the boreal forest (spruce-fir) and northernportions of the region include boreal species aslocally dominant,especially on wetter sites.

In 1992,the Upper Great Lakes region (Michigan,Minnesota,and Wisconsin) was about 42% (over 50million acres) forestland. Over 90% of the forestlandis used for commercial forestry, and more than halfof the commercial forest land is owned by the non-industrial private sector. The forestry sectoremploys about 200,000 people and produces over$24 billion dollars a year in forest products.Expectations in the industry are for sustained orincreased output of forest products,particularlygiven increasing demand for forest products,decreasing supply from the Pacific Northwest,and

SOCIOECONOMIC CONTEXTThe Midwest is a combination of the Manufacturingand Corn Belts,a region of manufacturing and agri-cultural production on which the entire countrydepends. Shaped by surface water systems andother emerging transportation networks,it devel-oped rapidly in the latter half of the 19th centurywith the arrival of both large numbers of settlersfrom eastern US and immigrants from Europe. It hasrelatively high population density, and numerouspockets of national excellence such as the MayoClinic in Minnesota. The region provides more than40% of the nation’s industrial output and is responsi-ble for 30% of the nation’s foreign agriculturalexports.

Historically, the Midwest’s image as the heart of theManufacturing Belt has been closely associated withthe automobile age,and its prosperity has tradition-ally been tied to the fortunes of the automobileindustry. Other heavy industry has been importantas well,including the production of chemicals,steel,paper, and medical products. Due to foreign compe-tition,an aging industrial infrastructure,and environ-mental issues,the Manufacturing Belt declined incompetitiveness during the 1970s,coming to bereferred to as the Rust Belt. However, the regionresponded with vigorous industrial restructuringrelated to modern technologies and componentssuch as electronics,and also to services and finan-cial industries. Meanwhile,continued improvementsin farming methods and seed stock from newresearch and development has pushed the yields ofcorn,soybeans,fruits, vegetables,and other crops upto previously inconceivable levels. This revitalizedeconomic base is expected to lead to continuedeconomic and demographic growth (see Figures 2and 3),but still lagging behind the rest of the coun-try (with the notable exception of Minnesota).

ECOLOGICAL CONTEXTLand use in the Midwest region is dominated bymanaged ecosystems such as farmland. However,the natural land cover of the region is characterizedby three prominent environmental gradients. First, asouthwest to northeast gradient from prairie to for-est in Minnesota is largely a function of water avail-ability. Second,a south to north gradient fromEastern deciduous (oak-hickory) to Northern mixedhardwood forests (beech,maple,hemlock) inMichigan and Wisconsin is a prominent landscapefeature. These patterns correspond to climatic andsoil gradients and a steep south-north land-use gradi-

Figure 2. Population trend estimate for the Midwest region usingthe baseline assumptions from the NPA Data Services estimates.Under this scenario, the population of the Midwest is expected toincrease by about 30% by 2050.

Figure 3. Percentage of economy by sector for (a) 1997, and (b)2050, estimated from the NPA Data Services using baselineassumptions (NPA 1999b). Under these estimates, by 2050 themanufacturing percentage of the economy decreases by 7%, andthe service sector increases by 5%. See Color Plate Appendix

Midwest Population Estimates and Projections

Midwest Industry Income

1997 Projected 2050

ment. This can affect the availability of seed sourcesand,therefore,slow the migration of species north-ward. This delay can contribute to dieback as com-munities make the transition from one type toanother. Second,the northern forests are stronglyinfluenced not only by climate,but also by the soilspresent,with conifers tending to dominate on thesandy soils,such as found in the north of the region,especially in lower Michigan. Sandy soils are moreprone to drought conditions. Although vegetationmodels consider this influence,the scale of the vari-ation in soil effects is much finer than can be repre-sented in the models. Therefore,soil effects con-tribute to uncertainties in projections.

CLIMATE VARIABILITYAND CHANGEThe climate of this region is typical of an interiorcontinental climate,although the Great Lakes exerta strong influence on nearby areas for both precipi-tation and temperature. Total annual precipitationvaries from a low of about 25 inches in westernMinnesota and Iowa to more than 40 inches peryear along the Ohio River. Rainfall in most of theregion is highly seasonal,with most falling in thesummer. Only in the Ohio River Valley is precipita-tion more consistent throughout the year.Temperatures range widely from winter to summer,year-to-year, and even decade-to-decade,withreduced variability in areas adjacent to the GreatLakes.

Annual mean temperature trends in this region overthe 20th century indicate that the northern portion,including the upper Great Lakes,is warming at arate of almost 4°F/100 years (2°C),but the southernportion along the Ohio River valley has a slight cool-ing trend of around 1°F/100 years (0.6°C). By theend of the 21 st century, the projected changes inmean annual temperature for much of the Midwestare between 5 and 10°F (3-6°C) with about twicethe rate of temperature increase in the Canadian cli-mate model scenario than in the Hadley scenario.In both scenarios,the mean daily minimum temper-ature rises more than the maximum temperature,often by a degree or two F by the end of the 21st

century, a characteristic also observed during the20th century.

The decrease in the annual mean temperatures overthe 20th century along the Ohio River is also associ-ated with a reduction in the number of days in thelate 20th century exceeding 90°F (32°C). Decreases

the already high production from the neighboringsouthern and southeastern regions of the US. Thesecond and third-growth forests of the Upper GreatLakes are maturing,and recent forest inventoriesreport substantial increase in the amount of forestedland and in stocking on those lands. The majority ofAmericans,including those in the region, express adesire for increased emphasis on non-commodityvalues in forest management (e.g., recreation,aes-thetics,and biodiversity). This desire often conflictswith the dependence of rural landowners on forestsfor employment and community development.While both standing volume and demand for forestproducts continue to increase in the Upper GreatLakes,the amount of land available for timber pro-duction continues to decrease due to conversion tourban and industrial uses,and development of sea-sonal and retirement homes.

Two trends in land use should be considered andare very likely to continue for the short term.Declines in the amount of farmland in Michigan,Minnesota,and Wisconsin (a 5% decline between1998 and 1997) was observed in the Census ofAgriculture. Forest cover increased by 3% between1980 and 1993 according to the USDA ForestService forest inventory, and urban sprawl has accel-erated,both replacing important agricultural landuse. Although the pressures causing these changesare still in place (declining agricultural productivityand increasing demand for recreational and aesthet-ic uses of land),it seems unlikely that the trends cancontinue long term. Increasing development anddeclining rates of agricultural abandonment are like-ly to lead to declines in forest area in the longerterm (Warbach and Norberg,1995). Furthermore,large-scale management of forests on private lands isbecoming increasingly difficult as ownership isbecoming increasingly fragmented among manymore and smaller parcels (Norgaard,1994;Brownand Vasievich,1996). Between 1960 and 1990, aver-age private parcel sizes declined by an average of1.2% per year across the region. While this“parcelization”associated with recreational and sea-sonal home development doesn’t necessarily resultin forest clearing,it does affect the management offorests and,therefore,the ability of foresters andthus forests to respond to changing climatic condi-tions.

Given the substantial potential expansion of thetemperate deciduous forests and savannas (oak andhickory dominant) it is important to consider twolimiting factors. First,between two-thirds and three-quarters of these two communities are under activehuman management for agriculture and/or develop-

172


of up to 14 days per year exceeding 90°F in theOhio Valley have been observed, owing in part tothe extreme heat in the first half of the centuryassociated with the intense droughts during the1930s. This is in contrast to the projected increasein the probability of temperatures exceeding 90°Fduring the 21st century. For example,in Cincinnati,the probability of more than half the days in Julyhaving temperatures at or above 90°F increasesfrom 1 in 20 now, to 1 in 2 by 2030 in the Canadianclimate scenario and to 1 in 10 for the Hadleyscenario.

Changes in extreme temperature can be even moredramatic. For example,the Canadian scenario sug-gests that for places like Chicago,the probability ofthree consecutive days with nighttime temperaturesremaining above 80°F (27°C) and daytime tempera-tures exceeding 100°F (38°C) increases from abouta once in 50 years occurrence today, to approxi-mately once every 10 years by the decade of the2030s. By the decade of the 2090s this probabilityincreases to one year in two or 50% for any givenyear. Analysis of the historical data shows no changeduring the 20th century in the number of daysbelow freezing for the southern portion of theregion,and only slight decreases of around 2 daysper 100 years around the Great Lakes. Again,this isexpected to change considerably by the end of the21st century. For example,in southern Wisconsin,the Canadian scenario projects an increase in thenumber of wintertime nights remaining above freez-ing from the current 5 nights per winter to 15 ofthe nights by the end of the 21st century.

The length of the snow season over the last 50years increased by about 6 days per year in thenorthern Great Lakes area,but decreased by asmuch as 16 days in the Ohio River valley and adja-cent areas. Both climate scenarios project adecrease of up to 50% in the length of the snowcover season by the end of the 21st century.Although it is possible that during the next fewdecades lake-effect snows might initially increasewith a reduction in lake ice cover, it is highly likelythat by the end of the 21st century, a reduction inthe conditions favorable for lake-effect snows wouldcause the frequency of lake-ef fect snows todecrease (Kunkel et al.,2000). Therefore,it is likelythat by the end of this century, sustained snowcover (more than 30 continuous days of snowcover) could disappear from the entire southernhalf of the region.

Observed trends for annual precipitation over the20th century show moderate to strong increases

almost everywhere in the region,often exceeding20% per century. On an annual basis,precipitationis projected to increase by 20 to 40% by the end ofthe 21st century in the upper Midwest and decreaseby up to 20% along the Ohio River in the Canadianscenario. However, these changes are not uniformthroughout the year. For the Hadley scenario,increases in precipitation occur everywhere in theMidwest with increases of 20 to 40% common.However, the magnitude of precipitation increasesin the Hadley scenario is unusual compared to other2xCO2 equilibrium models run over the past severalyears in that it does produce more precipitation atthe end of the 21st century (Quinn,et al.,1999 ).

Observed changes in soil moisture calculated fromthe Palmer Drought Severity Index indicate moder-ate to very strong increases in wetness in the east-ern portions of the region. In contrast, even withthe temperature increases there is a strong enoughincrease in precipitation to outweigh increasedevaporation in the Hadley scenario which leads tosmall positive changes of soil moisture content by2100. In the Canadian scenario however, despitethe increase in annual precipitation,a reduction insummer precipitation coupled with the largerincrease in temperature relative to the Hadley sce-nario leads to increased evaporation and reducedsoil moisture content,especially during summer.The frequency and intensity of droughts increase inthe Canadian scenario,but decrease slightly in theHadley scenario by 2100.

Changes in variability related to the changes inextremes are of particular interest. The interannualvariability of the annual mean temperature decreas-es slightly for both model scenarios by the end ofthe 21st century, but the magnitude of this decreaseis very small compared to the increase in the mean.As a result,new record high extremes of tempera-ture are common in both climate scenarios through-out the 21 st century (Karl and Knight,1998;NationalClimatic Data Center, 1999).

The interannual (year-to-year) variability of precipita-tion in both scenarios shows no significant changes,even by the end of the 21st century. But an increasein the mean annual precipitation would likely beaccompanied by even greater changes in heavy dailyprecipitation events (Groisman et al.,1999a). Bothclimate scenarios show increases in precipitation forthe Midwest region,and analysis indicates that thisincrease is occurring due to increases in the highestdaily precipitation amounts (Figure 4). However, inthe Canadian scenario, even with projected increas-es in precipitation,the increased evaporation due to

173


rising temperatures would result in both decreasedrunoff and increased direct lake evaporation, result-ing in decreased lake levels (see water resourcessection below).

KEY ISSUES1.Water Resources2.Agricultural Ecosystems3.Natural and Semi-natural Ecosystems4.Quality of Life

The key issues discussed here were chosen to berepresentative of those issues that are potentiallyaffected directly by climate change,in either posi-tive or negative ways. Moreover, they are issues ofkey economic and/or environmental importance. Itis clear that secondary and even tertiary effects due

174


to climate change among the various sectors arepossible,however these types of effects are verydifficult to quantify and are not considered here.

1. Water Resources

Some climate model scenarios project a reductionin available water with increasing atmospheric CO2

even with increased precipitation,due to enhancedevaporation with increased temperatures outweigh-ing the precipitation increase (Meehl et al.,2000).Decreased lake levels and river flow would posespecial problems for the region,affecting com-merce and recreation and altering entire ecosys-tems. Understanding potential impacts of climatechange on water resources is a key element inunderstanding impacts on a variety of other sec-tors. Issues with the region’s many lakes includingthe Great Lakes,and the river systems including theOhio and Mississippi basins,are critical to theregion’s economy and ecosystems. These issuesinclude water levels, water temperature,ice cover,and water quality.

Lake levels. For the Great Lakes,the potentialimpacts related to the various scenarios of climatechange span a wide range. The Canadian climatescenario projects a reduction in the levels of theGreat Lakes by as much as 4 to 5 feet because ofdecreases in the net basin supply of water causedby reduced land surface runoff and increased evap-oration of lake water. These reductions are project-ed to occur in spite of the projected increases intotal precipitation. Should these changes occur,they would be 3 to 6 times larger than the seasonalvariation of Lake levels,which normally reach aminimum during winter and a maximum in latesummer, unlike many smaller lakes which reachtheir minimum in late summer. The reduction inwater levels projected by the Canadian scenariowould result in a 20-40% decrease in outflow in theSt.Lawrence River. The Hadley scenario,with itssmaller temperature increases and greater precipita-tion increases,leads to smaller changes (a slightincrease of 1 foot) in Lake levels during the 21stcentury (see box,Quinn et al.,1999). However,Chao (1999) presents Great Lakes level changes cal-culated using a number of both transient and equi-librium climate change simulations. For each lakethe results show lake level declines ranging fromless than 1 foot (0.3 meter) to more than 5 feet (1.8meters) (see Figure 5). The differences in theresults from various model projections illustrate thedifficulty in planning long-term adaptation strate-gies because they require consideration of a broadrange of possibilities.

Figure 4. Annual trends in daily precipitation by percentile for the(a) CGCM1 (Canadian model) and (b) HadCM2 (Hadley model) sce-narios. Notice the largest trend is in the heaviest daily precipitationamount for both model simulations indicating that most of theincrease in annual precipitation is due to an increase in precipita-tion on days already receiving large amounts (analysis based onmethod in Karl and Knight, 1998). See Color Plate Appendix

Midwest Daily Precipitation/HadCM2

Midwest Daily Precipitation/CGCM1

175


ed by a change in shipping patterns favoring rail-roads and trucking.

Lake ice cover. Some offsetting changes related tothe water transportation problem are projected inboth the Hadley and Canadian climate scenarios.For example,with water temperature increasesthere would be a longer ice-free season (see box)resulting in a one to two month extension of theshipping season by the end of the 21st century. Thiscould translate into hundreds of millions of dollarsof additional business revenue,however the ship-

If reductions in runoff as large as those projected bythe Canadian climate scenario occur, this wouldhave a substantial ef fect on all the major rivers ofthe region. During the 1988 drought,hundreds ofmillions of dollars were lost due to transportationinefficiencies (Changnon,1989). On an annualbasis, over $3 billion in business revenue and per-sonal income,and 60,000 jobs relate to the move-ment of goods on the Great Lakes-St.Lawrencetransportation system (Allardice and Thorp,1995).Low water levels often lead to gouged ship hulls ordamaged propellers. Dredging operations can beused to offset these problems. In addition to theadded cost of dredging,another risk is the possibili-ty of re-suspending human-made inert toxins andheavy metals lying within the lake bottom sedi-ments,significantly impacting water quality. Such“surprises”are often difficult to anticipate. There isthe potential that the competitive advantage thisregion has had due to reliable and efficienttransportation of goods in and out of the areacould be lost.

This event illustrated the high likelihood for futurecontroversy if droughts of this magnitude becomemore frequent due to climate change. Clearly if thisoccurs the barge industry would be severely impact-

The Drought of 1988

Along with the droughts of the 1930s,the 1988 Midwest drought was one of the worst in the previous 100years and brought home the socioeconomic impacts of these types of short-period climate fluctuations.Major impacts occurred in most sectors of the economy including agriculture, recreation,and transportation.The National Climatic Data Center (1999) lists this drought and associated heat waves as one of the mostexpensive natural disasters in US history, with costs of over $30 billion.

One of the unforeseen impacts was on river-borne transportation on the lower Mississippi River. The droughtreduced water levels in the Mississippi River to the point that barge traffic was restricted,causing bulk com-modity shipping to be shifted to more expensive rail transport. A controversial proposed response was toincrease the diversion of Lake Michigan water into the Mississippi River system via the Chicago River. Thediversion is limited to 3,200 cubic feet per second (cfs) by a US Supreme Court decision and the proposedincrease was to 10,000 cfs. However, a number of factors led to a decision not to implement this increaseddiversion. First,it was determined to be too politically controversial,particularly because Great Lake waterlevels were rapidly falling from their record levels of the previous few years and the increased diversion couldaccelerate this fall. Furthermore, hydrologic studies indicated that the increased diversion would be insuffi-cient to improve conditions anyway.

As with many such impacts,there were winners and losers. Losers included producers such as farmers,petro-leum companies,the coal industry, and others who had to pay more to ship their products. In addition to theriver-based shipping industry itself, other losers were consumers of the commodities normally shipped viariver traffic that had to pay more due to higher shipping costs. There were also negative ecological impactsdue to low water levels including fish kills, wetland damage,and salt-water intrusion up the lower MississippiRiver past New Orleans. The major winners economically were the alternative shippers such as the rail andtrucking industries.

Figure 5. Change in water level for each of the Great Lakes under anumber of climate change scenarios, from Chao (1999). See ColorPlate Appendix

difficulties related to flash flooding events,taxingexisting storm water routing infrastructure. Alreadystates like Illinois have updated their 100- and 500-year rainfall return periods in recognition of thechanging climate (Angel and Huff 1997).Furthermore,Olsen et al.(1999) found evidence forincreased flood risk over the most recent decades inthe lower part of the Missouri basin,and on theMississippi River at St.Louis.

With a warmer climate that is drier due to enhancedevaporation,but also experiencing an increase inheavy precipitation events,it is likely that therewould be changes in eutrophication incidence inlakes. Increased surface water temperature reducesvertical mixing of nutrient rich water between thehypolimnion (the colder bottom layer in a thermallystratified lake) and the epilimnion (the warmer, toplayer) resulting in a reduction in the intensity ofsummer algae blooms. However, heavier rain eventswith associated greater runoff would likely increaseeutrophication because heavy rain events are mostimportant in transporting nutrients such as fertilizerfrom the watershed to the lake (Stephan et al.,1993). Also under this scenario it is likely thatgroundwater could also be affected through a reduc-tion in water tables due to both increased irrigationand a reduction in recharge with enhanced evapora-tion;there would also be an increased risk of well-head contamination during flooding events (Rose etal.,1999).

2. Agricultural Ecosystems

Agriculture in this region is vitally important to thenation and the world. With continued intensiveapplication of technological advances,the impact ofprojected climate changes on agricultural yieldswould likely increase the high productivity of theregion. Larger impacts could be related to changesin growing conditions abroad affecting crop prices.

Under climate change scenarios,a longer growingseason would likely translate into increased farmproduction. Although only 4% of the farms in theUpper Great Lakes region have irrigation capabili-ties, even non-irrigated crops are likely to increaseyields with an increased growing season length.Increasing irrigation capabilities would allow farm-ers to take advantage of the increased growing sea-son length and push yields even higher. Althoughsoil-types and farming practices in the Great Lakesregion are more suitable for current crop types thanthose grown further south,these limitations are like-ly to be outweighed by increases in growing seasonlength and warmer summer temperatures (Andresen

ping cost per ton would likely increase for reasonspreviously discussed due to low water levels (Chao,1999,see Figure 6).

Improved understanding of Great Lakes ice cover, itsclimatic variability, and the ensuing economic andecological impacts is needed for the development ofstrategies for adaptation to potential climate change.Reduced ice cover duration over the 21st century, asrecent studies suggest (see box,Quinn et al.,1999),would have feedbacks on lake evaporation,lake lev-els,and even lake-effect snowfall (by possibly affect-ing the seasonality of lake effect snowstorms).Ecosystem services that could possibly be lost witha reduction of lake ice include:1) storage of air-borne atmospheric particulates until their rapidrelease in the spring,2) enhancing overwinter sur-vival of fish and fish eggs,and 3) protecting theshore against erosion.

For the thousands of smaller lakes in the region,icecover is projected to form every year only in lakesin Minnesota,upper Michigan,and Wisconsin by theend of the 21st century (Fang and Stefan,1998),butshortened by nearly half of its normal duration.These changes could eliminate fish winterkill inmost shallow lakes,but possibly can endanger tradi-tional recreational users of these lakes because ofreduced ice thickness.

Storm-water routing and flood plains. Despite thelower water levels as temperatures increase,it islikely that a continuation of the increase in heavyprecipitation events would occur as the climatewarms (IPCC,1996;Groisman et al.,1999a;Groisman et al.,1999b;Karl and Knight,1998). Partsof the region,including Indiana and much of thearea around Chicago are already at a high risk relat-ed to the number of residents living in the 500-yearflood plain. The addition and expansion of impervi-ous surfaces such as pavement can compound the

176


Figure 6. Change in shipping costs under a number of climatechange scenarios, from Chao (1999). See Color Plate Appendix

et al.,1999). Figure 7 shows the projected changesin the climate of Illinois during the 21st centuryusing the two climate scenarios. For the Hadley cli -mate scenario, farming conditions during the sum-mer will be like those in West Virginia by 2030 andlike the eastern part of North Carolina by the end ofthe 21st century. In contrast,the much drierCanadian scenario projects Illinois’ climate to be

similar to that of Missouri and Arkansas by 2030 andeastern Texas and Oklahoma by the end of the 21st

century. Simulation results by Izaurralde et al.(1999) suggest that the Corn Belt in general,is likelyto experience an increase in corn yields due to boththe CO2 fertilization effect,and decreased plantstress due to low temperatures. In particular, resultsfrom the Hadley scenario show a simulated warm-

177


Great Lakes Water Diversion

In 1900,the city of Chicago built the Chicago Sanitary and Ship Canal to keep sewage from contaminating theChicago water supply intakes in Lake Michigan. The flow of water down the Chicago River was reversed.Sizeable amounts of water were diverted from Lake Michigan. This diversion launched a series of continuinglegal controversies involving Illinois as a defendant against claims by the federal government, various LakeStates,and Canada (which wanted the diversion stopped or drastically reduced). During the past 96 years,extended dry periods lowered the lake levels. Using these dry periods as surrogates for future conditions,their effects on the past controversies were examined as analogs for what might occur as a result of climatechange. The results suggest that changing socioeconomic factors,including population growth,will likelycause increased water use,with Chicago seeking additional water from the Great Lakes. New priorities forwater use will emerge as in the past. Future reductions in available water could lead to increased diversionsfrom the Great Lakes to serve interests in and outside the basin. Lower lake levels in the future could lead toconflicts related to existing and proposed diversions,and these conflicts would be exacerbated by the conse-quences of global warming. Costs of coping with the new water levels could also be significant. Should LakeMichigan levels drop as much as 5 feet by the end of the 21st century as predicted by the Canadian climatescenario,it is possible that a lowering of the level of the Canal would be needed. To lower the canal by 4feet,at least 30 miles of the canal would need to be dredged,and 15-17 miles would be rock excavation athuge financial costs (Injerd,1998). A warmer climate, even with modest increases in precipitation,will likelylead to a drier climatic regime and will tax the economy and challenge existing laws and institutions for deal-ing with Great Lakes water issues.

Great Lakes Physical Impacts

Hydrologic impact analyses for the Great Lakes were developed for 20-year periods centered about 2030,2050,and 2090 using the Canadian and Hadley climate scenarios. These scenarios were used with the GreatLakes Environmental Research Laboratory’s Advanced Hydrologic Prediction suite of models (Quinn et al.,1999) and freezing degree-day ice cover models to assess impacts to Great Lakes water supplies,lake levels,ice cover, and tributary river flows for the 121 Great Lakes major tributary basins. Relative changes in hydro-logic factors predicted by the two scenarios compared to the 1961-90 base period are summarized below.

1. Increases in precipitation are offset by increases in evaporation with higher temperatures in the Canadianclimate scenario resulting in decreases in Lake levels of between 4 and 5 feet by the year 2090. TheHadley scenario shows modest increases in Lake levels, generally less than 1 foot,largely due to smallerincreases of temperature accompanying the increase in precipitation.

2. The Canadian climate scenario leads to a decrease in mean annual outflow from each lake of -20 to -30% by2090,whereas the Hadley scenario produces a modest increase in outflow of between 2 and 7%.

3. Both scenarios produce increases in water surface temperatures between 4.5°F (2.5°C) and 9°F (5°C).

4. Ice cover duration for Lakes Superior and Erie show decreases ranging from 29 to 65 days depending onthe model scenario used.

ning 5-year mean. Occasionally, severe drought,heat,or flooding can cause larger yield reduc-tions,such as in 1983,1988,and 1993. It is pos-sible that severe drought will pose the biggestthreat to future crop production. Severedroughts,such as in 1988,can cut production by30% or more,and heavy flooding events canhave a similar effect (see Figure 8). If moredroughts or wet spells do become more com-mon,adaptations in farming practices and cropselection would be necessary to offset yieldreductions due to these types of climate events.

Dairy productivity is also directly affected bytemperature. For example,optimal ranges fordairy cattle are between 40 and 75°F (4 and24°C);dairy cattle are sensitive to heat stress and

high humidity (Wolfe,1997). As the climate warms,it is likely additional measures such as artificial cool-ing methods will be necessary to ensure that theproductivity of livestock is not reduced by extremesof heat. On the other hand,with warmer winters itis likely that productivity will be enhanced by areduction in cold stress. It is uncertain whetherenhanced productivity with the reduction in coldstress will offset the costs of dealing with increasedheat stress.

The types of grasses,including crops that grow inthe Midwest,are a mixture of C3 and C4 type grass-es. The C4 grasses include warm weather crops,such as corn,and the weedy grasses,such as crab-grass and many other creeping perennials. The C3grasses,which dominate in the northern portions ofthe region,include cultivated grasses such as wheatand ryegrass that thrive in cooler growing seasonconditions. Total biomass transient climate changeexperiments, examining the response of grasslandsto gradual changes in climate,show that total bio-mass in grassland regions of the western Midwestthat are initially dominated by C3 perennial grasseswould be replaced by C4 perennial grasses (Coffinand Lauenroth,1996). The change in grass types isstrongly determined by increased temperaturethroughout the year. Because many weedy creepingperennials are C4-like,this could pose problems foragricultural areas in the western and northernMidwest where the current dominant natural vege-tation is C3 grasses. Agriculture has contributed topollution in lakes and rivers of the region includingbioaccumulation of fertilizers and pesticides in fish.The use of herbicides is likely to increase as temper-atures increase and the dominant grass typebecomes the C4 grasses (Allardice and Thorpe,1995). Conflicting priorities between agriculturalyields and water quality are likely to be exacerbated.

ing in minimum temperatures of over 3°C (5.5°F) bythe end of this century that helps contribute to sim-ulated corn yield increases (Izaurralde et al.,1999).

The year-to-year variability found in corn and soy-bean yields is primarily driven by growing seasonweather conditions. The climate variability of theMidwest since 1960 has been rather small and yieldsin most years have been within 10 to 15% of a run-

178


Figure 8. The relationship between Midwest soybean yield and pre-cipitation is shown here. Soybean yields in thousands of bushelsare shown as the differences from the average yield in recentdecades. Precipitation is the difference from the 1961-90 averageprecipitation. Note that lower yields result from both extreme wetand extreme dry conditions. Soybean yields from NationalAgricultural Statistics Service, USDA

Figure 7. Illustration of how the summer climate of Illinois wouldshift under the (a) CGCM1 (Canadian model) scenario, and (b)HadCM2 (Hadley model) scenarios. For example, under the CGCM1Canadian scenario, the summer climate of Illinois would becomemore like the current climate of southern Missouri in 2030 and morelike Oklahoma’s current climate in 2090.

Summer Climate Shifts

Midwest Soybean Yield and Precipitation

Soils in the Midwest region contain significantamounts of carbon and the management of the vege-tation above the soil strongly influences the carbonthat is removed from and is stored in those soils.Land management activities that influence soil car-bon include deforestation and afforestation,biomassburning,cultivation crop residue management,appli-cation of inorganic fertilizer and organic manures,and the various farming systems (Lal et al.,1998).On agricultural lands,soil management practicessuch as conservation tillage, mulch farming, watermanagement,soil fertility management,liming,andacidity management have been shown to increasethe carbon stored in agricultural soils.However, theclimate change effects on soil,particularly soil car-bon storage,are difficult to predict. For example,inagricultural areas where crop residue is available andincorporated into the soil system,there tends to be abuild up of soil carbon. This is especially evident inthe productive soybean/corn rotations where nitro-gen fixation by the beans and the additional residueinput from corn rapidly builds soil carbon.

Adaptation OptionsAgriculture has exhibited a capacity to adapt to mod-erate differences in growing season climate. It islikely that agriculture would be able to adapt underthe moderate climate change scenarios produced bymany GCMs. There are several adjustment possibili-ties that are already used and could be employed inthe future to adapt to climate change. Some of thesepossibilities are farm practices (e.g.,earlier plantingand harvests, changed planting densities,integratedpest management,conservation tillage) and develop-ment of new varieties and farming technologies.There are already examples of individual farmers get-ting the benefit of double cropping (planting a sec-ond crop after the first is harvested) as the growingseason has noticeably increased in just the past fewyears,but these are isolated cases at present. The

potential for double cropping soybeans with awarmer climate was examined using the twoAssessment scenarios. If only temperature is consid-ered,simulation results using both the Canadian andHadley scenarios indicate that the potential existsfor double cropping even in the northern parts ofthe region. However, the limiting factor appears tobe a lack of adequate soil moisture. If a secondcrop of soybeans is grown,particularly in morenorthern sites,adequate yields would likely bedependent on the use of irrigation (Garrity andAndresen,1999).

If extreme climate conditions like the droughts ofthe 1930s or 1988 become more common,thenadditional adaptation measures would be necessary.Some possibilities include shifting the mix of crops,increasing irrigation,and removing marginal landsfrom production. Hybrid strains of corn, for exam-ple,have been developed that allow the corn to begrown under a wide range of conditions. If climateconditions in the Midwest become unsuitable forthe current hybrids to be grown, one adaptationmechanism would be a switch to a hybrid more suit-ed to future climate conditions. Furthermore,thereis evidence that carbon dioxide fertilization by itselfcan enhance crop production. Studies indicate thatthis effect,coupled with a warmer climate,out-weighs poor soil quality in the northern parts of theregion and could allow corn and soybean crops tobe grown in areas where they are not currentlyfound (Andresen et al.,1999).

3. Natural and Semi-naturalEcosystems

Both natural and semi-natural ecological systems onland and in water will be affected by the sustainedwarming projected by the climate models. Water-based ecosystems are currently under multiple

179


Agriculture and Drought

The agricultural impacts of the drought of 1988 illustrate potential impacts of one possible future climate sce-nario,one that is warmer and much drier. The temperature and precipitation conditions in 1988 were similarto conditions in the 1930s. Overall grain production was down by 31%,with corn production down by 45%.The supply of grain was adequate to meet demand because of large surpluses from previous years;however,the drought reduced surpluses by 60%. Interestingly, overall farm income was not reduced because grainprices increased substantially (35% for corn,45% for soybeans);however, these overall figures mask large loss-es in the heart of the drought region where yield reductions were much larger than the national average. Thereduced production caused slightly higher domestic food prices,estimated at a 1% increase in 1988 and a 2%rise in 1989 (in 1989,the drought persisted in some areas and surpluses were much reduced following 1988).In summary, the drought of 1988 demonstrated that the agribusiness sector remains vulnerable to severe cli-matic anomalies,despite decades of advances in agricultural technology.

and noise pollution,and in reducing storm waterrunoff. Furthermore,urban forests help reduce urbanair temperatures through increased shade and evapo-transpiration,which reduces air conditioning needsand summer-time energy demand (McPherson et al.,1997). The more than 70 million acres of urbanforests in the US are dwindling (American Forests,1998,Wong,1999). With almost 80% of the US popu-lation living in urban areas,the health of these smallpockets of forest is a growing concern (Sadof andRaupp,1999). Because of their urban environment,these forests are already stressed beyond what a com-parable rural forest would be in the same region,andclimate changes that could af fect the health of ruralforests would likely be magnified in the urban set-ting. Encouraging the planting of appropriate treespecies in urban areas wherever possible would bebeneficial.

Fragmentation of large forest tracts in the region isanother major stress on the forest industry, and thiscould become an even greater problem if scenariosof vegetation change are correct. Competing landuse,particularly between forest and agriculture,couldbecome a major problem,possibly significantly reduc-ing the current extent of forestland. Forest fragmen-tation also severely affects wildlife, by reducing pro-tective cover and natural migration stopover habitat.Furthermore,there is concern that natural responsesto climate change are likely to be nonlinear, with anapparent resistance to change up to a certain thresh-old,beyond which time a rapid or catastrophic transi-tion could occur (Moll,1998).

Model simulations under both the Hadley andCanadian climate scenarios project the disappearanceof the boreal forest from the region,and a consistentand substantial reduction in the amount of area cov-ered by both the temperate continental coniferousforest and cool temperate mixed forest types as well.This suggests that the northern hardwood forests thatsustain the regions forest products industry are likelyto undergo substantial conversion to temperatedeciduous forest and temperate deciduous savannas.The results for the grasslands are mixed,dependingon the moisture projections in the climate scenariosand the assumptions about water use in vegetationmodels. The fact is,however, that very little naturalgrassland remains and the fate of the grasslands hasmore to do with agricultural policies and economicconditions than climate (Brown ,1999).

Fish and WildlifeWildlife impacts are a critical issue. For example,since the mid-20th century songbird populations havedeclined in the Midwest for a variety of reasons

stresses such as eutrophication,acid precipitation,toxic chemicals,and the spread of exotic organisms.Climate change will interact with these existingstresses,often in not very well understood ways.

Forest types in the Midwest range from the decidu-ous forests of Missouri,southern Illinois,Indiana,and Ohio,to coniferous forests in northernMinnesota,Wisconsin,and Michigan. The upperMidwest has a unique combination of soil and cli-mate that allows for abundant coniferous treegrowth. For example,Michigan is second only toOregon in Christmas tree production. It is possiblethat regional climate change will displace boreal for-est acreage,and make current forestlands more sus-ceptible to pests and disease. A decrease in soilwater content as projected by the Canadian sce-nario would increase the potential for more droughtand excessive wet periods,increasing forest firepotential. Higher temperatures could also increasegrowth rates of marginal forestlands that are cur-rently temperature limited,but can also reducestands of aspen,the major hardwood harvested inthe northern Midwest. The southern transition zoneof the boreal forest is susceptible to expansion oftemperate deciduous forests,which in turn wouldhave to compete with other land use pressures.This could have major impacts on the boreal forestindustry, such as Christmas tree production.

Oak/hickory forests in the southern Midwest arealso a major resource for the forestry industry. Oakdecline or oak dieback is caused by a complex inter-action of environmental stresses. Trees weakenedby environmental stresses such as drought, flooding,or insect defoliation could then be invaded andkilled by insects or diseases that could not success-fully attack a healthy tree (Wargo et al.,1983).Although commonly called oak decline,this prob-lem is not confined to oak or other deciduous trees(see box). Because the initial stresses leading to thisproblem are either drought or excess wetness,theCanadian scenario’s combination of temperatureand precipitation change would likely reduce soilmoisture leading to more drought-like conditionsand increasing the likelihood of this kind of declinein both deciduous and coniferous species.

ForestsAnother area of concern is the urban forest. Theseare tracts of forest-type land in urban settings suchas parks or trees lining streets and are particularlyimportant for large urban areas such as Chicago andDetroit. Besides beauty, these urban forests providepractical benefits such as shade,wind breaks,habitatfor urban wildlife,and help in mitigating water, air,

180


including habitat destruction and fragmentation(Robinson,1997). However, bird populations,likeother wildlife,are particularly susceptible to extremeclimate events. Periods of excessive wetness andparticularly drought affect both habitat and food sup-plies, resulting in diminished reproduction and adecline in populations the next year. Furthermore,the locations of the climate extremes affecting migra-tory Midwest bird populations could very well beoutside the borders of the US, for example in CentralAmerica,where the birds spend the winter(Robinson,1997).

Climate changes could have large and unpredictableimpacts on aquatic ecosystems. Along with concernsregarding changes in water availability and ice cover,changes to other characteristics such as water tem-peratures are a concern. For example,Great Lakessurface water temperatures are likely to increase byas much as 5°C by the end of the 21st century (seebox). Water temperature increases in deep inlandlakes can lead to a decrease in dissolved oxygen con-tent and primary productivity, and a decline in cold-water fish populations. A rise in water temperaturewould change the thermal habitats for warm-,cool-,and cold-water fish. The thermal habitat for warm-and cool-water fishes would likely increase in size,but could be eliminated for cold-water fish in lakesless than 13 meters (43 feet) deep (especially thosein eutrophic states) and in many streams(Magnusson,1997). Species such as brown and rain-bow trout are at risk (EPA,1995;1997) as well asother cold water fish found in shallow lakes.Furthermore,Chao (1999) used a number of GCM-based scenarios of climate change to examinechanges in cold water habitat in Lake Erie,(definedas a well-oxygenated hypolimion),and found reduc-tions of 50 to 80% depending on the scenario.

WetlandsBecause Great Lakes’ water levels have an unusualseasonal cycle (related to winter precipitation beinglocked up as snow/ice cover with frozen soils),reduced ice cover and higher temperatures are like-ly to affect the winter Great Lakes ecosystem bychanging the annual cycle. The annual and interan-nual water level fluctuations act as a perturbation towetland biophysical systems and maintain a moreproductive intermediate stage of wetland develop-ment. Changes in the annual cycle could be com-pounded due to substantially lower water levels,which would lead to more pressure on wetlands.Projected lake-level changes would likely affect wet-land distributions, resulting in a displacement or lossof current wetlands (Mortsch,1998). These wet-lands serve a variety of important functions includ-ing waterfowl habitat, fish breeding areas,and natu-ral improvement of water quality. Wetlands particu-larly at risk are those in areas where the topographyinhibits successful wetland migration,such as areaswith irregular topography.

Invasive SpeciesInvasive species are another current stress thatcould be exacerbated under climate change. Forexample,the zebra mussel has become a majorproblem in the Great Lakes system. Zebra musselsare small, fingernail-sized mussels native to theCaspian Sea region of Asia. It is believed that theywere transported to the Great Lakes via ballastwater from a transoceanic vessel. The ballast water,taken on in a freshwater European port, was subse-quently discharged into Lake St.Clair near Detroit,where the mussel was discovered in 1988. Sincethat time,it has spread rapidly to all of the GreatLakes and waterways in many US states,as well asOntario and Quebec. Hydroelectric power plants,municipal drinking water facilities,and other water-

181


Drought, Insects, and the Decline of Forest Species

In the late 1930s,many hemlock trees showed signs of deterioration in the Menominee Indian Reservation ineast-central Wisconsin. Hemlock is an important forest tree at Menominee,which is located about 30 miles(50 km) north of the southwestern range limit of the tree. By the late 1930s,the hemlock borer, ordinarilynot a problem, reached epidemic proportions. Careful examination by Secrest et al.(1941) revealed thatextensive root damage had occurred during the drought years 1930-1937. Borer attacks in 1938 were suc-cessful only on trees that had 10% or less of their root system still alive. In 1939,attacks were successful onlyon trees with less than half of the main lateral roots alive. Obviously hemlock borer attacks were successfulonly on trees that were already heavily damaged by unfavorable climatic conditions. The 1930s droughtswere ultimately responsible for the loss of trees near the range limit,but insect attack was the proximatecause of death of trees already weakened by drought. The implication of this example is that unfavorable cli-matic conditions may not kill trees outright,but by stressing the trees, climate can contribute to death byinsect attack.

temperatures (at or above 100°F during the day and80°F during the night) are likely to increase in fre-quency, with 3-day consecutive heat waves occur-ring every other year by the end of the 21st century.If heat waves do increase in frequency and severitythis would likely lead to more temperature-relatedmortality and morbidity. This is particularly true inmost large Midwest cities that experience thesekinds of events relatively infrequently and wherethe population is not accustomed to these events(Kalkstein and Smoyer, 1993). Studies have shownthat during the first heat wave of a season,mostheat-related deaths begin to occur on the second orthird day of the heat wave (Kalkstein and Greene,1997). With a diverse population that ranges fromcrowded urban settings to dispersed farming com-munities,impacts are likely to vary considerablyamong different social and demographic groups.For example,under current climate change scenar-ios,more frequent extremely hot days are likely tooccur more often in urban areas where the “heatisland effect” reduces nighttime cooling. The recentsusceptibility to heat-related mortality in places likeChicago has been attributed to factors such as agrowing elderly lower income population that can-not afford or would not use air conditioning(Kunkel et al.,1999). However, recent adaptationand responses to potentially deadly heat waves,suchas enhanced warning and educational programs,have been markedly improved and would likely helpreduce severity of the impact on human mortality inthe future. Other concerns for both urban and ruralareas of the region include a possible increase inrespiratory disease due to excess air pollution. Poordispersion and high temperatures are closely relatedto elevated levels of ozone and other air pollutants,potentially compounding the health ef fects ofextremely high temperatures (Karl,1979,see Figure9).

Indirect effects of climate variability on infectiousdiseases are another possible area of health impact,but whether any change in disease rates mightoccur is uncertain and highly specific to each dis-ease. An increase in heavy precipitation eventscould also increase the potential for water-bornedisease outbreaks similar to the one in Milwaukeeduring the 1993 flood (Rose et al.,1999,see sidebarstory). Increases in water-borne infections such asCryptosporidiosis are possible if heavy precipitationevents become more frequent resulting in greatercattle- and human-derived Cryptosporidium con-tamination of surface water in combined sewer andstorm water drainage systems. The water-associateddisease known as “Swimmers itch”could becomemore prevalent if lakes are more often contaminated

using industries are most heavily impacted by zebramussel populations. Mussels colonize the surfacesof pipes,diminishing the flow rate through waterintake pipes. Unless preventive measures are taken,larval zebra mussels colonize the interior parts ofturbines and other equipment,leading to costlyrepairs. On the other hand,zebra mussels do act toimprove water quality through their natural filtra-tion of nutrients.

Temperature can limit the extent of zebra musselcolonization and has likely kept populations in LakeSuperior small. Each mature female produces sever-al hundred thousand eggs during the breeding sea-son,which occurs when the water temperature isabove 54°F (12°C). The longer this period,the moresuccessful colonization is likely to be. Adults areunable to survive prolonged exposure to tempera-tures above 90°F (32°C) and they can tolerate tem-peratures as low as 32°F (0°C),provided they do notfreeze. With warmer summer water temperatures innorthern lakes,such as Lake Superior, the small cur-rent infestation is likely to become as widespread asit is in the lakes in warmer parts of the region.

4. Quality of Life

The Great Lakes and surrounding region is theindustrial heartland of the United States. It is aleader in automobile,paper products,medicine,chemical,and pharmaceutical production.Recreation is also an important part of the regionaleconomy. Although most of these industries are notdirectly vulnerable to climate change,severe indi-rect effects are possible. For example, changes ingovernmental policy designed to address climatechange issues have the potential to dramaticallyimpact certain industries (e.g.,the automobileindustry) affecting the overall economic health ofthe region.

Human HealthIt is possible that human health will be affected byclimate change in the Midwest in a number of ways.Although our understanding of the relationshipbetween cold weather and mortality is weaker thanthe relationship between heat and mortality, milderwinters are likely to have beneficial effects oncold-related mortality. For example,Changnon(1999) argues that 850 lives were saved across theUS during the record warm winter of 1997-1998.But, even with generally milder winters,tempera-tures are still expected to go below 0°F everyother year even out to the end of the 21st centuryin places like Chicago. On the other hand,increases in dangerously high day and nighttime182


183


by larval Schistosoma from birds or mammals,andswimming becomes more frequent with hotter days.Various vector-borne diseases caused by infectiousagents transmitted by mosquitoes or ticks might alsobe effected,but the complex and multiple impactsof climate on transmission make prediction difficult.For example,increases in temperature and precipita-tion extremes can affect mosquito ecology andpotentially the transmission of viral agents such asLacrosse or St.Louis encephalitis,however climatepredictors of encephalitis outbreaks in the US arestill unclear (see Human Health,Chapter 15).

For most of the health effects discussed above,thereappears to be considerable opportunity for adapta-tion,either through increased prevention and/orgreater education. The development of better educa-tional and monitoring programs and increased sup-port for public health systems would help reduceweather-related mortality from heat waves orincreases in vector-borne diseases.

RecreationRecreation is clearly sensitive to climate change. Icefishing and snowmobiling are favorite pastimes inthe upper Midwest,and higher winter temperaturescoupled with a possible reduction in lake-effectsnows are a direct threat to these activities.Reductions in lake-ice cover would significantlyreduce the length of the season for recreationalactivities dependent on ice cover. Changes in theseasonal characteristics of precipitation resulting inmore winter precipitation falling as rain than snowand increased numbers of days above freezing wouldaffect both snowmobiling and skiing opportunities.

Reductions in cold-season recreational opportunitieswould likely be offset by increases in opportunitiesfor more warm-season recreational activities such asgolfing and boating in the fall and spring seasons asthe length of the warm season increases. But as theclimate continues to warm,the number of desirabledays for outdoor summer activities is likely to bereduced. Clearly, habits and preferences for outdoorrecreation options will need to change,and com-mercial interests will need to keep abreast of thesechanges.

1993 Midwest Flooding and Water-borne Disease

The heavy rainfall and major flooding on the upper Mississippi River basin in the spring and summer of 1993was an exceptional and unprecedented event in modern times. By any measure this was a climate anomalyunseen in the modern historical record for this region. The flooding resulted from pre-existing conditionsthat created circumstances ripe for flooding,and day after day of heavy rainfall over the basin.

One unforeseen impact of the flooding was an outbreak of water-borne Cryptospiridiosis in the city ofMilwaukee in the early spring of 1993. Unusually heavy spring rainfall and snowmelt washed fecal materialfrom agricultural non-point sources into streams. This fecal material is often infected with Cryptospiridiumoocysts. The streams emptied their infected load into Lake Michigan near the intakes for the Milwaukeewater supply that made its way into the water treatment system. Engineering malfunctions combined withthe massive flooding resulted in an outbreak of water-borne Cryptosporidiosis that infected over 400,000people and caused over 100 deaths. As a result of this outbreak,the Milwaukee water system was forced toextend the water intake pipes much further from the shoreline,and improve water treatment controls.

Figure 9. Relationship between average summer air temperatureand the number of air stagnation days for the Midwest showing thatas the mean summer temperature increases, the number of air stag-nation days also increases (air stagnation data from Wang andAngell, 1999).

Temperature and Stagnation

ing the construction of dams and levees on riversystems,or the development of new hydroelectricgeneration facilities in the face of potential changesin lake levels and river flow (Smith,1997).

On the other hand,a number of activities and sec-tors will likely be able to adapt easily with no out-side intervention. Agriculture and many types ofbusinesses and industries have benefited and willcontinue to benefit from consistent technologicaladvances and will easily adapt as conditions change.Advances in plant genetics suggest that most crop-ping activities will be able to cope with most fore-seeable climate shifts. Furthermore,the turnoverrate in most large capital items,such as manufactur-ing facilities,is rapid enough that these types offacilities will have turned over or been remodeled atleast once,if not two or more times by the year2050 (Ausubel,1991).

Another critical adaptation strategy is to develop apublic education program regarding the potentialrisks and consequences associated with rapid cli-mate change. For example,the potential for increas-ing fire danger associated with warmer and drierconditions should be communicated to homeown-ers in high fire-risk ecosystems. The increasedpotential for flooding with increases in the frequen-cy of heavy rain events should be communicated toflood plain landowners. With better information,the residents of the region would be better pre-pared to respond to a less certain climate.

CRUCIAL UNKNOWNS ANDRESEARCH NEEDS1. Perhaps one of the most important unknowns

with regard to future climate is reconciling pro-jections of temperature and precipitation changefrom various climate models. Both model scenar-ios used here project warming and increasedprecipitation. However, in the Canadian sce-nario, warming was great enough to increaseevaporation to the point that decreased soilmoisture becomes a critical issue. Since someclimate model simulations even project adecrease in precipitation for the region,thiscould severely impact water levels in both theGreat Lakes and Mississippi/Ohio River systems,ground water, and soil moisture levels. However,other simulations project increases with onlymodest effects on water levels,which leaves thisquestion an important one for future research.

ADDITIONAL ISSUESA number of additional consequences are also likelyto be of importance for the Midwestern US:• Warmer winters are likely to result in savings in

winter heating bills,potentially offsettingincreased cooling bills during summer. Fewerlake-effect snows would reduce transportationproblems,and could reduce building costs.

• Great Lakes issues include impacts on water sup-ply intake pipes due to reduced water levels,andincreased temperatures resulting in reducedwater cooling efficiencies for power plants.

• Erosion of agricultural soil is currently a majorproblem,which could be exacerbated by thepotential for increased heavy downpours.

• Quality of cool season vegetable crops could bereduced with brief high temperature events atcritical stages in crop development.

• With climate changes and changes in naturalecosystems,there are concerns for proper man-agement of game and wildlife and their respec-tive habitat resources in order to maintain cur-rent levels of hunting and other outdoor recre-ational activities that contribute to the regionaleconomy.

• Climate impacts on agriculture will have second-ary impacts on local economies and land use,resulting in winners and losers,and potentiallylarge community changes.

• With a drier climate,freshwater ecosystems maybe more susceptible to the ef fects of such cur-rent problems as acid rain increasing surfacewater acidity, and the effects of increased devel-opment of dams for reservoirs.

• Exotic species will likely continue to invade thisregion and some native species will start or con-tinue to decline. The climate change effects onthese kinds of biological changes are unknown.

ADAPTATION STRATEGIES Perhaps the most important approach to adapting tothe potential effects of climate change is to developand maintain flexibility in vulnerable activities andsectors. This would include, for example,develop-ing water resource policies that are flexible enoughto adapt to the potential for either an increase ordecrease in available surface and ground water.Other strategies include:improving forecasts andwarnings of extreme precipitation events and therelated impacts;as well as designing large infrastruc-ture projects to provide increased capability to copewith climate extremes. This might include rethink-184


2. Another crucial unknown is how extreme weath-er events might change. If precipitation variabili -ty increases, even with no change in averageannual amount,this would have large implica-tions for drought and wet-spell occurrence. It isclear that extreme events such as droughts andheat waves, flooding,and severe winter stormsall can have dramatic effects on agriculture,trans-portation,human health,etc. This is perhaps oneof the most dif ficult but critical issues to addressas far as future climate is concerned,particularlysince relatively little is known about pastchanges in extreme events,and the current gen-eration of climate models still does not resolvemany of these types of events well.

3. Also at issue are ef fects of CO2 fertilization oncrops, forests,and other flora and the ability offarmers to adapt to a changing,but uncertain,future climate in an economically competitiveworld agricultural market. For example,signifi-cant increases in growing season temperatureswill require shifts to new varieties that are moreheat tolerant,do not mature too quickly, andhave a higher temperature optimum for photo-synthesis. This may be achieved through plantgenetics,but it is uncertain as to how the publicin the US and abroad will react to these newgenetically modified food sources. Moreover,new crop types may require abandoning tradi-tional crops and this may be difficult to accom-plish. For example,an apple grower may need tochange varieties,taking years to grow new trees.

LITERATURE CITEDAllardice, D. R.,and S.Thorp,A changing Great Lakeseconomy: Economic and environmental linkages,SOLEC working paper presented at State of the LakesEcosystem Conference, U.S.Environmental ProtectionAgency, , EPA 905-R-95-017,1995.

American Forests (AF),Regional Ecosystem Analysis,Puget Sound Metropolitan Area,Final Report,UrbanForestry Center,American Forests,http://www.amfor.org/ufc/uea/report.html,8 pp.,1998.

Andresen, J.A.,W. B.Sea, G.Alagarswamy, and H.H.Cheng,National Assessment,Upper Great Lakes FinalReport,Agricultural Sector, US Global Change ResearchProgram,10 pp.,1999.

Angel, J.R.,and F.A.Huff, Changes in heavy rainfall inMidwestern United States, J.Water Resources Planning.Management, 123,246-250,1997.

Ausubel, J. H.,Does climate still matter?, Nature, 350,649-652,1991.

Botts,L.,and B.Krushelnicki, The Great Lakes: AnEnvironmental Atlas and Resource Book. USEnvironmental Protection Agency and EnvironmentCanada,Downsview, Ontario,44 pp.,1987.

Brown, D. G.,and J. M.Vasievich,A study of land owner-ship fragmentation in the Upper Midwest, Proceedings,GIS/LIS 96 Conference, Denver, Colorado.,pp.1199-1209,1996.

Brown, D. G. US National Assessment,Great Lakesregional summary report, Land Ecology, US GlobalChange Research Program, pp.51-57,1999.

Changnon, D.,1997:Damaging storms in the UnitedStates: Selection of quality data and monitoringindices,in Preprints of the Workshop on Indices andIndicators for Climate Extremes, NOAA/NationalClimatic Data Center,Asheville,North Carolina,USA,June 3-6,1997,24 pp.,1997.

Changnon,S.A.,The drought,barges,and diversion,Bulletin of the American Meteorological Society, 70(9)1092-1104,1989.

Changnon,S.A.,Impacts of 1997-98 El Nino-generatedweather in the United States, Bulletinof the AmericanMeteorological Society, 80,(9),1819-1928,1999.

185


Izaurralde,R.C.,R.A.Brown,and N. J. Rosenberg,.USregional agricultural production in 2030 and 2095:Response to CO2 fertilization and Hadley ClimateModel (HADCM2) projections of greenhouse-forced cli-matic change, Pacific Northwest National Laboratories,Richland,Washington, Rep.No.PNNL-12252,42 pp.,1999.

Kalkstein,L.,and J. S.Greene,An evaluation of cli-mate/mortality relationships in large US cities and thepossible impacts of a climate change, EnvironmentalHealth Perspectives, 105(1),84-93,1997.

Kalkstein,L.,and K.Smoyer,The impacts of climatechange on human health: Some international implica-tions, Experientia, 49, 969-979,1993.

Karl,T. R., Potential application of model output statis-tics (MOS) to forecasts of surface ozone concentra-tions, Journal of Applied Meteorology, 18,254-265,1979

Karl,T. R.,and R.W. Knight,Secular trends of precipita-tion amount,frequency, and intensity in the USA,Bulletin American Meteorological Society, 79, 231-241,1998.

Kunkel,K.E.,R.A.Pielke,Jr.,and S.A.Changnon,Temporal fluctuations in weather and climate extremesthat cause economic and human health impacts: Areview, Bulletin of the American MeteorologicalSociety, 80, 1077-1098,1999.

Kunkel,K.E., N.Westcott,and D. Kristovich,Assessmentof potential effects of climate change on heavy lake-effect snowstorms near Lake Erie,preprints, 11thSymposium on Global Change Studies, Long Beach,California, January 10-13,2000,AmericanMeteorological Society, Boston,Massachusetts,pp.50-53,2000.

Lal,R., J. M.Kimble,R. F. Follett,and C.V. Cole,ThePotential of US Cropland to Sequester Carbon andMitigate the Greenhouse Effect,Sleeping Bear Press.Ann Arbor, Michigan,1998.

Lew,A.A.,The Midwest: The Nation’s Heartland,inGeography USA:A Virtual Textbook,(http://www/for.nau.edu/~alew/midwest7.html).,1998.

Magnuson, J. J., Potential effects of climate changes onaquatic systems: Laurentian great lakes andPrecambrian shield region, chap.2,in Assessment, edit-ed by C.E.Cushing, John Wiley & Sons,Chichester,England,pp.7-53,1997.[Papers in this volume were

Chao, P.,Great Lakes water resources: Climate changeimpact analysis with transient GCM scenarios, Journalof American Water Resources Association, 35, 1499-1507,1999.

Coffin D. P.,and W. K.Lauenroth,Transient responses ofNorth American grasslands to changes in climate,Climatic Change, 34,269-278,1996.

EPA,Climate Change and Wisconsin (230-F-97-008ww), Climate Change and Michigan (230-F-97-008 v),Climate Change and Minnesota (230-F-97-008 w),1997. available at http://www.epa.gov/globalwarm-ing/publications/impacts/state/index.html

EPA,Ecological Impacts from Climate Change: AnEconomic Analysis of Freshwater RecreationalFishing (220-R-95-004),1995. available athttp://www.epa.gov/globalwarming/publications/

Fang,X.,and H. G. Stefan, Potential climate warmingeffects on ice covers of small lakes in the contiguousUS, Cold Regions Science and Technology, 27, 119-140,1998.

Garrity, C.M.,and J.A.Andresen,Climatological con-straints of a double-cropping system in the UpperGreat Lakes region: An assessment of risk and poten-tial,paper presented at 1999 Annual Workshop of theUS National Assessment: The Potential Consequencesof Climate Variability and Change,Atlanta,Georgia,April12-15,1999.

Groisman, P.Ya., et al., Changes in the probability ofheavy precipitation:Important indicators of climaticchange, Climate Change, 42,243-283,1999.

Groisman, P.Ya.,R.W. Knight,T. R.Karl.Relationsbetween changes of heavy precipitation and stream-flow in the United States,in press, Bulletin of theAmerican Meteorological Society, 2000.

Injerd, D., Panel Presentation,Impacts and risks of cli-mate change and variability:stakeholder perspectivesin Adapting to Climate Change and Variability in theGreat Lakes-St.Lawrence Basin: Proceedings of aBinational Symposium, edited by L. D. Mortsch,S.Quon,L.Craig,B.Mills,and B.Wrenn,EnvironmentCanada,Downsview, Ontario,1998.

IPCC (Intergovernmental Panel of Climate Change),Climate Change 1995: The Science of ClimateChange, edited by J.Y. Houghton, F. G. Meira Filho,B.A.Callander, N. Harris,A.Kattenberg,and K.Maskell,Cambridge University Press,Cambridge,UnitedKingdom,570 pp.,1996.

186


originally published in Hydrological Processes, 11(8),819-1067,1997.

McPherson,E. G., D. Nowak, G. Heisler, S.Grimmond,C.Souch,R.Grant,and R.Rowntree,Quantifying urbanforest structure,function,and value: The ChicagoUrban Forest Climate Project, Urban Ecosystems, 1(1),49-61,1997.

Meehl, G.A., F. Zwiers, J. Evans,T. Knutson,L.Mearns,and P.Whetton,Trends in extreme weather and climateevents: Issues related to modeling extremes in projec-tions of future climate change, Bulletin of theAmerican Meteorological Society, 81, V. 81,No.3 pp.427-436.

Moll,R.,Reports of the working groups: Ecosystemshealth in Adapting to Climate Change and Variabilityin the Great Lakes-St.Lawrence Basin: Proceedings ofa Binational Symposium, edited by L. D. Mortsch,S.Quon,L.Craig,B.Mills,and B.Wrenn,EnvironmentCanada,Downsview, Ontario 1998.

Mortsch,L. D.,Assessing the impact of climate changeon the Great Lakes shoreline wetlands, ClimaticChange, 40, 391-416,1998.

National Climatic Data Center (NCDC),Billion DollarUS Weather Disasters 1980-1999,Asheville,NorthCarolina,US Department of Commerce,NOAA,Website,(http://www.ncdc.noaa.gov/ol/reports/billionz.html),1999.

Norgaard,K. J.,Impacts of the Subdivision Control Actof 1967 on land fragmentation in Michigan’s town-ships,Ph.D. dissertation, Michigan State University, EastLansing,Michigan,1994.

NPA Data Services,1999a, Demographic Databases,Household Databases Three Growth Projections 1967-2050, 1424 16th Street,NW, Suite 700,Washington,DC20036

NPA Data Services,1999b, Economic Databases ThreeGrowth Projections 1967-2050, 1424 16th Street,NW,Suite 700,Washington,DC 20036

Olsen, J., J. Stedinger, N. Matalas,and E.Stakhiv, Climatevariability and flood frequency estimation for theupper Mississippi and lower Missouri rivers, Journal ofAmerican Water Resources Association, 35, 1509-1523,1999.

Quinn, F.H.,B.M.Lofgren,A.H.Clites,R.A.Assel,A.Eberhardt,and T. Hunter, Great Lakes water resourcessector report,Great Lakes Environmental ResearchLaboratory,Ann Arbor, Michigan,6 pp.,1999.

Robinson,S.K.,The case of the missing songbirds,Consequences, 3, 2-15,1997.

Rose, J. B.,S.Daeschner, D. R.Easterling, F. C.Curriero,S.Lele,and J.A. Patz, . Climate and waterborne outbreaksin the US, Journal of American Water WorksAssociation,, 77-87, Sept.2000.

Sadof, C.,and M.Raupp,Biological controls in theurban forest, Midwest Biol Cont.News, (5)12,1-5,1999.

Secrest,H.C.,H. J. Mac Aloney, and R.C.Lorenz,Causesof the decadence of hemlock at the Menominee IndianReservation,Wisconsin, Journal of Forestry, 39, 3-12,1941.

Smith, J. B.,Setting priorities for adapting to climatechange, Global Environmental Change, 7(3),251-264,1997.

Stefan,H.G.,M.Hondzo,and X. Fang,Lake water qualitymodeling for projected future climate scenarios,Journal of Environmental Quality, 22,417-431,1993.

Wang, J. X.L.,and J. K.Angell,Air stagnation climatologyfor the United States (1948-1998),NOAA/Air ResourcesLaboratory ATLAS No.1,Silver Spring,Maryland,73 pp.,1999.

Warbach, J. D.,and D. Norberg, Michigan Society ofPlanning Officials Trend Futures Project:Public Landsand Forestry Trends Working Paper, Michigan Societyof Planning Officials,Rochester, Michigan,1995.

Wargo, P.M., D. R.Houston,L.A.LaMadeleine,Oakdecline, Forest Insect & Disease Leaflet 165, USDepartment of Agriculture, Forest Service,8 pp.,1983.

Wolfe, D.W.,Potential climate change impacts on NewEngland agriculture,in Workshop Summary Report,New England Regional Climate Change ImpactsWorkshop, Institute for the Study of Earth,Oceans,andSpace,University of New Hampshire,Durham,NewHampshire,1997.

Wong,K.,A pixel is worth 1000 words:Satellite imagesreveal startling tree loss in American cities, US Newsand World Report, July 19,1999.

187


John T. Lehman,University of MichiganJohn D. Lindeberg,Center for Environmental Studies,

Economics & ScienceBrent M.Lofgren,Great Lakes Environmental

Research LaboratoryJames R.Nicholas,USGS,Lansing,MichiganJamie A.Picardy,Michigan State UniversityJeff Price,American Bird ConservancyFrank H.Quinn,Great Lakes Environmental Research

LaboratoryPaul Richards,University of MichiganJoe Ritchie,Michigan State UniversityTerry Root,University of MichiganWilliam B.Sea,University of MinnesotaDavid Stead,Center for Environmental Studies,

Economics & ScienceShinya Sugita,University of MinnesotaKaren Walker, University of MinnesotaEleanor A.Waller, Michigan State UniversityNancy E.Westcott,Illinois State Water SurveyMark Wilson,University of MichiganJulie A.Winkler, Michigan State UniversityJohn Zastrow, University of Wisconsin

Additional ContributorsStanley Changnon,Illinois State Water SurveyByron Gleason,National Climatic Data Center

* signifies Assessment team chairs/co-chairs

ACKNOWLEDGMENTS

Many of the materials for this chapterare based on contributions from participants on and those working with the

Eastern Midwest Workshop TeamJ. C.Randolph,Indiana UniversityOtto Doering,Purdue UniversityMike Mazzocco,University of Illinois,Urbana -

ChampaignBecky Snedegar, Indiana University

Great Lakes Workshop and Assessment TeamPeter J. Sousounis*,University of MichiganJeanne Bisanz*,University of MichiganGopal Alagarswamy, Michigan State UniversityGeorge M.Albercook,University of MichiganJ. David Allan,University of MichiganJeffrey A.Andresen,Michigan State UniversityRaymond A.Assel,Great Lakes Environmental

Research LaboratoryArthur S.Brooks,University of Wisconsin-Milwaukee,

WisconsinMichael Barlage,University of MichiganDaniel G.Brown,Michigan State UniversityH.H.Cheng,University of MinnesotaAnne H.Clites,Great Lakes Environmental Research

LaboratoryThomas E.Croley II,Great Lakes Environmental

Research LaboratoryMargaret Davis,University of MinnesotaAnthony J. Eberhardt,Buffalo District,Army Corps of

EngineersEmily K.Grover, University of MichiganGalina Guentchev, Michigan State UniversityVilan Hung,University of MichiganKenneth E. Kunkel,Illinois State Water SurveyDavid A.R.Kirstovich,Illinois State Water Survey

188


191

CHAPTER 7

POTENTIAL CONSEQUENCES OF CLIMATE VARIABILITY AND CHANGE FOR THE GREAT PLAINSLinda A. Joyce1,2

, Dennis Ojima3, George A.Seielstad4

, Robert Harriss5, and Jill Lackett3

Contents of this ChapterChapter Summary



Ecological Context


Key Issues

Changes in Timing and Quantity of Water

Weather Extremes

Invasive Species and Biodiversity

Quality of Rural Life on the Great Plains

Additional Issues


Literature Cited

Acknowledgments

1Coordinating author for the National Assessment Synthesis Team; 2USDA Forest Service, 3Colorado State University,4University of North Dakota, 5National Center for Atmospheric Research


Over the 20th century, temperatures in the Northernand Central Great Plains have risen more than 2˚F(1˚C),with increases up to 5.5˚F (3˚C) in someareas. There is no evidence of a trend in the histori-cal temperature record of the Southern Great Plains.Over the last 100 years,annual precipitation hasdecreased by 10% in the lee of the RockyMountains. Texas has seen significantly more highintensity rainfall.


Air temperatures will likely continue to risethroughout the region,with the largest increases inthe northern and western parts of the Plains.Seasonally, more warming is projected to occur inwinter and spring than in summer and fall.

A pattern of decreasing precipitation appears likelyin the lee of the Rocky Mountains while other sec-tions of the Great Plains may experience slightincreases. Although precipitation increases are pro-jected for parts of the Great Plains,increased evapo-ration due to rising air temperatures are projectedto surpass these increases, resulting in net soil mois-ture declines for large parts of the region.


The Great Plains produces much of the nation’sgrain,meat,and fiber, and in addition provides recre-ation,wildlife habitat,and water resources. Thoughmore rural than the rest of the United States,theurban areas of the Great Plains provide housing andjobs for two-thirds of the people of the Great Plains.Soil organic matter is a major resource of the GreatPlains as it provides improved soil water retention,soil fertility, and the long-term storage of carbon.

192


Key Findings

• Productivity of crops and grasses of the regionwill likely respond positively to additional atmos-pheric carbon dioxide,especially those systemswith adequate water and nitrogen such as alfalfaor soybeans.

• The warmer and predicted longer growing sea-sons will likely change the life cycles of all bio-logical organisms and these changes will haveprofound impacts on the ecology of the GreatPlains native ecosystems.

• The projected climate changes will likely alterthe current biodiversity, resulting in a new com-position of plant and animal species that may ormay not be detrimental to society. A possiblemigration of invasive species across the GreatPlains is a concern to stakeholders in the regionbecause the rapid rate of change in climate maybe disadvantageous to native species.

• Extreme temperatures and heat stress events,where the temperature remains over 90˚F(32˚C) for 3 consecutive days,will likely increasein the Southern Great Plains. These events willincrease the heat stress on humans and livestock.

• Changes in the demand for irrigation water varyby crop type and the changes in the seasonality

of precipitation. For example,in northeasternColorado the consumptive demand for water forperennial crops such as grass and alfalfa was esti-mated to increase at least 50% over current use,however, consumptive water use for corn wasnot expected to increase substantially.

• Increased air temperatures may reduce soilorganic matter affecting soil fertility, water-hold-ing capacity, and the storage of carbon in thesoil.

• The intensity of rainfall events may increase inthe Southern Great Plains, resulting in more rain-fall in shorter periods of time,with implicationsto urban and rural flood controls and soil ero-sion.

• Rural communities,already stressed by theirdeclining populations and shrinking economicbases,are dependent on the competitive advan-tage of their agricultural products in domesticand foreign markets. A changing climate willlikely bring additional stresses that will dispro-portionately impact family farmers and ranchers.

• Stakeholders in the region thought that commu-nity-based adaptive management was an impor-tant component for future planning.

193

Chapter 7 / The Great Plains


194

PHYSICAL SETTING ANDUNIQUE ATTRIBUTESThe Great Plains is often pictured as an agriculturallandscape dotted with many small rural towns. Infact,the Great Plains is more rural than the rest ofthe US;nearly 40% of the region’s counties havepopulations less than 2,500 people in contrast tothe 20% of rural counties in the remainder of the US(Skold,1997;Gutmann,1999). These landscapes ofgrasses, forbs,and shrubs,interspersed with a vari-ety of crops, give the Plains a sense of wide-openspace. Increasing sizes and declining numbers offarms and ranches, fewer rural trade centers,anddeclining rural populations are recent socio-

economic trends that further contribute to theremoteness of some rural counties. In contrast,other Great Plains’counties with large urban centersor with scenic amenities are experiencing popula-tion increases and economic growth (Drabenstottand Smith,1996).

Distributions of the naturally occurring vegetationand the planted agricultural crops (Figure 1) arestrongly linked to the gradients of temperature(north to south) and precipitation (west to east)within the Great Plains (Figure 2). Cool-seasongrasslands in the North give way to warm-seasongrasslands in the central and southern parts of theregion,which give way to drought-adapted shrubsin the southwestern parts and trees in the southeast-ern parts. As precipitation increases from west toeast across the Great Plains,the native vegetationincludes more mixed-grass and tall-grass species,andfinally tree species. This precipitation gradient influ-ences the use of irrigation in agriculture with ahigher percentage of crops grown under irrigationin the western Great Plains and more crops depend-ent on growing season precipitation in the easternGreat Plains. The extreme western Great Plains isdominated by dryland cropping because of thereduced availability of water for ir rigation. Croptypes tend to follow the temperature gradient,withcool-season grains such as wheat,barley and oatsdominating in the north and warm-season cropssuch as corn,sorghum,sugar beets,and cotton dom-inating in the south. However, the ability of the agri-cultural sector to expand across climatic zones isseen in that both corn and wheat can be foundfrom North Dakota to Texas. Both natural andhuman systems cope with the natural variability inclimate that characterizes the Great Plains. Periodsof drought,a result of variability in climate,heightenthe importance of water in this region.

Though dominated by grasslands,the Great Plains isalso home to a diversity of plants and animals inshrublands, wetlands, woodlands,and forest commu-nities. Riparian vegetation including deciduousforests,woodlands,and shrublands trace theMississippi,Missouri,Platte,Kansas,Arkansas,Colorado,and Rio Grande Rivers. Juniper wood-lands and conifer forests are found on the escarp-ments of the South Dakota badlands. Wind-deposit-ed material forms extensive sand dune systems in

POTENTIAL CONSEQUENCES OF CLIMATEVARIABILITY AND CHANGE ON THE GREAT PLAINS

Great Plains Vegetation Map

Figure 1. Distributions of the naturally occurring vegetation and thecurrent planted agricultural crops are strongly linked to the gradi-ents of temperature (north to south) and precipitation (west to east)within the Great Plains. Outlines show the federal land holdings inthe region (vegetation map from Natural Resource Ecology Lab,Colorado State University. Potential natural vegetation according toVEMAP members, 1995). See Color Plate Appendix


195

of soil fertility. Soil organic matter in grasslandecosystems is estimated to be an important reser-voir of terrestrial carbon (Anderson,1991). Forexample,the average aboveground plant biomassproduction for a cool-season grassland in Havre,MTis 33.9 g m -2. For a short-grass steppe ecosystem atthe Central Plains Experimental Range in Colorado,aboveground plant biomass production is 45.9 g m-2

(Haas et al.,1957;Cole et al.,1990). Soil C valuesare 3230 and 2310 g m-2, for Havre and the CentralPlains Experimental Range, respectively. The conver-sion of Great Plains grasslands to croplands hasresulted in major changes in soil organic matter andnutrient supplying capacity on these lands (Haas etal.,1957;Tiessen et al.,1982;Burke et al.,1990).

SOCIOECONOMIC CONTEXTMore than 70% of the Great Plains landscape (Figure1) is used to produce a large proportion of thenation’s food. Over 60% of the nation’s wheat isproduced in Montana,North Dakota,South Dakota,Nebraska,Kansas,Oklahoma,and Texas (Skold,1997). The states of Texas,Oklahoma,New Mexico,Nebraska,Kansas,and Colorado produce 87% of thenation’s grain sorghum. Over 54% of the nation’sbarley and 36% of the nation’s cotton are producedin the region (Skold,1997). Livestock in the GreatPlains constitutes over 60% of the nation’s total,including both grazing and grain-fed cattle opera-tions. Nearly 75% of the grain-fed cattle in the USare from the Great Plains,using the readily availablesupply of feed grains, over 50% of which is also pro-duced in the region. Great Plains’ farmers andranchers have excelled by being adaptive and byincorporating new technologies to buffer their pro-duction against the highly variable climate. Onetenth of Great Plains’cropland is irrigated. New

the Great Plains. The Nebraska Sandhills contain thelargest dune system covered by grassland in theUnited States (Ostlie et al.,1997). The rich grass-lands of the region have been the basis of a largegrazing system for thousands of years and currentlysupport a diversity of cattle,bison,and other mam-mals,as well as a diversity of insect and birdspecies. Nearly 60% of the bird species that breedwithin the US are considered regular breeders with-in the Great Plains (Ostlie et al.,1997). Endemicplants and animals are found throughout uniquehabitats in the Great Plains. The wetland basins inthe Prairie Pothole region of the Northern GreatPlains and the playa lakes of the Central andSouthern Great Plains are important habitat andbreeding grounds for migratory waterfowl. Fishhabitats include large streams with erratically vari-able flow, prairie ponds,marshes and small streams,and residual pools of highly intermittent streams.The largest and most diverse class of animals in theGreat Plains is insects (Ostlie et al.,1997).

Throughout this region,soil organic matter is amajor resource as it provides improved soil waterretention,soil fertility, and the long-term storage ofcarbon. Soil organic matter integrates climate, geol-ogy, topography, and ecosystem dynamics (Jenny,1980). The historical levels of grassland productivityand soil organic matter varied widely from north tosouth across the Great Plains, reflecting rainfall andtemperature gradients (Jenny, 1980; Parton et al.,1987;Schimel et al.,1990; Peterson and Cole,1995).Ecosystem processes associated with productivityand decomposition influence the net changes in soilcarbon (C) and these net changes serve as an index

Figure 2. Great Plains climate is characterized by a strong north-south temperature gradient and a strong east-west precipitationgradient (averages based on the 1961-1990 period; data fromDennis Ojima, VEMAP climate). See Color Plate Appendix

Great Plains Climate

When the Great Plains was being settled over 100years ago,county seats were established about 30miles apart in the eastern portion and somewhat far-ther apart in the central and western parts of theregion (Drabenstott and Smith,1996). These countyseats became the centers for government,com-merce,and finance. Consolidation in agriculture,expansion of telecommunications,improved trans-portation,and discount retailing in rural areas aretrends that have significantly reduced the number ofviable rural economic centers (Drabenstott andSmith,1996),placing an increasing burden on ruralgovernments to provide health care and educationwith dwindling financial resources. In contrast,those rural counties with large trade centers or sce-nic attractions,as in western Montana and Colorado,have seen growth in their local economies.

Agriculture,nationally and regionally, is becoming afarm-to-grocery integrated industry in response toconsumer’s demands for conveniently prepared andhighly nutritious food products (Barkema andDrabenstott,1996). While the total area in farmlandhas remained fairly stable in the Great Plains, farmsare getting bigger and the number of farmers fewer.Within the Central Great Plains,the total number offarms has gone from nearly 200,000 in 1930 to lessthan 100,000 in 1990 and big farms have increasedfrom 10% to over 30% of the total number of farms(Gutmann,1999). More crops are grown under con-tracts with rigid production guidelines. Contract

technologies in agriculture,crop genetics,and live-stock production have facilitated the expansion anddiversification of Great Plains agriculture.

However, Great Plains’communities are undergoingdramatic changes as the industries of agriculture,commerce,and finance consolidate (Drabenstottand Smith,1996;Barkema and Drabenstott,1996).

196


Multiple Stresses on Urban and Natural Environments

The Rio Grande River is life itself to cities,industries,wildlife,and rare vegetation on both sides of the US-Mexico border. However, environmental stresses and socioeconomic changes are overwhelming the urbaninfrastructure in this area of the Southern Great Plains. The Rio Grande supplies water to the rapidly expand-ing human population as well as the rapidly expanding manufacturing facilities. Total population in citiesalong the US-Mexico border is projected to double in less than 30 years. More than 60% of all US-Mexicotrade passes through Laredo,Texas in trucks,making it the largest land port in the US. Mexico will soon dis-place Japan as the number two trading partner of the US,despite the fact that Mexico’s economy is one-twelfth the size of Japan’s.

The rapid increase in trade flowing through border cities, following passage of the North American Free TradeAgreement (NAFTA),has resulted in a complex array of costs and benefits to border cities. Industrial pointsource pollution as well as automobile emissions are a significant problem in the industrialized areas of theLower Rio Grande Valley and Northern Mexico. In the unincorporated shanty-towns along both sides of theRio Grande border, infrastructure for water supply and wastewater treatment is generally absent or totallyinadequate. Reported cases of hepatitis and other viral diseases are typically two times greater in bordercounties compared to the statewide average.

The state of human health and environmental quality are very closely tied to land use, climate variability, airquality, and water supply. Climate change is likely to exacerbate these multiple stresses on human and naturalsystems;both by changes to local weather and to the flow regime of the river.

Great Plains Agicultural Exports

Figure 3. Agricultural exports are an important percentage of thetotal agricultural production within each state. (USDA, 1997 Censusof Agriculture). See Color Plate Appendix

production varies across crop types,but even thesmall proportion of contracted wheat and feed grainis rising. Cattle feeders without contracts reportthat their markets have shrunk (Barkema andDrabenstott,1996).

One consequence of changing agricultural econom-ics is that larger farms’ volume of business often jus-tifies searching greater distances to seek lowerprices on purchased items or higher prices on itemssold (Barkema and Drabenstott,1996). Local mar-kets decline as the large rural centers or urban areasattract consumers.

Great Plains’ agriculture has also been limited bytwo other global trends. First,the largest increase inthe agricultural market is in food products that areready or nearly ready for human consumption,suchas cereals and snack foods,meats,fruits,and vegeta-bles. The main agricultural crops of the Great Plainsare bulk commodities such as wheat and corn. Withthe exception of meat processing,minimal food pro-cessing is done within the Great Plains because ofprohibitive transportation costs to reach the majormarkets. Secondly, the international markets for USagricultural production have shifted from the grainpurchases of the Soviet Union and the EuropeanUnion to the more diversified markets of Asia. In1996, over 40% of US agricultural exports went toAsian markets (Barkema and Drabenstott,1996).From 4 to 38% of each Great Plains state’s crop pro-duction — grain and livestock — is exported out-side of the US (Figure 3). Thus,Great Plains agricul-ture is highly sensitive to changing consumer prefer-ences and the global economy.

Although agriculture dominates land use in theGreat Plains,the percent share of agriculture issmall,2% of the 1997 gross state product of all GreatPlains states (US Department of Commerce 1998).The Northern Great Plains states are more depend-ent on agriculture than the Central Great Plainsstates,which are more dependent on agriculturethan the Southern Great Plains states (Figure 4).Agriculture, forestry, and fisheries comprise 11% ofNorth Dakota’s gross state product,in contrast to 1%in Texas. In 1996,North Dakota,South Dakota,andNebraska ranked one,two,and three nationally interms of the farm share of the gross state product.

Changes in agricultural economies and lifestyles inthe Great Plains are altering demographic patternsof the region. Populations are declining in mostrural areas. In North Dakota, over 65% of the coun-ties have fewer than 6 residents per square mile.The average age of a farm operator in the Central

Great Plains is nearly 54 years,52 years in theNorthern Great Plains,and over 56 years in theSouthern Great Plains (USDA Census of Agriculture,1997). Urban population in the Central Great Plainshas gone from 27% of the total population in 1930to 67% in 1990 (Gutmann,1999). Some 80% of thepeople in Texas live in cities. This shift in popula-tion increases the demand for services in the urbanareas and introduces urban problems such as deteri-orating air quality. In addition, agricultural to munic-ipal water transfers are becoming increasingly com-mon as more water is needed for urban inhabitants(NRC,1992).

ECOLOGICAL CONTEXTWater and temperature strongly influence the struc-ture and function of the Great Plains grasslands. Thetransition from short-grass steppe in the lee of theRocky Mountains to mixed prairie and finally tall-grass prairie at the eastern edge of the Great Plainscorresponds to a precipitation gradient of low rain-fall east of the Rockies to relatively high and moreevenly-distributed rainfall in eastern parts of theregion. The temperature gradient from north tosouth in the Great Plains represents another impor-tant gradient that determines plant type distributionand local abundance (e.g., warm-season versus cool-season grasses).

197


Agricultural Land Comparisons

Figure 4. The Northern Great Plains are more dependent on agricul-ture than the Central which is more dependent on agriculture thanthe Southern Great Plains, yet agriculture dominates land use in allregions of the Great Plains. (Economic data from US Dept ofCommerce, Bureau of Economic Analysis, Regional EconomicAnalysis Division, June 1998, and land use data from USDA, 1997Census of Agriculture.) See Color Plate Appendix


198

Organic matter in Great Plains’soils also varies withtemperature and precipitation gradients (Jenny,1980; Parton et al.,1987;Schimel et al.,1990;Peterson and Cole,1995). The net changes in soil Cintegrate ecosystem processes associated with pro-ductivity and decomposition rates that change withweather and with land management. Changes insoil organic matter ser ve as an index of soil fertility.Soil carbon storage is the largest carbon pool in thegrassland ecosystems. Soil carbon storage and flux-es are influenced by vegetation, characteristic soilproperties,inherent climate regime,and land usepractices. The current carbon level is not onlydetermined by the current state of these factors butis also dependent on the land use history. Changesin rates of decomposition,plant production,andrespiration from climate and land managementaffect the storage and flux of carbon from the soil.Rainfall amounts tend to control the amount ofplant production that occurs whereas temperaturedetermines decomposition rates. However, bothrainfall and temperature interact in both processesto a certain degree. Changes in soil carbon areresponsive to agricultural practices that alter theamount of plant material entering the soil. Changesin tillage practices, grazing patterns,and manure dis-tribution all affect the storage of soil carbon.Conversion of grasslands or forests to cropland canresult in a rapid decline in carbon stores. Up to50% of the soil carbon and the woody biomass ofthe forest can be lost due to cropland conversion(Haas et al.,1957;Cole et al.,1989,1990).Ecosystems are often sensitive to changes inextreme events. Changes in low or high tempera-ture extremes or in drought occurrence can rapidlychange the structure of an ecosystem,especiallywhen these modifications in extremes coincidewith a disturbance event,such as fire,that resetsthe ecosystem to a different vegetation composi-tion. In these instances,large fluxes of carbon andlong-term changes in carbon storage may result.

Three environmental parameters of global change— increased carbon dioxide,increased tempera-ture,and altered precipitation — will likely affectGreat Plains ecosystems primarily through theircombined effects on plant and soil water interac-tions,photosynthesis,and other aspects of plantmetabolism. Responses will result from directeffects of the changing environment on individualplants (e.g.,productivity),as well as from changesin the mix of plant communities and cropping sys-tems that occur due to different sensitivities of indi-vidual species. In order to assess these changes atthe regional scale,the ecosystem model CENTURY(Parton et al.,1987,1994) was used to simulate

plant productivity and carbon storage in natural andin agricultural ecosystems.

Both of the two climate scenarios used in thisassessment have a marked increase in temperatures,with the Canadian scenario being the warmer of thetwo scenarios (see Climate chapter). Both scenariosare associated with increases in precipitation,how-ever, the regional pattern and the magnitude of theincrease in precipitation is highly varied. Of thetwo,the Hadley scenario simulates a slightly wettergrowing season compared to the Canadian scenario.The timing of precipitation,seasonal changes intemperature,and the increased atmospheric carbondioxide concentrations determine the impact of thetwo climate scenarios on the plant productivity andsoil carbon levels of the Great Plains.

In order to capture the impact of land use practiceson carbon storage and fluxes,the dominant naturalor managed vegetation type was assigned to individ-ual grid cells across the region. The CENTURYmodel simulated net carbon storage in these gridcells for each climate scenario. These simulationswere based on current dominant agricultural prac-tices and do not represent adaptive agriculturalmanagement practices being tested for maximumcarbon storage. Recent no-till or reduced-till prac -tices have been developed to lessen the impact onthe soil carbon losses from croplands. The use ofadaptive cropping management strategies in the USis estimated to account for 0.08-0.2 PgC/y (Lal et al.,1998),although these cropland areas are vulnerableto large carbon losses due to removal of vegetationand soil disturbances on an annual basis. The histor-ical land management usage greatly determines thecurrent carbon storage and potential fluxes fromthese managed systems.

For both climate scenarios,the increased atmospher-ic carbon dioxide associated with the climatechanges partially ameliorates the negative warmingtrends on crop and grassland productivity (Figure5). Productivity of most crops and grasses respondspositively to additional carbon dioxide,especiallythose with adequate nitrogen such as the nitrogen-fixing alfalfa or soybean systems. These croppingsystems,when simulated under dryland manage-ment,perform much better with increased carbondioxide since these cropping systems are able touse the available soil moisture more effectively thanunder conditions in the absence of elevated carbondioxide.

Soil organic matter responds to changes in vegeta-tion as well as atmospheric carbon dioxide levels

Soil Carbon


199

tant metabolic responses that bear on forage quality,including reduced crude protein content andincreased total non-structural carbohydrates. Thisdirect effect on forage quality may be overshadowedby quality changes due to altered plant communi-ties,such as changes in the balance of dominantwarm- and cool-season grasses,more legumes,ormore shrubs. The warmer and predicted longergrowing seasons will likely alter life cycles of all bio-logical organisms and these changes will likely haveprofound impacts on the ecology of the Great Plainsnative ecosystems.

(Figure 5). Where moisture levels and productivitydecline,soil carbon may actually increase as decom-position processes become limiting as seen alongthe lee of the Rocky Mountains and especially in theSouthern Great Plains. The southeastern parts of theGreat Plains lose soil organic matter due to theincreased soil moisture levels resulting fromincreased water use efficiency associated with theelevated carbon dioxide levels. In the dryland sys -tems with little excess crop residue,soil organicmatter continually declines over the duration of theclimate change scenario. Systems such as wheat-fal-low continue to mine soil carbon as temperatureand moisture conditions facilitate release of carbonby microorganisms during the fallow period. In sys-tems where excess crop residue is available andincorporated into the soil system,there tends to bea build up of soil carbon. This is especially evidentin the productive irrigated soybean/corn rotationswhere nitrogen fixation by the beans and the addi-tional residue input from corn rapidly builds soilcarbon.

Some changes, such as enhanced forage productionin response to elevated carbon dioxide,may be ben-eficial. However, plant growth under elevated car-bon dioxide conditions directly influences impor-

Figure 5. The productivity of the Great Plains increases from westto east and from north to south, following the precipitation and thetemperature gradients. Land uses are strongly influenced by pro-ductivity. Both climate scenarios increase the moisture stress onthe central parts of the Great Plains and productivity declines inthis region. Soil organic matter in the Great Plains is an importantreservoir of terrestrial carbon. The amount of carbon stored in thesoil is strongly influenced by past and present land managementpractices and weather patterns. Where moisture levels and produc-tivity decline, soil carbon may actually increase as decompositionprocesses become limiting. Where soil moisture levels increasefrom increased water use efficiency, soil carbon levels may decline.(CENTURYresults from VEMAP analysis, Natural Resource EcologyLab, Colorado State University.) – See Color Plate Appendix

Net Primary Productivity (NPP)1961 - 1990

0-100100-200200-300300-400400-500500-600600-700700-800800-900900-1000>1000No Data

NPP (g/sq m)

<-100-100- -40-40- -20-20- -00-2020-4040-6060-8080-100100-150>150No Data

Difference (g/sq m)

Canadian Model Difference from 1961-90 by 2100

Hadley Model Difference from 1961-90 by 2100

Soil Carbon

800-16001600-24002400-32003200-40004000-48004800-56005600-64006400-72007200-80008000-88008800-9600No Data

Soil C (g/sq m) <-1500-1500- -500-500- -250-250- -200-200- -150-150- -100-100- -50-50- 00-5050-100100-500>500No Data

Soil C (g/sq m)

Canadian Model Difference from 1961-90 by 2100 1961 - 1990

Hadley Model Difference from 1961-90 by 2100

CLIMATE VARIABILITY AND CHANGEThe Great Plains climate is characterized by a strongnorth-south temperature gradient and a strong east-west precipitation gradient (Figure 2). Annual pre-cipitation ranges from less than 7.8 inches (200mm) on the western edge to over 43 inches (1,100mm) on the eastern edge of the Great Plains.Average annual temperature is less than 39˚F (4˚C)in the Northern Great Plains and exceeds 72˚F(22˚C) in the Southern Great Plains.

The spring and summer peaks in precipitation pro-vide growing season moisture for the prairies. Forexample,75% of the annual precipitation falls dur-ing the growing season at the Konza Prairie,Kansas(Hayden,1998). Growing seasons range from 110days in the Northern Great Plains to 300 days in theSouthern Great Plains (Donofrio and Ojima,1997).Monthly average maximum temperatures exceed91˚F (33˚C) in most places on the Great Plains dur-ing one of the summer months. Three consecutivedays of temperature over 90˚F (32˚C) can signalheat stress for both humans and livestock.

The variability of weather in the Great Plains is acharacterizing feature,as “normal” years are rare and

200


extremes are most often the common experience.Blizzards, floods,droughts,tornadoes,hail storms,thunderstorms,high winds,severe cold,andextreme heat often arrive suddenly, disrupt normaldaily activities,and can be life-threatening.

Analysis of weather data for the last 100 years in theGreat Plains indicates a warming pattern in theNorthern and Central Great Plains,but no evidenceof a trend in the historical temperature record forthe Southern Great Plains (see Climate Chapter).Temperatures have risen in the Northern andCentral Great Plains by about 2˚F (1˚C) in the 20th

century (Karl et al.,1999),with increases of 5˚F(3˚C) in Montana,North Dakota,and South Dakota.This warming trend in the Northern Great Plains isreflected in 6 fewer days with temperatures lessthan 32˚F (0˚C). Meteorologists use indices,such asheating degree days and cooling degree days,todescribe likely changes that would be required tomaintain the human quality of life — additionalheating needed in the winter or cooling in the sum-mer. Heating Degree Days (HDD) are the degreesthe average daily temperature is below 65˚F (18˚C).Cooling Degree Days (CDD) are the degrees theaverage daily temperature is above 65˚F (18˚C). TheNorthern Great Plains has had significantly fewerheating degree days (-628 HDD) and significantlymore cooling degree days (40 CDD) over the 20th

Temperature Change - 20th & 21st Centuries

Precipitation Change - 20th & 21st Centuries

Observed 20th Canadian ModelThe Canadian model projectsdecreases in precipitation inthe southern Plains andincreases in the north. TheHadley model projectsincreases over almost theentire region, but somedecreases are also evidenteast of the Rockies.

Figure 7.Precipitation aver-ages over the lastcentury indicate adecrease in precipi-tation to the east ofthe Rockies.Several areas, mostnotably Texas, haveprecipitationincreases (data thisassessment).

100%

75%

50%

25%

0

-25%

-50%

-75%

-100%

100%80%60%40%20%0-20%-40%-60%-80%-100%

Hadley Model

Observed 20th Canadian ModelThe model scenarios indi-cate an additional 5˚F(Hadley scenario) to 12˚F(Canadian scenario) increas-es over the next century formuch of the Plains (data thisassessment).

Figure 6. Theobserved changesin air temperaturefor the Great Plainsover the last centu-ry indicates agreater warming inthe north than inthe south on aver-age.

Hadley Model+15ºF

+10ºF

+5ºF

0º

-5ºF

+15ºF

+10ºF

+5ºF

0º

-5ºF

See Color Plate Appendix

century (Chapter 1). Daily minimum temperaturesthroughout the year have increased more than maxi-mum temperatures,indicating greater nighttimewarming. In the Northern Great Plains, over the last50 years the mean date of the last measurable snow(greater than 1 inch on the ground) has occurred 4days earlier. This significant change indicates awarming trend in the winter-spring months.Over the last 100 years,annual precipitation hasdecreased by 10% in eastern Montana,easternWyoming, western and central North Dakota,andColorado and has increased by 5 to 20% in SouthDakota,Oklahoma,Texas,and parts of Kansas (Karlet al.,1999). Texas has seen an increase in highintensity precipitation events,with significantlymore area reported in severe wetness and signifi-cantly less area reported in drought conditions overthe last 100 years (see Chapter 1).

The two climate scenarios used in this Assessment(see descriptions in Climate Chapter) project a con-tinuation of the trends seen in Great Plains histori-cal climate:higher temperatures,and for some areas,greater precipitation. The Canadian scenario proj-ects greater increases in temperature than theHadley scenario (Figure 6). In both scenarios,theannual average temperature rises more than 5˚F(3˚C) by the 2090s. Increases in temperature aregreatest along the eastern edge of the RockyMountains. Minimum temperatures rise more in thewinter (December-January-February) than the sum-mer (June-July-August). By the 2090s winter temper-atures increase 14˚F (7˚C) compared to 9˚F (5˚C)for summer in the Canadian scenario,and 9˚F (5˚C)versus 7˚F (4˚C) for the Hadley scenario.

Great Plains’annual precipitation increases by atleast 13% in both scenarios by the 2090s (Figure 7).A pattern of decreasing precipitation trends appearsin the lee of the Rocky Mountains and is greatlyaccentuated in the Canadian scenario. The annualincreases are greatest in the eastern and northernparts of the Great Plains. Winter precipitationincreases slightly more than summer precipitationin both scenarios. Precipitation is likely to occur inmore intense rainfall events,especially in theSouthern Great Plains. Although precipitationincreases are projected for parts of the Great Plains,increased evaporation from rising air temperatureswill outweigh the surplus of moisture fromincreased precipitation and soil moisture will likelydecline for large parts of the region.

Drought conditions within 8 climate divisions in theGreat Plains have been described using the PalmerDrought Severity Index (PDSI): values between 1.99

201


and –1.99 are considered near normal. Values of –2to –2.99 are considered moderate drought; values of–3 to –3.99,severe drought,and values less then –4are considered extreme drought. The droughts ofthe 1930s and 1950s are shown as years or periodswhere the PDSI was less then –2 and both climatescenarios suggest future periods where droughtconditions appear likely (Figure 8). If vegetationcover is not maintained in the sand dune areas ofthe Great Plains,there is some chance that increaseddroughtiness may result in mobilizing sand dunes(Muhs and Maat,1993).

The number of times that a climate division in eachof 8 Great Plains states experiences three consecu-tive days exceeding 90˚F (32˚C) also increases in

Annual Average Palmer Drought Severity Index (PDSI)

Figure 8. The droughts of the 1930s and 1950s are shown as yearsor periods where the Palmer Drought Severity Index (PDSI) wasless then –2 and both climate scenarios (CCC = Canadian, Hadley)suggest future periods in each of these 8 climate divisions in theGreat Plains where drought conditions appear likely. The 95% con-fidence interval for the historical period of 1960 to 1990 is shown astwo dashed lines (VEMAP data, Natural Resource Ecology Lab,Colorado State University.) See Color Plate Appendix


202

both scenarios (Figure 9). For the climate divisionsin Colorado and Oklahoma,this represents morethan a doubling of the number of times such heatstress would occur.

KEY ISSUESWorkshops held in the Great Plains identified stake-holders’concerns about climate change and climatevariability in the context of other current stresseson the environment and society (Ojima et al.,1997;Seielstad,1998;National Aeronautics and SpaceAdministration and University of Texas at El Paso,1998). The key issues explored in detail belowapply equally to the Northern,Central,and Southern

Great Plains. Additional issues were identified andthese are discussed as a group following the sec-tions below. The key issues impact the full spec-trum of industries and ecosystems in the GreatPlains and were chosen partly because they arecommon to a great deal of the economy and ecolo-gy of the region; for example,all of the key issuesrelate to agriculture.

• Changes in the timing and quantity of watercould exacerbate the current conflicts surround-ing water allocation and use in the Great Plains.

• Potential shifts in climate variability may increasethe risks associated with farming, ranching,andwildland management.

• I nva s i ve species may have unanticipated indire c timpacts on the Great Plains ecology and economy.

• Rural communities,already stressed by theirdeclining populations and shrinking economicbase,are dependent on the competitive advan-tage of their agricultural products in domesticand foreign markets. A changing climate willbring additional stresses that will disproportion-ately impact family farmers and ranchers.

• Soil organic matter is a critical resource of theGreat Plains as it provides improved soil waterretention,soil fertility, and the long-term storageof carbon.(see Agriculture Solutions box)

1. Changes in Timing and Quantityof Water

Current IssuesWater supply and demand,allocation and storage,and quality are all climate-sensitive issues affectingthe region’s economy. Competing uses for waterinclude agriculture,domestic and commercial uses,recreation,natural ecosystems,and industrial usesincluding energy and mining. Texas leads the GreatPlains states in the total amount of water withdrawnfrom either surface or groundwater sources (Figure10) although Wyoming has the highest per capitause of water, the result of a small human populationand a large agricultural use. Agriculture (irrigationand livestock) withdraws the largest share of waterin every Great Plains state, except North Dakota(Figure 10). “Consumptive use” refers to that part ofwater withdrawn that is evaporated,transpired,incorporated into products and crops,consumed byhumans or livestock,or otherwise removed from theimmediate water supply (Solley, 1997). Agricultureis the largest consumptive use category in everystate (Figure 10),accounting for over 40% of thetotal water used in most states.

3+ Consecutive Days exceeding 90ºF (32° C)

Figure 9. Heat stress events can be triggered for livestock and forhumans when the temperature exceeds 90˚F (32˚C) for three ormore consecutive days. The number of times that a climate divi-sion in each of the 8 Great Plains states experiences three consec-utive days where temperatures exceed 90˚F (32˚C) increases inboth scenarios. The 95% confidence interval for the historic periodof 1960 to 1990 is shown as two dashed lines (VEMAP data, NaturalResource Ecology Lab, Colorado State University). See Color PlateAppendix

Sources of water include precipitation,surfacewater in rivers,streams and lakes,and groundwaterin aquifers. Surface water supplies most of thewater withdrawals in Montana,North Dakota,andWyoming whereas groundwater sources are greatestfor Nebraska (Solley, 1997). Seasonality of precipita-tion also influences user dependency on thesesources (Miller, 1997). Irrigated agriculture alongthe eastern edge of the Rocky Mountains is moredependent upon snowmelt runoff from the RockyMountains than on spring and early summer rains.Non-irrigated agriculture in the eastern parts of theGreat Plains is less dependent upon snowmeltrunoff and more dependent upon the spring andearly summer rains.

Water quality is a current constraining factor in theproductive use of water. The management of waterquality problems,such as salinity, nutrient loading,turbidity, and siltation of streams,is tied to the avail-ability of water to accommodate agricultural andhuman demands. Dams,diversions, channelizations,and groundwater pumping have influenced nearlyall freshwater ecosystems in the Great Plains byaltering riparian habitats,aquatic ecosystems, hydro-logical cycles,and recreational opportunities.

In each state of the region,the allocation of wateramong competing uses depends on the ownershipof water rights,and on the contracts and operatingrules governing federal and other public water proj-ects. Initial allocations can be modified by markettransactions,but the cost of transferring water orwater rights through markets varies considerablyfrom state to state. For example,laws restrict trans-fers from agricultural to non-agricultural uses insome states such as Nebraska. Water apportionmentdecision-making among the various sectors is amajor challenge and is currently marked by con-flict,negotiation,or cooperation,depending on thesetting. While interstate allocation issues have gen-erated significant conflict,there are promising signsof cooperation in the tri-state effort betweenColorado,Wyoming,and Nebraska to address endan-gered species preservation on the Platte River. Inaddition,negotiations between urban centers andirrigators who are willing to sell their water rightsor rent water owned by the city are becoming com-monplace in some states. Management of both sur-face water and groundwater to meet diverse andincreasing water needs is a major political and man-agement concern in the region.

Potential ImpactsThe projected increases in temperature anddroughts are expected to exacerbate the current

competition for water among the agricultural sector,urban and industrial users, recreational users,andnatural ecosystems,as well as within each user com-munity. As water needs and available resources dif-fer across use categories and within categories,changes such as a shift in the seasonality of precipi-tation will impact users differently. For example,alteration in the timing of snowmelt runoff from theRocky Mountains would impact the current systemof water management in the Platte Basin. In theSouth Platte Basin, water is taken from thestreams/rivers in the spring and pumped strategical-ly into nearby shallow groundwater aquifers suchthat the release of the recharge meets downstreamuser needs at the appropriate time. Winter warmingcoupled with an increase in winter precipitationcould result in earlier and greater snowmelt fromthe Rocky Mountains. This storage system along theSouth Platte is adapted to the current climate andtemporal needs of water users;under projectedchanges,the release of recharge may not meet thetiming or amount of the downstream users’ waterneeds.

Irrigated cropland has allowed for a diversificationin agriculture in the Great Plains. Lack of soil mois-ture can greatly reduce yield of crops and forage.Ojima et al.(1999) found that,under the Canadianclimate scenario,consumptive demand for water forperennial crops such as grass and alfalfa wouldincrease at least 50% by the 2090s over current use(average of 1981-90). However, because growingseason precipitation increases,consumptive water

203


Great Plains Water Use

Figure 10. Surface waters are important sources for the westernand northern Great Plains. Ground water, such as the Ogalallaaquifer, supplies large shares of the water for Nebraska, Kansas,and Oklahoma. Although the total amount of water withdrawalvaries across the Great Plains, agriculture is the dominant con-sumptive use in all states (Solley, 1997). See Color Plate Appendix

Water availability in both dryland and irrigated sys-tems could be improved by existing and new tech-nologies for residue management and tillage prac-tices. Various techniques have been used toincrease storage and availability of water:manage-ment of groundwater aquifers;enhanced snowpackstorage in mountains through forest management;crop management practices that enhance soil mois -ture retention through crop stubble,wind breaks,and mulches;and snow management strategies onthe Plains. Such techniques could increase thequantity of stored water to provide resilience to achanging climate. Irrigation scheduling,adjustingyield target to match available water, and/or chang-ing cropping systems or land use in the event thatirrigation costs exceed the worth of increasedproduction are other options. More efficient irri-gation application methods in agriculture (such asprecision farming) could decrease water con-sumption. The need for better, non-consumptivewater use in urban areas could be achievedthrough conservation, xeriscaping (low water-uselandscaping),and the use of gray water systemsfor landscapes. However, the effectiveness of suchmeasures and distribution of impacts across vari-ous water users would depend on a number ofboth natural and institutional factors. For example,more efficient non-consumptive water use in onelocation does not necessarily result in less overallwater demand as diminished use at one point maysimply allow increased use downstream. In addi-tion,potential water quality issues may arise withrepeated use.

Water trading is another drought managementresponse that is more developed in some statesthan in others. In Colorado, for example,there areactive water rental markets along the Front Rangeof the Rocky Mountains,and many cities that haveacquired water rights in advance of need routinelyrent them back to agricultural users in normalwater years. This practice provides a drought bufferfor the urban uses because the city can decide notto lease the water during drought years. “Waterbanking”is a term applied either to conjunctive useof groundwater and surface water supplies or to aformal mechanism to facilitate voluntary watertransfers. This is a relatively new concept in theGreat Plains region,but both Texas and Kansas haveinstituted programs to encourage water banking.

2. Weather Extremes

Current IssuesExtreme weather events include severe wintersnow storms,ice storms,high winds,hail,torna-

use for corn does not increase substantially (Figure11). Thus,depending upon the future changes intemperature and precipitation,both irrigated landsand farmed acreage not currently irrigated couldcompete for scarce water resources in the future.

Potential increases in drought and/or storm intensi-ty could severely impact water quality. Non-pointsource pollution can contain contaminants fromfertilizers,herbicides,pesticides,livestock wastes,salts,and sediments that reduce the quality of bothsurface water and groundwater drinking water sup-plies. Many small towns in the Great Plains strug-gle to meet current drinking water standards. Theprojected increase in intense rainfall in theSouthern Plains may increase problems with runoffin urban areas or runoff of livestock wastes fromfeedlots.

Adaptation OptionsThere are existing strategies that deal withdroughts, chronic water shortages, extreme weath-er events and year-to-year climate variability.Thesestrategies favor improvement and maintenance ofsoil, water, biotic,and land resources. Numerouscultural,economic,political,and social factors ofteninhibit a rapid and widespread adoption of moresustainable practices (Wilhite,1997). These existingpractices and new technologies are possibly onlymarginal in effecting the climate change impacts.

204


Consumptive Water Use

Figure 11. Lack of soil moisture can greatly reduce yield of cropsand forage. Under both climate scenarios, the consumptivedemand for water on grass pasture increases more than 50% whilethe water needs for irrigated corn change little. Perennial cropsexperience an increase in consumptive demand for water; the sizeof the increase depends on the climate scenario (Ojima et al., 1999).See Color Plate Appendix

does,lightning,drought,intense heavy rain, floods,heat waves, extreme cold snaps,and unexpectedfrosts. Natural systems have adapted to thisvariability, but climate extremes have significanteconomic impacts on farmers and ranchers aswell as the human communities in the GreatPlains. For example,in May of 1999,an out-break of F4-F5 tornadoes hit the states ofOklahoma,Texas,Kansas,and Tennessee, result-ing in at least $1 billion in damages and 54deaths. In fall 1998,severe flooding in south-east Texas from 2 heavy rain events with 10-20inch rainfall totals caused approximately $1 bil-lion in damages and 31 deaths. The severedrought from fall 1995 through summer 1996 in theagricultural regions of the Southern Great Plainsresulted in about $5 billion in damages.

Urban and industrial infrastructures have also beenimpacted by extreme weather events. For example,the summer 1998 heat wave and drought severelyimpacted roads and pipelines in Texas. In addition,this extreme event resulted in over $6 billion indamages from Texas/Oklahoma eastward to theCarolinas and at least 200 deaths. The April 1997flood put nearly 90% of Grand Forks,North Dakotaunder water and caused over $1 billion in damages.

The extremes of hot and cold,as well as wet anddry, pose challenges for livestock enterprises. In thewinter of 1996-97,eight blizzards in North Dakotaresulted in the deaths of over 120,000 cattle,9,500sheep,and several thousand hogs and poultry(Junkert,personal communication via Seielstad)with direct losses of $250 million. Heat,particularlyhot-humid conditions,impacts the performance ofintensive livestock operations more than cold,andcattle are impacted to a greater degree than sheep(Hahn and Morgan,1999). In August 1992,a 3-dayheat wave after a relatively cool period in centraland eastern Nebraska caused several hundred feed-lot cattle deaths. In a July 1995 heat wave, over4,000 feedlot cattle died in the central US (Hahnand Morgan,1999). Weather is the primary factor inmanagement decisions on the timing of calving sea-sons because newborns and neonatal animals arevulnerable to extremes of both heat and cold.

Potential Impacts Analyses of the historical record (1895-1995) identi-fy increases in high intensity rainfall events (greaterthan 2 inches/day) in the Southern Great Plains(Karl et al.,1999). Model results suggest that thefrequency of high intensity rainfall will continue toincrease in the Southern Great Plains, resulting inmore rainfall in shorter periods of time. The histori-

cal record also indicates that 7-day extreme precipi-tation totals are increasing in the Great Plains(Kunkel et al.,1999). In both the Canadian and theHadley scenarios,the frequency of very high tem-peratures and heat stress events appear likely toincrease.

Stakeholders in the region view slow changes in cli-mate averages as less of an issue than the possibilityof greater or more frequent extremes. Hail suppres-sion programs, rainfall enhancement programs,anddrought mitigation programs currently operate indifferent areas in the Plains. These current concernsabout extreme events were also highlighted as con-cerns under a changing climate in the Great Plains.

Adaptation OptionsStakeholder groups recognized that climate in theGreat Plains is inherently variable,and that decisionsregarding land use,land management,and develop-ment are made with this variability in mind. Thisvariability underscores the need for adaptationstrategies that reduce risk and uncertainty throughbetter access to more timely, accurate,easily accessi-ble information about near-term weather (weeks togrowing season), extreme weather events,and fore-casts for weather 6 to 18 months in the future.Physically, emotionally, and psychologically, peoplecope better with a disaster or an abrupt change ifthey are prepared or if a response plan is in place.

Making decisions under uncertainty requires thatstrategic (long-term) planning decisions be made sothat a tactical (short-term) response can be initiatedwhen needed. For example,the stress of a heat

205


THI and Wind Speed at Rockpor t, MO., in July, 1995

Figure 12. The hourly temperature-humidity index (THI) values andwindspeeds at a mid-central US location in the highest-risk area ofan early summer, 1995 heatwave which caused large-scale live-stock deaths and performance losses (Hahn and Mader, 1997). Thetemperature-humidity index averaged above 80 for more than 3days here, and nighttime relief (THI less than 74) from the extremedaytime heat did not occur naturally. See Color Plate Appendix

event and the vulnerability of the human communi-ty (Wilhite,1997). If increasing demand for waterin the Great Plains increases the social vulnerabilityto water supply disruptions,then the impact offuture droughts will be greater even if climate pat-terns were to remain the same. Thus,policies thatdevelop regionally appropriate drought mitigationmeasures today will likely reduce the social impactsassociated with future droughts,whether or notthey are the result of climate change (Wilhite,1997).

3. Invasive Species andBiodiversity

Current Issues Natural systems in the Great Plains are currentlystressed by a variety of agents including fragmenta-tion of grasslands through land conversions to agri-culture,cities,and roads;sedimentation and waterpollution from fertilizer and pesticide runoff;intro-duction of invasive species through human activi-ties and natural encroachment;altered hydrologydue to the impoundment and diversion of water;and changes to natural runoff from watersheds byhuman activities that alter the natural efficiency ofwatersheds and the permeability of soil surfaces.Increasing human demands on natural systems forconsumable wildlife opportunities (such as huntingand fishing) and other recreational opportunitiesare also likely to continue to stress natural systems.

The pattern of persistence of native species in theGreat Plains is likely associated with the regionalpattern of agriculture and urban development. Inthe agriculturally rich,eastern portions of the GreatPlains,the native habitats are absent or highly frag-mented. In the central portion of the Plains,thesehabitats may still be present but are largely discon-nected. In the western edges,the native habitatsare often more continuous although the recentrapid expansion of urban areas along the RockyMountains is increasingly fragmenting these nativecommunities. Persistence of species depends onsufficient habitat,sufficient core areas,or sufficientconnectivity between habitat patches.

Invasive species are currently a significant issue onthe Great Plains. Invasive species are plant or ani -mal species that have been introduced into an envi-ronment in which they did not evolve and theyusually have no natural enemies present to limittheir reproduction and spread (Westbrooks,1998).Invasive species typically have high reproductiverates, fast growth rates,and good dispersal mecha-nisms. The costs and weed-associated losses in

event can be minimized if livestock operators have3-4 days to reduce feed intake (and therefore,meta-bolic heat production) in cattle. Livestock can copewith high heat during the day if nighttime tempera-tures are sufficiently low for the animals to cooldown. In the 1995 event,the heat wave was suffi-ciently extreme (Figure 12) and extensive to causeheat loads which exceeded the stress threshold andled to impaired performance for many animals anddeath for those most vulnerable (Hahn and Morgan,1999). If livestock operators knew that the heatindex was going to exceed a certain threshold moreoften,operators could make strategic decisions toprepare for extreme heat events through some typeof cooling management,such as fans,shade,or theuse of water sprays. Without such forethought,animmediate response to cool animals is difficult toarrange. Information about climate and real-timeweather would allow farmers and ranchers to makestrategic decisions and be able to respond quicklywhen necessary.

In the Northern Great Plains,the stakeholders dis-cussed how they could build learning communitiesin which stakeholders share information with eachother;in other words,information flows in all direc-tions (Seielstad 1998). Information outreach couldinclude short television segments containing infor-mation pieces on climate and other aspects of theenvironment,a web site with interactive informa-tion,newsletters,and partnerships with the NationalPark Service to develop information kiosks aboutenvironmental changes.

Nearly all Great Plains states have drought manage-ment plans in place or are developing them. Theseplans tend to be reactive, focusing on the emer-gency response to drought rather than a mitigationplan (Wilhite,1997). Such plans may need to beupdated to reflect the changing demographic andeconomic conditions within Great Plains states.Colorado’s drought plan incorporates a monitoringsystem,an impact assessment,and a response sys-tem. The monitoring system allows for the earlydetection of a drought. In the assessment plan,thepotential impact of a drought on 8 sectors is evalu-ated: municipal water, wildfire protection, agricultur-al industry, tourism,wildlife,economics,energy loss,and health. The response phase deals with theunmet needs identified by the assessment andassists local communities when their capabilities areexceeded. In the aftermath of a drought,an evalua-tion of the response provides suggestions to reviseand improve the drought response system. Theimpact of a drought—economic,social or environ-mental—is the combination of the meteorological

206


crop and forage production in the agricultural sec-tor are nearly $15 billion annually. Introduced in1827 as a contaminant in seed,leafy spurge occursin all of the Great Plains states except Oklahomaand Texas. Grazing capacity of areas with morethan 10 to 20% leafy spurge cover is significantlyreduced (Westbrooks,1998). The direct and sec-ondary economic impacts of leafy spurge infesta-tions on over 1.6 million acres in North and SouthDakota,Montana,and Wyoming were approximately$129 million (Leitch et al.,1994). Crop losses inKansas are annually $40 million from fieldbindweed (Westbrooks,1998). With an ability toreduce wheat yield by 25%,jointed goatgrass hasinfested 5 million acres of winter wheat and isspreading at a rate of 50,000 acres or more a year

(Westbrooks,1998).

In native ecosystems,invasive species may compro-mise the ecosystem’s ability to maintain its structureor function (Stohlgren et al.,1999,Mack andD’Antonio,1998,Vitousek et al.,1996). In grasslandecosystems, riparian areas next to streams and riversare rich in native plant species and highly produc-tive. These riparian areas are also critical habitat forspecies associated with the surrounding drierecosystems,offering shelter and forage. These ripari-an ecosystems may be easily invaded by invasivespecies and may facilitate the establishment andmigration of exotic plant invasions to upland sites ingrassland ecosystems (Stohlgen et al.,1998). Natural

207


Risk Management in the Northern Great Plains:Predicting Wheat Yield and Quality

Farmers have always monitored their crops closely for signs of nutrient deficiency, heat and water stress,insect infestation,and disease. With large acreages,however, it can be difficult to find the time to monitor allfields and even more difficult to follow change over a growing season and from year to year. Remote sensinghas the advantage of being able to view large areas over one or several growing seasons. It can also detectradiation—for example,near-infrared (IR) radiation—which is invisible to the human eye. Plants absorb radia-tion in the visible red (R),but radiate it strongly in the near-infrared;how strongly depends on the plants’vigor and health. When vegetation indices,calculated from IR and R observations,are measured frequentlyover a season,the progression of canopy emergence,maturity, and senescence can be seen over large areas.The indices are also related to crop yields and, for grains,protein content. Real-time access to such data couldbe useful in agricultural and ranching operations.

Dryland wheat farmers and researchers in the Northern Great Plains used one vegetation index,theNormalized Difference Vegetation Index (NDVI),to see if protein content of wheat could be determinedremotely during the 1998 growing season (Seielstad,1999). Protein content,a measure of wheat quality, isrelated to existing soil nutrients, chemical inputs,plant biomass,and weather. The amount of nitrogen fertiliz-er required to produce a specific grain protein content is related to rainfall. In a wet year, early rapid plantgrowth will deplete soil nitrogen, resulting in a deficit late in the season and if additional nitrogen is notapplied,the high yield will be of low quality. Conversely, late-season nitrogen application would not likelyimprove a crop’s quality in a dry year.

With early information on weather and protein content of the crop, farmers could apply part of the total fer-tilizer typically needed for the entire growing season. The decisions about further fertilizer applications couldbe made as the growing season progresses. If less than average precipitation occurs or appears likely tooccur, the initial fertilizer application may be sufficient. If growing season precipitation is predicted to beadequate for wheat production,then additional fertilizer to meet the growing needs could be applied. Thismanagement practice could save money, reduce runoff of unnecessary nitrogen,and lower the eventual emis-sion of nitrogen into the atmosphere as a greenhouse gas.

Producer response was positive. In an article published in the Precision Agriculture Research Association(PARA) Newsletter, cooperating producers Carl and Janice Mattson write “...these NDVI images showed usthe broad picture and gave us a great deal of optimism when it comes to satellite images. Perhaps withthe more detailed images possible with the new satellites going up,there may be a vast amount of knowl-edge that we can apply to our management decisions.” Several other producers also had positiveresponses. A similar study was applied to rangelands to develop efficient methods of assessing foragequantity and quality.

composition. In the short-grass steppe, for instance,the slight warming of nighttime temperatures seenin the last 20 years has been linked to the decline ofblue grama grass,the dominant grass of the short-grass prairie (Alward et al.,1999).Increased averagetemperature and annual precipitation in the CentralGreat Plains may make it possible for invasiveplants,such as kudzu and Johnson grass,now foundfurther south,to migrate north. Additional land-usepressure on these native systems is likely if agricul-tural practices extend into these areas as a result ofmore favorable climate or demands for agriculturalproducts.

Changes,such as nutritional value of plants andchanges in timing of insect emergence,may imply adecline in avian populations in the Great Plains,while longer growing seasons and the possibility ofincreased productivity may mitigate the declines(Larson,1994). Grassland bird species are currentlydeclining,a function of loss of habitat. Further alter-ation of their habitat from climate change would bea likely additional stress on these declining popula-tions.

Changes in precipitation,temperature,and thehydrological cycle are likely to impact aquatic sys-tems and the terrestrial animals dependent on theseecosystems. The abundance of wetlands is closelytied to the interannual variation in climate,both pre-cipitation and temperature (Malcolm and Markham,1998). Changes in annual precipitation in the north-

disturbance regimes such as fire and grazing areseen as important in maintaining the native species.Alteration of the frequency, intensity, spatial pattern,or scale of these disturbances may expedite thereplacement of native species with exotics(Stohlgen et al.,1999). The loss of native speciesand the increasing presence of exotic or invasivespecies are current and continual challenges for nat-ural resource managers of the Great Plains.

Potential ImpactsThe projected climate changes will likely alter thebiodiversity of the Great Plains. The “new”composi-tion of species that might arise in response to the“new” climate may or may not be detrimental tosociety, but the rapid rate of change in climate islikely to be disadvantageous to native species.Invasive species exploit habitats left vacant bynative species susceptible to multiple stresses. Asclimate changes,the indirect impacts of weeds andpests are likely to bring surprising challenges.

Some native species will be unable to adapt fastenough to new climate regimes, resulting in a low-ered competitive edge and weakened resistance toinfestations by invasive plants and animals. Potentialimpacts include shifts in the relative abundance anddistribution of native species,significant changesin species richness and assemblages,and localextinctions of native species. Subtle changes inthe diurnal or seasonal patterns of temperaturehave been shown to affect plant community

208


Agriculture Solutions to Global WarmingBy Martin Kleinschmit, Farmer and Rancher, Bow Valley, Nebraska

Farmers have a lot at risk as global climate heats up,but they also have a lot to gain by participating in thesolution to climate change. By conserving soil organic matter, farmers can improve soil health and productivi-ty as well as capture and store (sequester) carbon in the extensive crop and rangelands of the Great Plains.The higher temperatures and greater numbers of droughts and floods projected for the region could threatencrops, raise production expenses,and increase the risk of failure. To protect our food supply, healthy soilsable to withstand erratic weather patterns are needed. Increasing the carbon content of the soil will help tomitigate global warming by keeping carbon dioxide out of the atmosphere,but it will do even more to bufferthe soil against the threats of climate change. Presently most US farmland has only half or less of its historicallevel of organic matter. Soil scientists have established that a 6-inch (15 cm) block of soil with 1-2% organicmatter can hold only about one inch (2.5 cm) of rain before it runs out of the bottom. With 4-5% organicmatter, that same soil can hold 4-6 inches (10-15 cm) of rain before it leaves the root zone and takes with itthe water-soluble nutrients. Increasing soil organic matter also reduces the risks of flooding and erosion,andretains moisture longer so plants have access to it during periods of dry weather. It lessens the need for (andexpense of) irrigation, reduces ground water pollution,and reduces the amount of run-off, lessening thethreat of stream pollution. It also lowers the cost of fertilization since nutrients not lost to erosion and leach-ing need not be replaced. Agricultural incentives that encourage carbon sequestration in soil provide anopportunity to promote food security in a changing climate and reduce the threat of climate change at thesame time.

ern regions are likely to have significant impacts onthe breeding duck populations (Sorenson et al,.1998;Bethke and Nudds,1995). Changes in precipi-tation and riverine flow regimes are likely to exacer-bate land-use conflicts and competition for watersupply including conservation needs.

Maintaining natural biodiversity—the full array ofnative plants,animals,natural communities,andecosystems that occur within the Great Plains—maybe difficult as climate changes. In the NorthernGreat Plains,the relatively undisturbed landscapes ofthe Indian reservations contain a variety of micro-environments supporting a wide range of indige-nous plant and animal species. The scattered federalland holdings within the Northern and CentralGreat Plains also offer refuges for species andopportunities to enhance native vegetation. In theSouthern Plains,however, there is very little protect-ed land (Figure 1). The challenging issue in this arearelates to private sector land management for eco-tourism,and whether this is an appropriate route tomaintaining some degree of biodiversity.

Adaptation Options Rather than identify specific strategies,stakeholdersin the Central Great Plains proposed a set of generalprinciples to guide strategic development of socialresponses to climate change. These five principleswere articulated as follows:

First,“no regrets”strategies,those that respond toexisting stresses while making the system moreresilient to climate change without incurring signifi-cant costs (OTA,1993),should be vigorouslyexplored. There is a high level of uncertainty inregional climate projections and even greater uncer-tainty associated with how natural systems willrespond to those changes. Developing detailedadaptation strategies based on predictions of futurebehavior of natural systems is currently not tenable.Instead,no-regrets strategies are particularly appro-priate for natural systems because current environ-mental stresses could be addressed and mitigatedthrough such strategies. The implementation of ben-eficial strategies today could have a positive influ-ence on future stresses/impacts that may accruefrom a changing climate.

Second,the key to developing effective copingstrategies for present and future stresses is to pro-vide organisms with alternatives for adaptation,suchas landscape heterogeneity and high levels of con-nectivity in aquatic and terrestrial systems.Landscape heterogeneity is the diversity amongecosystems,such as grasslands and forests,and with-in ecosystems,such as different successional stages.

This latter type of diversity depends on maintainingappropriate disturbances to the ecosystems,such asperiodic fire. In many cases,disturbance activitiesneed to be created by management actions,such aslivestock grazing and prescribed burning,in order tocompensate for the loss of natural disturbanceregimes,such as buffalo herds and wildfire. The ideaof enhancing land stewardship by private landown-ers is central to the success of this managementprinciple.

The third principle focuses on preserving currentland uses that promote integrity in natural systems.This would entail,to the extent possible,encourag-ing conservation and restoration through properland management. A fundamental need in imple-menting this principle is to identify actions that fos-ter long-term economic vitality while at the sametime enhancing ecosystem resiliency.

The fourth principle is adaptive management. Thestakeholders felt that it is critically important tolearn by doing and to evaluate what works andwhat fails to work in an attempt to lessen theimpact(s) of climate change on natural systems.There will be surprises no matter how well pre-pared stakeholders may be;therefore managementthat is flexible and responds quickly will be mosteffective for dealing with uncertainties.

Finally, effective coping strategies depend on inform-ing the public and decision makers about the impli-cations of climate change for natural systems,andwhat these effects mean to the quality of humanlife. For example,why is the role of wetlands inflood control important to society? What couldchanges in this natural system mean to a communityor to natural systems on local and regional bases?

The stakeholders saw these principles as fundamen-tal to the discussions of climate change in allregions and an effective way to educate the generalpublic and decision-makers about the related issuesinvolved with climate change.

4. Quality of Rural Life on theGreat Plains

Current IssuesCurrent demographic changes are imposing chal-lenges to the rural areas of the Great Plains.Declining rural populations,the aging of remainingresidents,and the increased remoteness of neigh-bors place rural communities at increased socioeco-nomic risk. The growing urban areas are magnets

209


Potential ImpactsThe projected changes in climate – increases in tem-perature, reductions in soil moisture,and moreintense rainfall events – will likely require changesin crop and livestock management in the GreatPlains. Because rural populations and their commu-nities are highly dependent on the natural resourcesof the Great Plains,they are at risk from climatechange,and from potential increases in climate vari-ability. Rural economies in semi-arid regions areeconomically vulnerable due to lower marginaleconomies (lower profits and tax bases, fewerresources available) and their reliance on livestockand cropping systems that are often stressed.International exports will reflect the climate changeimpacts on the global agricultural markets.

Increases of warm-season forages may be a welcomeaddition to a forage mix in the Northern GreatPlains,but the loss of the current diversity of warm-and cool-season forages in the Central Great Plainsmay pose limitations in grazing management. Theelevated atmospheric concentration of carbon diox-ide will possibly lower forage quality of native grass-es. Legumes,a potential source of nitrogen,couldbe a new and important part of farm and ranchmanagement. Changes in the seasonality of precipi-tation,particularly the growing season precipitation,will likely impact plant growth of native vegetationand crops with and without irrigation in the GreatPlains. Warmer winters will mean some chance of

for jobs and people. This shift in populationincreases the demand for services in urban areas,while increasing the burden on rural governmentsto provide health and education services with adeclining economic base. The vulnerability of therural human populations on the Great Plains willaffect their ability to marshal resources,both natu-ral and societal,to cope with increased risk anduncertainty. As the urban centers continue to grow,problems such as air quality will likely compromisethe quality of life in urban areas,including thehealth-related aspects (see Additional Issues).

The consequences of weather and change on agri-cultural economics,beef cattle production,andgrassland use can be subtle and complex due toindirect impacts from international trade,cost offeed,and markets. Most agricultural commoditiesare subject to production/price cycles,with thetime between peaks and troughs of production con-trolled largely by the producer’s ability to respondto price signals and consumer behavior. However,climate variability, especially drought,can signifi-cantly modify the dynamics of cattle inventoriesand production/price cycles, resulting in losses toproducers. For example,the 1995-96 Texas droughtresulted in larger numbers of cattle being sent tomarket due to poor range condition,increased cornfeed prices,and the largely diminished winterwheat feed crop.

210


Table 1. Soil and Water Conservation Strategies and their Benefits(Soils Working Group,Ojima et al.,1997).

Adaptation Action Benefits to Benefits to Climate Change IssueFarmer/Rancher

Soil organic matter management Increase in water-holding capacity Increase in soil carbon storageIncrease in soil fertility

Precision Farming Reduction in N2O emissionsTargeted fertilizer application Cost savingsTargeted water application Cost savings/reduced

salinizationTargeted pesticide application Cost savings/reduced

toxification

Energy from biomass Diversified Reduction in carbon dioxide emissions from fossil fuel burning

Managing livestock wastes to Usable energy Reduction in methane emissions capture methane

more rain than snow with resultant deeperrecharge,enhancing the competitive advantages ofshrubs. More intensive storm activity and anincreased frequency of heat waves will likely be anincreasing problem for the Southern Great Plains.Whether or not the plant community will be able toaccommodate changes in growing season climate orhydrological patterns is a matter of concern amongstakeholders who depend on these weather pat-terns for their livelihood.

Adaptation OptionsThe stakeholders in the Great Plains proposed thatthe most effective adaptation strategies would be“no regrets”actions,developed from the bottom-upthrough community-based efforts,with an emphasison risk reduction and increasing diversification.Because each community has dif ferent needs andvalues,a community-based approach was stronglysupported to address issues related to adaptationand mitigation of climate change and variability.Stakeholders identified that any government policydirected towards responding to climate change atthe national scale should focus on the long-term,notshort-term economic incentives;should be flexible,allowing for local implementation and shortresponse times;and should promote adaptationstrategies that are sustainable and economicallyviable.

In the agricultural sector, various strategies haveevolved to cope with drought and soil erosion(Ojima et al.,1997). Many of these coping strategiesnot only provide direct benefits to the farmer orrancher, but are also beneficial to the environment(Table 1). The loss of carbon and water from crop-lands can be minimized through practices such asreduced tillage. Cover crops and residue manage-ment can facilitate soil conservation by suppressingsoil loss from wind and water erosion. Precisionagricultural practices that integrate specialized cropvarieties, fertilizer inputs,and irrigation schedulesinto crop management may provide technology tocope with climate changes.

Stakeholders also spoke about diversification,howthis strategy had helped in coping with other cli-mate or economic events in the Great Plains,andwhat institutional factors limited the ability of peo-ple in the Great Plains to diversify their operationsand their local and regional economies. Livestockenterprises are often a mix of range managementand planted forage or crop activities. Stakeholdersidentified diversification of land use as an importantstrategy to increase profits and/or reduce risk.Examples included a strategy that some operators

have already adopted such as diversifying ranchoperations to include recreation,or a new strategythat policy makers could implement such as carboncredits. In addition,stakeholders identified theimportance of diversification within a land use,such as ranching,to cope with the effects of cli-mate change. Diversification strategies included 1)mix or change animal species to fit the new envi-ronment,2) change genetics of the animal species,3) change seasonality of production (e.g.,calving,lambing, weaning),or 4) reduce production prac-tices in stressful environments.

ADDITIONAL ISSUES• Industrial and urban infrastructures designed for

historical climate extremes may be inadequateunder a different climate. As the Great Plainscontinues to change due to social and economicfactors,demands on the existing water resourcestructures and other social infrastructures willalso change and further challenge resource allo-cations in the future for land and water. In addi-tion, climate change will affect long-term plan-ning regarding current capacity or future designof additional infrastructure needs to sustainablyutilize water and land resources. For example,in the Southern Great Plains, researchers havebeen working with urban engineers to assessunexpected impacts on urban infrastructuresfrom recent extreme events,and to developadaptation strategies for future events.

• Across the Great Plains,the poor living in com-munities of all sizes are disconnected from someof the most essential safety nets to cope withnatural hazards. The human mortality thatresults from severe heat waves, floods,and natu-ral hazards is a problem that eludes the best-intentioned policies. For example,in theSouthern Plains,the poor along the Lower RioGrande are extremely vulnerable to bothdroughts and flash floods. The relationshipbetween poverty and carrying capacity of theenvironmental system is not a typical part of acommunity’s approach to developing socialservice systems. The current struggle to achievea sustainable economy that properly takesaccount of ecosystem services and social servic-es will likely become an increasingly importantissue.

• Rising air temperatures will exacerbate the cur-rent air quality issues in the urban areas of theGreat Plains. Denver (and much of the ColoradoRocky Mountain Front Range) suffers serious airquality problems on a seasonal basis. Every

211


more informed decisions about their cropping plansunder the forecasted climatic conditions.

For native species,there is a need to synthesize cur-rently available information about the potentialimpacts of climate change on native species and nat-ural systems. This would include quantifying theecological and physiological thresholds of nativeorganisms and their tolerances to changes in envi-ronmental factors such as temperature,salinity, andsedimentation,and developing coordinated crop-ping/grazing systems that minimize impacts at criti-cal periods for wildlife reproduction. Given thatmany natural systems in the region are substantiallyaltered by human activities, research is needed onrestoration techniques that will be effective torestore biological diversity and ecosystem servicesto degraded systems.

A better understanding of the relationship betweenlivestock dynamics and rangeland condition,and therole that the diversity of both plant and animal com-ponents of rangeland ecosystems play in maintain-ing good rangeland condition is a needed area ofresearch. Studies of climate change and elevatedcarbon dioxide levels on vegetation and animaldynamics are needed to understand ecosystem-levelresponses. Climate change will interact with theother current stresses on these native ecosystemsand it is the cumulative effects of multiple stresseson natural systems that needs greater understandingand the development of management tools.

For agriculture, research is needed on the best feasi-ble methods of diversification under certain climaticconditions and in specific localities. Soil types, rain-fall,and growing seasons limit agricultural diversifi-cation. For example,including leguminous crops,like field peas,lentils and Austrian winter peas,incrop rotations has been successful in the NorthernPlains but of limited value in the Central Plains.

Research on new crop and crop variety develop-ment for a future climate involves long lead times.Waiting until the seeds are actually marketable willmean that there will be crop failure in the first fewyears of a changed climate until new seeds whichare better adapted to the new climate are produced,tested,and marketed. With new crops come newpests,and these need to be taken into considerationwhen new seed is developed.

The attainment of viable management practices forstoring carbon and conserving Great Plains landsand aquatic areas is a complex issue,and will

major city in Texas is out of compliance with cur-rent air quality standards. Houston is the secondmost polluted city in the nation. Typically, inTexas,a longer warm season will produce moredays exceeding air quality standards. This is animmediate issue in Texas because these citieswill either lose federal funds in the next decadedue to their inability to control emissions,or willspend millions of dollars in courts trying to chal-lenge regulatory penalties. The problems areexacerbated by rising urban populations frompopulation growth as well as rural migration tocities for jobs. The issue of air quality has thepotential to be an international issue if situationssuch as the 1998 wildfires in Mexico that inun-dated the Great Plains with smoke and dustbecome more frequent under an altered climate.Major air quality impacts occurred in the RioGrande and Southern Great Plains and plumesfrom these wildfires were transported as farnorth as Canada.

• Advancements in technology have improvedfarming capability. Partnership in global changeresearch and advancement of these technologiesshould provide a greater buffer against perturba-tions to the agricultural ecosystems and econo-my.

• Dissemination of information on adaptive vari-eties of crops and livestock suitable for changingconditions is needed from reliable sources forstakeholders.

CRUCIAL UNKNOWNS ANDRESEARCH NEEDSResearch needs identified for the Great Plainsfocused on improvement of weather forecasts,enhancement of diversification in agriculture,man-agement of the biodiversity of native ecosystems,improved water allocation decision-making tools,improved pest management strategies,carbonsequestration techniques,and human dimensionsresearch.

Monitoring,early detection,and distributed warningof extreme weather events would allow for prepara-tion and responses to minimize damages.Interannual and seasonal forecasts of weather wouldimprove advanced planning of many activities:whatcrops to plant,how many cattle to graze,and othermanagement activities. Increased research on theincidence and possible timing of hail in the GreatPlains would benefit farmers by helping them make

212


require new knowledge. Research is needed tounderstand the carbon cycle in the context of theecology of Great Plains’ecosystems and agro-ecosys-tems,with the objective of using that information todevelop conservation systems that lead to carbonstorage and soil conser vation. The role of wetlandsin sequestering carbon needs to be better under-stood. Economic research is also needed to start tounderstand how best to achieve the goal of promot-ing sustainable management practices including car-bon sequestration.

Research Needs Related toHumans in the Great Plains

All of the previously mentioned research will alsoenhance human-managed systems in the GreatPlains,as the health of natural systems is vital forhuman quality-of-life and for recreational activities.Water quality issues are important for rural settle-ments,as are water supply issues. Diversificationmay be an important tool for survival of many familyfarms and ranches in the Plains,and better forecastsmay encourage human preparedness for extremeevents in the Plains.

There are however, several additional research needsthat focus on the human populations of the region(cf Stern and Easterling,1999). These include com-prehending the perception of, understanding of, andawareness of climate change and its impacts in dif-ferent parts of the Great Plains. A survey should beconducted to determine respondents’perceptionsof whether and how climate has changed,how theymonitor it,how it has affected their livelihoods,andif they have made any changes as a result.

Involving local people in designing and implement-ing the monitoring of climate change is important.Stakeholders can provide input to scientists regardingwhat types of information and forecasting are neededat the local level. This can make research more valu-able to the farmer/rancher and provide an opportuni-ty for scientist-practitioner interaction. Variables thatmay require study include soil organic matter, soilmoisture,plant productivity, and extreme event fre-quency.

Rural residents other than farmers or ranchers,thoseliving in urban areas,and those that depend on thePlains for products may also be affected by climatechange. These affects should be a further area ofresearch on human dimensions. How climate changeaffects demographic patterns in the region,bothurban and rural (specifically age structure andemployment) would also be an important researchtopic. Many people enjoy recreation in the Plains,and people all over the country depend on farmproducts from the Plains. Therefore, changes thataffect the Plains will have indirect effects spreadwidely throughout the US and the world.

213


Rural Heartland , Federal Reserve Bank of Kansas City,http://www.kc.frb.org/publicat/heartlnd/hrtmain.htm,1996.

Gutmann,M.,Social and demographic changes in theCentral Great Plains,paper presented at the CentralGreat Plains Regional Assessment of Climate ChangeImpacts Workshop,Loveland,Colorado,March 22-24,1999,Sponsored by the Natural Resource EcologyLaboratory, Colorado State University, and the USDepartment of Energy, 1999.

Haas,H. J.,C.E.Evans,and E. F. Miles,Nitrogen and car-bon changes in Great Plains soils influenced by crop-ping and soil treatments,USDA Technical Bulletin 1164,1957.

Hahn, G. L.,and T. L.Mader, Heat waves in relation tothermoregulation, feeding behavior, and mortality offeedlot cattle,Proceedings of 5th InternationalLivestock Environment Symposium.Vol. I,pp.563-571,American Society of Agricultural Engineers,Minneapolis,Minnesota,1997.

Hahn, G.L.,and J.A.Morgan, Potential consequences ofclimate change on ruminant livestock production,paper presented at the Climate Change Impacts onRangeland Systems Symposium,Society for RangeManagement annual meeting,Omaha,Nebraska,February 25-26,1999,Sponsored by the Central GreatPlains Assessment and the Department of Energy, 1999.

Hayden,B. P.,Regional climate and the distribution oftall grass prairie,in Grassland Dynamics, edited by A.K.Knapp, J. M.Briggs, D. C.Hartnett,and S.L.Collins,Oxford University Press,Inc,New York,New York,US,p.19-34,1998.

Jenny, H., The Soil Resource, Origin and Behavior,Ecological Studies, vol.37,Springer-Verlag,New York,1980.

Karl et al.,Trends in average annual temperature andaverage precipitation over the 1895 to 1995 period.Data are the US Historical Climatology Network sta-tions averaged in each of the 344 Climate Divisions.Analysis prepared for National Assessment SynthesisTeam,1999.

Kunkel,K.,K.Andsager, and D. R.Easterling,Long-termtrends in extreme precipitation events over the conter-minous United States and Canada, Journal of Climate,12, 2515-2527,1999.

LITERATURE CITEDAlward,R. D., J. K.Detling,and D. G. Milchunas,Grassland vegetation changes and nocturnal globalwarming, Science, 283,229-231,1999.

Anderson, J. M.,The ef fects of climate change ondecomposition processes in grassland and coniferousforests, Ecological Applications, 1,326-347,1991.

Barkema,A.,and M.Drabenstott,Consolidation andchanges in heartland agriculture,in Economic ForcesShaping the Rural Heartland, Federal Reserve Bank ofKansas City, http://www.kc.frb.org/publicat/heartlnd/hrtmain.htm,1996.

Bethke,R.W.,and T. D. Nudds,Effects of climate changeand land use on duck abundance in Canadian Prairie-park lands, Ecological Applications, 5, 588-600,1995.

Burke,I.C.,T. G. F. Kittel,W. K.Lauenroth, P. Snook,C.M.Yonker, and W. J. Parton, Regional analysis of theCentral Great Plains: Sensitivity to climate variability,BioScience, 41,685-692,1990.

Cole,C.V., J.W. B.Stewart, D. S.Ojima,W. J. Parton,andD. S.Schimel,Modeling land use effects on soil organicmatter dynamics in the North American Great Plains,inEcology of Arable Land - Perspectives and Challenges,Developments in Plant and Soil Sciences, edited by M.Clarholm and L.Bergstrom,Kluwer AcademicPublishers,Dordrecht,the Netherlands,39,pp.89-98,1989.

Cole,C.V.,I.C.Burke,W. J. Parton, D. S.Schimel, D. S.Ojima,and J.W. B.Stewart,Analysis of historical changesin soil fertility and organic mater levels of the NorthAmerican Great Plains,in Challenges in DrylandAgriculture,A Global Perspective, Proceedings of theInternational Conference on Dryland Farming,August 1988,edited by P.W. Unger,T.V. Sneed,W. R.Jordon,and R. Jensen,Texas A&M University, CollegeStation,Texas,pp.436-438,1990 .

Donofrio,C.,and D. S.Ojima,The Great Plains today, inClimate Change Impacts on the Great Plains, Officeof Science and Technology Policy and United StatesGlobal Change Research Program Workshop,compiledby D. S.Ojima,W. E.Easterling,and C.Donofrio,Colorado State University, Fort Collins,Colorado,pp.1-24,1997.

Drabenstott,M.,and T. R.Smith,The changing economyof the rural heartland,in Economic Forces Shaping the

214


Lal,R., J. M.Kimble,R. F. Follett,and C.V. Cole,ThePotential of US cropland to Sequester Carbon andMitigate the Greenhouse Effect,Sleeping Bear Press,Chelsea,Michigan,1998.

Larson, D. L., Potential effects of anthropogenic green-house gases on avian habitats and populations in theNorthern Great Plains, American Midland Naturalist,131, 330-346, 1994.

Leitch, J.A., F. L.Leistritz,and D.A.Bangsund,Economiceffect of leafy spurge in the Upper Great Plains:Methods,models,and results,Agricultural EconomicsReport No.316,North Dakota State University,Agricultural Experiment Station,1994.

Mack,M.,C.,and C.M.D’Antonio,Impacts of biologicalinvasions on disturbance regimes, Trends in Ecologyand Evolution, 13, 195-198,1998.

Malcolm, J. R.,and A.Markham,Climate change threatsto the National Parks and protected areas of the UnitedStates,World Wildlife Fund,Washington,DC,1998.

Miller, K.,Climate variability, climate change,and west-ern water, in Report to the Western Water PolicyReview Advisory Commission, National TechnicalInformation Service,Springfield,Virginia,1997,see alsohttp://www.den.doi.gov/wwprac/reports/west.htm.

Muhs, D. R.,and P. B.Maat,The potential response ofGreat Plains eolian sands to greenhouse warming andprecipitation reduction on the Great Plains of the USA.Journal of Arid Environments, 25, 351-361,1993.

National Aeronautics and Space Administration andUniversity of Texas at El Paso,Tilting the balance: cli-mate variability and water resource management in theSouthwest,Southwest Border Regional Workshop onClimate Variability and Change,University of Texas at ElPaso,March 2-4,1998,University of Texas at El Paso,ElPaso,Texas,1998.

National Research Council, Water Transfers in the West:Efficiency, Equity, and the Environment, NationalAcademy Press,Washington,DC,1992.

Natural Resource Conservation Service,NorthernPlains Regional Office, State of the Land for theNorthern Plains Region, USDA NRCS,Northern PlainsRegional Office,Lincoln,Nebraska,1996.

Office of Technology Assessment, Preparing for anUncertain Climate, US Government Printing Office,Washington,DC,1993.

Ojima, D. S.,L.Garcia,E.Eigaali,K.Miller,T. G. F. Kittel,and J. Lackett, Potential climate change impacts onwater resources in the Great Plains, Journal of theAmerican Water Resources Association, 35, 1443-1454,1999.

Ojima, D. S.,W. E.Easterling,C.Donofrio (Compilers),Climate change impacts on the Great Plains,WorkshopReport,May 27-29,1997,Sponsored by Department ofEnergy, Office of Science and Technology Policy, USGlobal Change Research Program,and Colorado StateUniversity, College of Natural Resources,Colorado StateUniversity, Fort Collins,Colorado,1997.

Opie, J., The Law Of The Land: Two Hundred Years OfAmerican Farmland Policy, University of NebraskaPress,Lincoln,Nebraska,1987.

Ostlie,W. R.,R.E.Schneider, J. M.Aldrich, T. M. Faust,R.L.B.McKim,S. J. Chaplin, The Status of Biodiversity inthe Great Plains, The Nature Conservancy,Arlington,Virginia,1997.

Parton,W. J., D. S.Schimel,C.V. Cole,and D. S.Ojima,Analysis of factors controlling soil organic matter levelsin Great Plains grasslands, Soil Science Society ofAmerica Journal, 51, 1173-1179,1987.

Parton,W. J., D. S.Schimel, D. S.Ojima,and C.V. Cole,Ageneral model for soil organic matter dynamics:Sensitivity to litter chemistry, texture,and management,in Quantitative Modeling of Soil Forming Processes,edited by R.B.Bryant and R.W.Arnold,pp.147-167,SoilScience Society of America Special Publication 39,SoilScience Society of America,Madison,Wisconsin,1994.

Peterson, G.A.,and C.V. Cole, Productivity of GreatPlains Soils: Past, Present, and future, inConservation of Great Plains Ecosystems, edited by S.R. Johnson,and A.Bouzaher, pp.325-342,KluwerAcademic Publishers,Dordrecht,the Netherlands,1995.

Schimel, D. S.,W. J. Parton,T. G. F. Kittel, D. S.Ojima,andC.V. Cole,Grassland biogeochemistry: Links to atmos-pheric processes, Climate Change, 17, 13-25,1990.

Schimel, D. S.,T. G. F. Kittel,and W. J. Parton,Terrestrialbiogeochemical cycles: global interactions with theatmosphere and hydrology, Tellus, 43AB, 188-203,1991a.

Schimel, D. S.,T. G. F. Kittel,A.K.Knapp,T. R.Seastedt,W.J. Parton,and V. B.Brown,Physiological interactionsalong resource gradients in a tallgrass prairie, Ecology,72, 672-684,1991b.

215


VEMAP members,Vegetation/ecosystem modeling andanalysis project:comparing biogeography and biogeo-chemistry models in a continental-scale study of terres-trial ecosystem responses to climate change and CO2doubling,Global Biogeochemical Cycles,9,477-437,1995.

Vitousek, P. M.,C.M.D’Antinio,L.L.Loope,and R.Westbrooks,Biological invasions as global environmen-tal change, American Scientist, 84, 468-478,1996.

Westbrooks,R. G.,Invasive plants: changing the land-scape of America: fact book, Federal InteragencyCommittee for the Management of Noxious and ExoticWeeds,Washington,DC,1998.

Wilhite, D.,Improving drought management in theWest,in Report to the Western Water Policy ReviewAdvisory Commission, National Technical InformationService,Springfield,Virginia,see also http://www.den.doi.gov/wwprac/reports/adrought.htm,1997.

Seielstad, G.,(Ed.), Proceedings, Northern Great PlainsRegional Workshop on Climate Variability andClimate Change, University of North Dakota,UpperMidwest Aerospace Consortium,1998.

Seielstad, G.,Progress report for a public accessresource center (PARC) empowering the general pub-lic to use EOSDIS – Implementation,University ofNorth Dakota,Grand Forks,North Dakota,1999.

Skold,M. D.,Agricultural systems and economic charac-teristics of the Great Plains,in Climate ChangeImpacts on the Great Plains, compiled by D. S.Ojima,W. E.Easterling,and C.Donofrio,Office of Science andTechnology Policy and US Global Change ResearchProgram Workshop,pp.85-88,Colorado StateUniversity, Fort Collins,Colorado,1997.

Solley,W. B.,Estimates of water use in the westernUnited States in 1990 and water-use trends 1960-90,inReport to the Western Water Policy Review AdvisoryCommission, Western Water Policy Review AdvisoryCommission, National Technical Information Service,Springfield,Virginia,1997.

Sorenson,L.G.,R.Goldberg,T. L.Root,and M. G.Anderson, Potential effects of global warming on water-fowl populations breeding in the northern GreatPlains, Climatic Change, 40, 343-369,1998.

Stern, P.C.,and W. E.Easterling,(Eds.), Making ClimateForecasts Matter, National Academy Press,Washington,DC,1999.

Stohlgren,T. J.,K.A.Bull,Y. Otsuki,C.A.Villa,and L.Lee,Riparian zones as havens for exotic plant species in thecentral grasslands, Plant Ecology, 138, 113-125,1998.

Stohlgen,T. J., D. Binkley, G.W. Chong,M.A.Kalkhan,L.D. Schell,K.A.Bull,Y. Otsuki, G. Newman,M.Bashkin,and Y. Son,Exotic plants species invade hot spots ofnative plant diversity, Ecological Monographs, 69, 25-46,1999.

Tiessen,H.,L.W.B.Stewart,and J.R.Bettany, Cultivationeffects on the concentration and amounts of carbon,nitrogen,and phosphorus in grassland soils, AgronomyJournal, 74, 831-835,1982.

USDA Census of Agriculture 1997,http://www.nass.usda.gov/census/

US Department of Commerce,Bureau of EconomicAnalysis,Regional Economic Analysis Division, June1998

216


ACKNOWLEDGMENTSMaterials for this chapter are based in large partOn the products of workshops and assessments car-

ried out by the

Central Great Plains Assessment TeamDennis Ojima,Colorado State University, co-chairJill Lackett,Colorado State University, co-chairLenora Bohren,Colorado State UniversityAlan Covich,Colorado State UniversityDennis Child,Colorado State UniversityCeline Donofrio,Colorado State UniversityWilliam Easterling, Pennsylvania State UniversityKathy Galvin,Colorado State UniversityLuis Garcia,Colorado State UniversityJim Geist,Colorado Corn Administrative CommitteeMyron Gutmann,University of Texas at AustinTom Hobbs,Colorado State University/Colorado

Division of WildlifeTim Kittel,National Center for Atmospheric

ResearchMartin Kleinschmit,Center for Rural AffairsKathleen Miller, National Center for Atmospheric

ResearchJack Morgan,USDA Agricultural Research ServiceGary Peterson,Colorado State UniversityBill Parton,Colorado State UniversityKeith Paustian,Colorado State UniversityRob Ravenscroft,Rancher, NebraskaLee Sommers,Colorado State UniversityBill Waltman,Natural Resources Conservation

Service

Northern Great Plains Assessment TeamGeorge Seielstad,University of North Dakota,co-

chairLeigh Welling,University of North Dakota,co-chairKevin Dalsted,South Dakota State UniversityJim Foreman,Ten Sleep,WyomingBob Gough,Intertribal Council on Utility PolicyJames Rattling Leaf, Sinte Gleska UniversityJanice Mattson,Precision Agriculture Research

AssociationPatricia McClurg,University of WyomingGerald Nielsen,Montana State UniversityGary Wagner, Climax,MinnesotaPat Zimmerman,South Dakota School of Mines and

Technology

Southern Great Plains Workshop SteeringCommittee

Robert Harriss,Texas A&M University (currentlyNCAR), chair

Tina Davies,Houston Advanced Research CenterDavid Hitchcock,Houston Advanced Research

CenterGerald North,Texas A&M University

Southwest-Rio Grande Workshop SteeringCommittee

Charles Groat,University of Texas-El Paso (currentlyUSGS), chair

Honorable Silvestre Reyes,US House ofRepresentatives,Texas,honorary chair

217


CHAPTER 8

POTENTIAL CONSEQUENCES OF CLIMATE VARIABILITY AND CHANGE FOR THE WESTERN UNITED STATESJoel B.Smith1 , Richard Richels2,3, and Barbara Miller3,4


Chapter Summary

Physical Settings and Unique Attributes


Ecological Context


Key Issues

Water Resources

Natural Ecosystems

Crop Productivity

Ranching

Coastal Resources

Tourism and Recreation

Additional Issues

Mining

Air Quality

Health Effects



Literature Cited

Acknowledgments

1Stratus Consulting,Inc.;2EPRI; 3Coordinating author for the National Assessment Synthesis Team;4World Bank 219


• In the 20th century, temperatures in the West rose2 to 5˚F.

• The region generally became wetter, with someareas having increases in precipitation greaterthan 50%. A few areas,such as portions ofArizona,became drier and experienced moredroughts. The length of the snow season inCalifornia and Nevada decreased by about 16days from 1951 to 1996.


• During the 21st century, temperatures are verylikely to increase throughout the region,at a ratefaster than that observed,with the Hadley andCanadian General Circulation Models (GCMs)projecting increased temperatures of about 3 toover 4˚F by the 2030s and 8 to 11˚F by the2090s.

• The two climate model scenarios projectincreased precipitation,particularly during win-ter, and especially over California. However, partsof the Rocky Mountains are projected to getdrier and the Canadian model projects most ofthe region getting drier by the 2030s. Otherchanges in climate are possible and there is somechance that that climate over much of the Westcould become generally drier during the 21st cen-tury.

• Under the Hadley and Canadian scenarios, runoffis estimated to double in California by the 2090s,though the climate models also suggest thepotential for more extreme wet and dry years inthe region.

• This chapter considers the effects of warmer andwetter conditions,based on the climate modelscenarios used in this Assessment. It also consid-ers a scenario of generally warmer and drier con-ditions.


The West is characterized by variable climate,diverse topography and ecosystems,an increasinghuman population,and a rapidly growing andchanging economy. Western landscapes range fromthe coastal areas of California,to the deserts of theSouthwest,to the alpine tundra of the Rocky andSierra Nevada Mountains. Since 1950,the region’spopulation has quadrupled,with most people nowliving in urban areas.The economy of the West hasbeen transformed from one dominated by agricul-ture and resource extraction to one dominated bygovernment,manufacturing,and services. Nationalparks attract tourists from around the world. Theregion has a slightly greater share of its economy insectors that are sensitive to climate than the nationas a whole;these include agriculture,mining,con-struction,and tourism,which currently representone-eighth of the region’s economy.

As a result of population growth and development,the region faces multiple stresses. Among these areair quality, urban sprawl,and wildfires. Perhaps thegreatest challenge,however, is water, which is typi-cally consumed far from where it originates.Competition for water among agriculture,urban,recreation,environmental,and other uses is intense,with water supplies already oversubscribed in manyareas.

The combination of continued development of theWest and climate change is likely to introduce somenew stresses, exacerbate some existing stresses,andease other stresses.

220


Key Findings

Water Resources• The potential for flooding is very likely to

increase because of earlier and more rapid melt-ing of the snowpack and more intense precipita-tion. Even if total precipitation increases sub-stantially, snowpacks are likely to be reduced.However, it is possible that more precipitationwould also create additional water supplies,reduce demand and ease some of the competi-tion among competing uses.

• In contrast,a drier climate is very likely todecrease water supplies and increase demand forsuch uses as agriculture, recreation,aquatic habi-tat,and power, thus increasing competition forscarcer supplies.

• Improved technology, planting of less water-demanding crops,pricing water at replacementcost,and other conservation efforts can helpreduce demand and vulnerability to drought.Advanced planning for potentially larger floods isneeded to reduce flood risks.

Natural Ecosystems• Vegetation models estimate that under wetter

conditions there is likely to be an increase in bio-mass,a reduction in desert areas,and a shifttoward more woodlands and forests in manyparts of the West. However, should the climatebecome drier, forest productivity would likely bereduced and arid areas would expand. It is possi-ble that fire frequency could increase whetherthe region gets wetter or drier.

• Human development of the West has resulted inhabitat fragmentation,creation of migration barri-ers such as dams,and introduction of invasivespecies. The combination of development,pres-ence of invasive species,complex topography,and climate change is likely to lead to a loss ofbiodiversity in the region. However, it is proba-ble the mountains will enable some species tomigrate to higher altitudes. It is also possiblethat some ecosystems,such as alpine ecosystems,would virtually disappear from the region.

• Human interventions to aid adaptation byspecies will be challenging,but reducing thepressures of development on ecosystems andremoving barriers to migration could be themost effective strategies.

Agriculture and Ranching• Higher CO2 concentrations and increased precip-

itation are likely to increase crop yields anddecrease water demands while milder wintertemperatures are likely to lengthen the growingseason and result in a northward shift in crop-ping areas. However, there is some chance thathigher temperatures will inhibit growth of cer-tain fruits and nuts that require winter chilling,and changes in the rainfall and humidity canharm some crops,such as grapes, by increasingpotential for disease.

• It is possible that higher temperatures andincreased precipitation will increase forage pro-duction and lengthen the growing and grazingseason for ranching,but flooding and increasedrisk of animal disease can adversely affect theindustry.

• Increasing crop diversity can improve the likeli-hood that some crops will fare well under vari-able conditions,while switching to less water-demanding crops and improving irrigation effi-ciency would conserve water. Improved weatherforecasting could aid farmers in selecting crops,timing harvests,and increasing irrigation efficien-cy;and aid ranchers in timing cattle sales andbreeding.

Tourism and Recreation • Higher temperatures are very likely to result in a

longer season for summer activities such as back-packing,but a shorter season for winter activi-ties,such as skiing. Ski areas at low elevationsand in more southern parts of the region arevery likely to be at particular risk from a shorten-ing of the snow season and rising snowlines.

• Adaptation strategies for tourism and recreationinvolve diversification of income sources.

221

Chapter 8 / The Western United States

222


PHYSICAL SETTING ANDUNIQUE ATTRIBUTESThe West region spans from California to the RockyMountains in Colorado and south to the Mexicanborder. The region contains 19% of the land areaand 17% of the population in the United States. Onaverage,the West has low precipitation,althoughsome parts are quite wet. It also has some of thegreatest variance in topography and climate in thelower 48 states. The West includes the lowest point(Death Valley, which is 282 feet below sea level) andthe highest point (Mt.Whitney, 14,494 feet abovesea level) in the lower 48 states. Among its majormountain ranges are the Sierra Nevada,the Wasatch,and the Rockies. The region also contains the GreatBasin in Nevada and Utah;in which most of therivers do not run to the sea. Especially because ofits varied topography, climate zones in the Westrange from deserts to alpine.

SOCIOECONOMIC CONTEXTThe West underwent a dramatic transformation in the20th century in its human population,economy, andlandscape. Since the middle of the century, the popu-lation has increased fourfold (see Figure 1). Althoughmore than two-thirds of the West’s 46 million peoplelive in California,more recently the intermountainstates have become one of the fastest-growing areasin the nation. Most people in the West live in urbanareas. To the large cities of California — SanFrancisco,Los Angeles,San Diego,and Sacramento —the West has now added Denver, Salt Lake City,Albuquerque,Phoenix,and Las Vegas as major metro-politan areas (see Figure 2). Thus,once predominant-ly rural states are now among the most urban in thecountry. The regional population is projected togrow by about one half, reaching 60 to 74 millionpeople, by 2025 (NPA Data Services,Inc.,1999).

The economy of the West has been transformed fromone dominated by agriculture and resource extractiveindustries in the 19 th century to one dominated bygovernment,manufacturing,and services such astourism. Figure 3 displays the relative value of allgoods and services produced in the region in 1996.About 11% of the region’s output is currently in sec-tors considered relatively sensitive to climate,includ-ing agriculture,mining,construction,and the tourismrelated sectors of hotels and amusement/recreation.This share of the region’s output in these sectors isprojected to increase to 12% by 2045,mainly becauseof increases in tourist related activities,but alsobecause of increases in agricultural services. Theshare of total output in agriculture is projected todecrease,although the total value of agricultural pro-duction is projected to increase (US BEA,1999a).

ECOLOGICAL CONTEXTAlthough much of the West is semi-arid grassland orshrubland,the region’s diverse ecosystems containalpine tundra,coniferous and mixed forests, chapar-ral, wetland,and coastal and estuarine areas (USGS,1993).

POTENTIAL CONSEQUENCES OF CLIMATEVARIABILITY AND CHANGE FOR THE WESTERNUNITED STATES

Figure 1: The West’s population grew from less than 10 million in1940 to 46.2 million in 1998 (US Census Bureau, 1998). California’spopulation mushroomed from less than 7 million in 1940 to morethan 33 million in 1998 (California Trade and Commerce Agency,1997; California Department of Finance, 1998). Although more thantwo-thirds of the West’s population lives in California, in recentdecades, the intermountain states have become the fastest-growingin the nation. For example, Arizona’s population grew from 1.3 mil-lion in 1960 to 4.5 million in 1998 (CLIMAS, 1998). Six of the 10fastest-growing states in the US are projected to be in this region,with Arizona, Nevada, and Utah being the fastest. California’s pop-ulation is projected to rise from its 1998 level of 33 million to about45 million (NPA Data Services, Inc., 1999). See Color PlateAppendix.

Historic and Estimated Population for the West

223


larly in the Southwest, receive most of their precipi-tation from summer monsoons,highly variable win-ter precipitation provides most of the annual runoffin the rest of the region (Bales and Liverman,1998).

Water and land in the West have been substantiallyaltered by people. In the West, water is typicallyconsumed far from where it originates. ForCalifornia users, water is extracted from natural sys-tems primarily in the northern part of the state,andfrom the Colorado River. More than one-third of thewater Arizona uses is from the Colorado River (CLI-MAS,1998). Western water tends to be subsidized(by the federal government and states) and sold toconsumers at prices effectively below what it coststo make supplies available. Irrigation is the majorconsumer of Western water (see Figure 4).

The federal government owns more than half of theland in the West,including 83% of Nevada. Most ofthe federally owned land is managed by the Bureauof Land Management, Forest Service, Park Service,and Department of Defense (Riebsame,1997).Indian reservations are scattered throughout theregion,and are most concentrated in Arizona,wherethey comprise about one-third of the state’s landarea (estimated based on Riebsame,1997). Betweentwo-thirds and three-quarters of the land in the Westis used for pasturelands, agriculture,and forests,with ranching using most of that land (USGS,1999).However, the amount of land used for farming(including cultivated and non-cultivated land such aspastureland) in the West decreased by 8% between1992 and 1997 (USDA,1997).

Continued population and economic growth couldresult in more demand for water, wood products,and minerals;more roads,and conversion of land tourban uses (which could increase runoff and,incoastal areas,vulnerability to sea-level rise);poten-tially more automobile emissions (although thisdepends on future technology and transport prac-tices);and increased demands for recreation. All ofthese could put more pressure on the remainingundeveloped areas. However, protecting open spacecould ease the current pressures of development onecosystems and enhance the ability of species tocope with climate change.

CLIMATE VARIABILITY AND CHANGEThe West experiences great temporal and spatialvariation in precipitation and temperature.Temperature regimes range from hot desert environ-ments to cold alpine environments. Precipitationranges from up to 40 inches per year in northernCalifornia to less than 10 inches in the deserts ofNevada,southeastern California,and westernArizona. Although many parts of the region,particu-

Urban Population Growth in the West

Figure 2: Over 93% of California’s residents live in cities, includingSan Francisco, Los Angeles, San Diego, and Sacramento, and theirsurrounding metropolitan areas. In intermountain areas, populationgrowth is also largely concentrating in cities, such as Denver, SaltLake City, Albuquerque, Phoenix, Las Vegas, Santa Fe and Provo.Much of the future population growth is expected to occur in urbanareas. Source: NPA Data Services, 1999. See Color Plate Appendix.

Figure 3: The West produces 18% of US Gross National Product. The region has a slightly greater share of its economy in relativelyclimate-sensitive sectors such as agriculture, mining, construction,and tourism, than the nation as a whole. While 1.8% of the nation’seconomic output is from agriculture (which includes forests andfisheries), 2.0% of the West’s economic output is from the agricul-ture sector. The West has 4.1% of its gross product from hotels,amusement/recreation, restaurants, and museums, which arestrongly affected by tourism, while the nation as a whole has 1.6%(US BEA, 1999a). With its Gross State Product of $962 billion,California comprises 72% of the total Regional Product of $1.3 tril-lion in 1996 (US BEA, 1999a). Ranked as a nation, California wouldbe the seventh largest economy in the world (California Trade andCommerce Agency, 1997). See Color Plate Appendix.

The Relative Value of EconomicActivitiy in the West

Over the 20th century, annual precipitation overmost of the region generally increased 10 to 40%.However, precipitation in the Central Valley ofCalifornia,southeastern California,south-centralUtah,northeastern Arizona,and western Coloradodecreased and some areas have experienced moredrought (Karl et al.,1990;USHCN, 1999). Thelength of the snow season decreased by about 16days from 1951 to 1996 in California and Nevada,and stayed about the same elsewhere (DavidEasterling,National Climatic Data Center, personalcommunication,2000). Since the late 1940s,snowmelt has come earlier in the year in manynorthern and central California river basins(Dettinger and Cayan,1995). The proportion ofannual precipitation from heavy storm events hasincreased in the 20 th century (Karl and Knight,1998).

The region is quite vulnerable to climate variability,as the 1998 El Niño event demonstrated,particularlyin California. El Niño storms during February 1998brought as much as three times the average rainfallfor the month,causing numerous deaths in additionto damages to homes,businesses, roads,utilities,andcrops (Willman,1998). On the other hand,anadvanced forecast for El Niño resulted in many pro-tective measures being undertaken (see Figure 5).

With its complex topography, developing reliableprojections of climate change in the West is particu-larly difficult. General Circulation Models (GCMs)tend to be least reliable projecting changes incoastal areas and in mountains,two features preva-lent in the West. However, it is possible to developGCM-based scenarios that give an indication of howincreased greenhouse gas concentrations couldchange the climate.The limitations of GCMs are dis-cussed in more detail in Chapter 1.

Average annual outputs from the Hadley andCanadian GCMs are shown in Figure 6. The Hadleymodel projects a 3.8˚F (2.1˚C) winter warming anda 3.1˚F (1.7˚C) summer warming by the 2030s1

over 1961-1990 temperatures and an 8.8˚F (4.9˚C)winter and an 8.3˚F (4.6˚C) summer increase bythe 2090s. The Canadian model projects more win-ter warming,with a 4.8˚F (2.7˚C) winter and a2.5˚F (1.4˚C) increase in summer temperature bythe 2030s and a 12.8˚F (7.1˚C) winter and 7.7˚F(4.3˚C) summer increase by the 2090s (NCAR,1999a).

Both models project a doubling of winter precipita-tion over California. However, the Hadley and

In many areas of the West,paleoclimatic data sug-gest that on some occasions droughts and floodswere more extreme over the past few thousandyears than was observed during the 20 th century(Bales and Liverman,1998). Since 1900,tempera-tures in the West have been rising,with increases of2 to 5˚F per 100 years in all areas except southernColorado, western New Mexico,and eastern Arizona(See Climate Chapter). Averaged over the region,the number of days with high temperatures over90˚F increased in the 20th century while days belowfreezing decreased (David Easterling,NationalClimatic Data Center, personal communication,1999).224


Relative Water Use in the West

Figure 4: In 1995, 87% of the water consumed in the West was forirrigation (Solley et al., 1998; see Figure 4). However, water usefor irrigation has declined slightly since 1980, while municipaluses have grown (Diaz and Anderson, 1995). For example, agri-culture accounts for 81% of all water used in Arizona, down from93% in 1963, while municipal demand currently accounts for 14%of water used, up from 5% in 1963 (CLIMAS, 1998). In addition,irrigated land in the region fell by 8% from 1982 to 1992, althoughacreage may have increased in recent years (USDA, 1997). Totalwater use in the region appears to have been declining since 1980(Templin, 1999). See Color Plate Appendix.

Figure 5: The 1997-1998 El Niño had quite strong effects in theWest, with particularly large winter precipitation events. Theheavy precipitation lead to such localized consequences asflooding and landslides. See Color Plate Appendix.

El Niño and Events 1997-1998

1The results for the 2030s are an average for 2025-2034.

Canadian models also show the potential fordecreased precipitation in some parts of the RockyMountains. The Canadian model shows no changein summer precipitation,while the Hadley modelprojects that summer precipitation would decrease.

The models do not project a significant change ininterannual variation of precipitation. Should inter-annual variation of precipitation increase,therewould be more extreme wet years and moreextreme dry years. It is likely that many areas in theWest could have wetter winters and drier summers.It is very unlikely that changes in precipitation willbe uniform across the West;some areas will likely bewetter while it is possible that others will be drier.Wet periods will very likely be followed by dry peri-ods because, even with climate change,there willstill be variability — seasonally, from year to year,and from place to place.

California has experienced relatively less sea-levelrise than the eastern United States because manyareas are being uplifted by moving of geologicalplates (Neumann et al.,2000). The coast south of LaJolla,California has been experiencing a relative sea-level rise of approximately 8 inches (20 cm) percentury;the coast from Los Angeles to San Franciscohas had a 0 to 6 inches (15 cm) per century of sea-level rise;and the coast in far northern Californiahas experienced a relative reduction in sea level of2 to 6 inches (5 to 16 cm) per century. TheIntergovernmental Panel on Climate Change esti-mates that sea level will rise 6 to 37 inches (15 to95 cm) by 2100 (Houghton et al.,1996),whichwould result in net sea-level rise for the entireCalifornia coast.

KEY ISSUES

The key issues in the West involve those systemsthat are sensitive to climate and,in a number ofcases,are already stressed by current developmentpatterns. All of these systems will be affected by cli-mate change.

1. Changes in seasonality and amount of waterresources

2. Plant and animal changes in natural ecosystems3. Changes in agricultural crop productivity4. Precipitation and forage changes for ranching5. Sea-level rise effects on coastal resources6. Changes in tourism and recreation

1. Water Resources

The more than fourfold increase in the populationof the West since the middle of the 20th century hasdramatically changed the use of natural resources inthe West and imposed stresses on these resources.One of the more stressed resources is water.Although agricultural water use is declining,2 watersupplies are tight because of growth in environmen-tal, municipal,and industrial demands and couldbecome tighter as the population and economy con-tinue to grow and unresolved water rights claimsare settled. For example, over the last ten years,California consumed more than its normal yearapportionment of Colorado River water, but surpluswater and water unused by Arizona and Nevada wasavailable to meet California’s needs (US Bureau ofReclamation,1997;US Bureau of Reclamation,1999).3 Meanwhile, rapidly growing urban areassuch as Las Vegas are demanding more water. Inaddition,many aquifers are being depleted at ratesfaster than their recharge,and high-volume ground-water mining has caused land subsidence (sinking)and fissuring (cracking).

225


Figure 6: Changes in annual mean temperature and precipitationfor the West as projected by the Hadley and Canadian models com-pared to 1961-90 Base Period.

Changes in Annual Mean Temperature and Precipitation

2 Although total water use for irrigation is declining, agricultural pro-duction is sensitive to changes in precipitation and subsequentchanges in water allocation. For example,in 1991,during the fifth yearof a drought, water supplies to California agriculture were severely cur-tailed.Overall economic losses were approximately $400 million –about 2% of total agricultural revenues.In spite of the drought, agricul-tural revenues in 1991 reached an all-time high (Gleick and Nash,1991).3Use of Colorado River water is allocated between the Upper Basinand the Lower Basin.The Lower Division states of California,Arizona,and Nevada are guaranteed a delivery of 75 maf (million acre feet) ineach 10 year period.Also,the Upper Division states (Colorado,NewMexico,Utah,and Wyoming) are to supply one-half of the waterrequired to be delivered by treaty to Mexico,that is,0.75 mafy (millionacre feet per year),if waters over and above the quantities of useapportioned to the Upper Basin (7.5 mafy) and the Lower Basin (8.5mafy) are insufficient. (House Document No.717,1948) Nevada’sapportionment of Colorado River water is 0.3 mafy plus 4 percent ofthe surplus water made available.The Upper Basin states receive thefollowing shares:Arizona 0.05 mafy, Colorado 3.855 mafy, Utah 1.713mafy,Wyoming 1.043 mafy, and New Mexico 0.84 mafy (NYT, 1999).

226


Brown (2000) forecast that by 2040 net water with-drawals in the region will increase,with with-drawals for domestic and public use increasingmost,and irrigation withdrawals declining slightlyexcept in the Upper Colorado Basin. As water useshifts from agriculture to municipal uses,the abilityto reduce withdrawals during droughts declines.

It has become increasingly difficult to build any sig-nificant new water resources infrastructure becauseof economic,environmental,and social constraints.In addition,institutional factors such as water rights,local planning and zoning,and regulations influenceand can limit the nature and level of response thatwater managers can make to changes in supply ordemand. Reserved and Native American waterrights claims are senior to those of many otherwater consumers,and many of these rights are notcurrently being exercised (see box on NativeAmerican water claims).

Because of its semiarid climate, water supplies inthe West are considered to be more vulnerable toclimate change than water supplies in other regions(Gleick,1990;Hurd et al.,1999a). Detailed hydrolog-ic modeling conducted for the western US projects

a significant change in snowfall and snowmeltdynamics because of higher temperatures. Risingtemperatures are likely to shorten the snowpackseason by delaying the autumnal change from rain-fall to snow and advancing the spring snowmelt. Alarger proportion of winter precipitation in moun-tainous areas is also very likely to fall as rain ratherthan snow, even if overall precipitation amounts donot change. McCabe and Wolock (1999) found thatunder the two GCM scenarios,April 1 snowpack inthe major western mountain ranges would bereduced, except that under the Hadley scenario,snowpack in the Rocky Mountains would have littlechange.4 Peak runoff is very likely to occur earlierin the year (see Figure 7) (Gleick and Chalecki,1999). Jeton et al.(1996) found that snowmeltwould occur more than two weeks earlier than cur-rently in the East Fork of the Carson River andNorth Fork of the American River in the SierraNevada under a 2.2˚C (4˚F) warming,which theHadley and Canadian scenarios suggest would occurby the 2030s.5

Wolock and McCabe (1999) projected changes inrunoff for the region using the Hadley and Canadianclimate models (see Table 1). They estimate thatCalifornia runoff will increase by the 2030s byabout three-fifths and double by the 2090s.6 Theirstudy projected small changes in runoff in the restof the West by the 2030s,and no change to approxi-mately 30% increases in runoff outside of Californiaby the 2090s. The changes in runoff for the areasoutside California are not considered to be statisti-cally significant because there is so much variancein year to year runoff. Soil moisture under both sce-narios is projected to increase,but in many loca-tions outside of California conditions could be drierduring some periods,particularly in the summer(NCAR,1999b).

These changes in runoff have important conse-quences for water management. Any changes inrunoff timing or variability could possibly causeproblems (Gleick,1987). Earlier spring runoff islikely to increase risk of spring flooding,complicateseasonal allocation schedules,and create problemsfor matching supply and demand and meeting envi-

Figure 7: Natural runoff (solid line) peaks in May as winter snowmelts. Under conditions of climate change (dashed line), runoffpeaks earlier and higher, but is lower in the summer. Source: Gleickand Chalecki, 1999.

Hypothetical Change in Runoff for a WesternSnowmelt Basin

4The article does not state at what altitude snowpack is measured.5Jeton et al.(1996) also found that total annual flow was insensitive tochanges in temperature and much more sensitive to changes in precip-itation.6In contrast,Miller et al.(1999a,1999b) found that total streamflow inthe Russian River in northern California,which is not snowmelt driven,would not change significantly under the Hadley 2090s scenario,butpeak runoff may occur one month earlier because of a potentialchange in winter storms.In contrast,snowmelt driven streamflow inthe Sierra Nevada would likely happen earlier and peak streamflowwould rise.Miller et al.(2000) found that the American River in theSierra Nevada,which is snowmelt driven,showed both an increase inmagnitude and earlier peak flow (see also Hay et al.,2000).

227


If there is reduced or even only small increases ofprecipitation, runoff is very likely to be reduced. Inaddition,both groundwater recharge and reservoirsupplies are very likely to be reduced as higher tem-peratures increase evaporation (Wilkinson andRounds,1998a).

Reduced runoff, particularly if combined with high-er demands due to hotter and drier conditions,would very likely make allocation of water suppliesa more critical issue for the West. It is likely thatinstream uses such as hydropower and recreationwould be among those most affected by a reductionin runoff. It is also likely that urban and industrialusers would be less vulnerable to supply reductions.Hurd et al.(1999b) found that urban and industrialusers of Colorado River water would have verysmall reductions in supplies if runoff is reduced. Ingeneral,it is very likely that those with more juniorwater rights claims (those who receive their alloca-tions after the senior claims are met) would be atgreatest risk should runoff decline (Miller et al.,1997). In addition it is possible that NativeAmericans will more fully exercise their rights towater (see box). Furthermore,during droughtsthere is likely to be increased dependence ongroundwater, causing increased overdraft,subsi-dence,and reduced baseflow of rivers. On theother hand,it is possible that drier conditions wouldresult in a decrease in flood potential and mudslidesin California.

With less runoff, water quality is likely to decline ifstronger pollution control measures are not under-taken. Higher temperatures alone would decreasedissolved oxygen levels in water while lowerstreamflow would concentrate pollutants. Lowerflows in the Colorado River are likely to result inincreased salinity levels,unless additional steps aretaken to control the problem (Gleick and Nash,1991). Lower lake levels could also increase waterquality problems. For example,salinity concentra-tions in the Great Salt Lake are likely to increasewith lower lake levels (Grimm et al.,1997).

ronmental in-stream flow requirements in the sum-mer. It is likely to be problematic for the currentreservoir system to store earlier spring runoff foruse in the summer unless new operating rules andregimes are implemented (Lettenmaier and Sheer,1991),and it is not clear that such a change wouldbe sufficient to reduce spring flooding and increasesummer supplies. This may be especially true inCalifornia,where both climate models used in thisAssessment show a substantial increase in runoff,particularly in the winter. In addition,more intenseprecipitation events (such as the extreme event inLas Vegas on July 8,1999 that caused extensiveflooding in the city) could increase flooding. Therisk of increased flooding is exacerbated by contin-ued urban development,which increases surfacerunoff during storms. Development in floodplainsand expansion of areas that could be floodedbecause of increased runoff could result in morepeople and property at risk to the effects of climatechange. In addition,higher runoff can increasemudslides.

On the other hand,it is possible that increasedrunoff would create more water supplies for theWest. Presumably, this could contribute to an easingof many current stresses on the water managementsystem because there would be relatively morewater available for users. A wetter climate wouldalso likely reduce the demand for surface water andgroundwater for such purposes as ir rigation andwatering lawns.

There is some chance that higher runoff could easewater quality problems although it could also resultin more runoff of pollutants from farms and streets,which can degrade water quality. It is likely thathydropower production would increase with morerunoff. However, earlier runoff is likely to result inmore electricity production in winter time,whendemand for heating is very likely to be falling,andless electricity production in summer when demandfor cooling is very likely to be rising.

Table 1: Estimated Changes in RunoffCurrent and Estimates Changes in Runoff from the Canadian and Hadley Models (mm)

Region Historical Runoff Change in Annual Runoff Change in Annual Runoff1961-90 (mm/yr) 2025-2034 (mm/yr) 2090-2099 (mm/yr)

Canadian Hadley Canadian Hadley

Upper Colorado 43 -15 3 2 28

Lower Colorado 2 -1 6 0 33

Great Basin 21 -1 4 16 29

California 232 60 63 320 273

Adaptation Options Although building additional flood controls or stor-age infrastructure to address the need to store earli-er runoff for the summer may bemore attractiveunder climate change,environmental and cost con-straints could serve as impediments. Where bothlocal and imported supplies are available,there willbe greater flexibility to deal with changes in watersupply availability. If groundwater supplies aremaintained as a buf fer against drought,local areasare likely to have better coping ability.

Adaptation to potentially increased demand andreduced supply may focus on the demand side ofwater use. Here too,the development path for theWest is critical. Should the increased populationcontinue to use water at the same or an increasingrate, agriculture water allocations could be furtherreduced.As noted above,this can make it more diffi-cult to reduce demand during droughts.

One source of adaptation lies in changing waterpricing structures. Pricing water closer to itsreplacement cost would discourage wasteful uses.While market-based solutions would increase effi-ciency, it is possible there will be equity problems:users with limited resources,such as the poor andsome farmers,may have to cut back on water usemore than others.

Water transfers (between users and across riverbasins) will almost certainly play some role inaddressing future water demand. These transfers

228


include water savings derived from system enhance-ment measures such as canal lining and other wastereduction measures,and transfer of water currentlyused in agriculture for use in urban areas. In addi-tion,institutions to manage groundwater quantityand quality may need to be strengthened (Knox,1991).

The efficiency of municipal and industrial wateruses can be significantly improved. Increased appli-cation of conservation technologies such as ultralow flush toilets and landscaping practices such asxeriscaping can reduce the growth of urbandemand for water and lower the vulnerability ofurban areas to drought. Use of treated effluentcould be increased (Wong et al.,1999).Municipalities near the ocean can also reduce waterdemand by desalting seawater, which is an expen-sive option. For example,Santa Barbara recentlybuilt a desalinization plant.

Increasing flood storage or flood control measuresis likely to be an adaptation to increased risk offlooding. However, flood control management isshifting away from reliance on physical structures toeffective management of floodplains,includingrestricting development,using wetlands,and tryingto re-create the ability of rivers to spread floods toavoid concentrated downstream impacts (Wong etal.,1999). These adaptations may be ef fective ifimplemented in response to climate change,butwould be more effective if implemented in anticipa-tion of climate change. If annual precipitation

Native American Water Claims

Indian water rights remain an unresolved and important issue for water allocation in the West in a number ofcases. Under the legal doctrine established by the 1908 Winters case (Winters v. United States [207 US 564(1908)]),Indian tribes have reserved water rights that could amount to 45-60 million acre-feet (Western WaterPolicy Review Advisory Commission,1998). However, the vast majority of those claims have never been clear-ly quantified or developed for the benefit of the tribes. In many cases,non-Indian water users have alreadyfully appropriated and used the sources of water potentially available to satisfy tribal rights. Tribal efforts toprotect and develop their water rights have encountered resistance from other water users and state waterauthorities. There is substantial ongoing litigation (approximately 60 pending cases as of 1995) and about 20ongoing negotiation efforts aimed at achieving settlements of Indian water rights claims. The low availabilityof financial resources in certain cases makes it difficult for tribes to develop their water rights or to contestcompeting uses that interfere with Indian water rights,including instream flow rights for fishery purposes.

Historically, tribes often made significant concessions of their reserved water rights to obtain water develop-ment on reservations. Yet,many Indian irrigation projects have fallen into disrepair for lack of project fund-ing. Some projects such as the Navajo Irrigation Project remain uncompleted,and others such as the Animas-La Plata Project have yet to be built despite Congressionally approved water settlements. Recently, theSecretary of the Interior promoted a comprehensive dialogue on a government-to-government basis withtribes in an attempt to develop a water rights negotiation process that responds to the concerns of tribes.

increases,but summers become hotter and drier,there is likely to still be a need for additional stor-age to provide more water in the summer or fordemand reduction measures to lessen the need forwater in the summer.

2. Natural Ecosystems

The wide diversity of natural ecosystems in theWest ranges from low-elevation deserts to alpinetundra (see Figure 8). In addition,productivityvaries considerably. Most of the West is grassland,shrubland/grassland,and desert shrubland. Themountains contain coniferous forests, woodlands,deciduous forests (mostly aspen),and mixedforests. California has a wide diversity of ecosys-tems,including mostly coniferous forests in thenorth and in the Sierras,oak savanna and chaparralalong the central coast,and shrubland and grass-land along the southern coast and interior. Thecentral and southern Rocky Mountains are domi-nated by ecosystems associated with mountains:alpine,coniferous forests interspersed with grass-lands,and,at lower elevations, woodlands. Thevery dry environments in the Great Basin supportshrublands,some grasslands,and deserts. The wet-ter parts of the Great Basin support woodland veg-etation (USGS,1993). Aquatic habitats range fromcool mountain to desert streams and rivers,includ-ing reservoirs which have substantially altered theaquatic ecology of the West. In addition, wetlandsin the west,particularly in arid areas,are importanthabitat for endangered species, fish rearing,andmigratory waterfowl.

With this wide diversity of ecosystems and topogra-phy comes a wide diversity of species,many ofwhich are in isolated habitats. California’s climatezones,from coastal to desert to alpine regions,sup-port a wide variety of plants and animals,as doesthe area near the New Mexico-Arizona-Mexico bor-ders and Utah,with its deserts,canyons,and alpinepeaks (Wilkinson and Rounds,1998b;US EPA,1998a and b).

Development has taken its toll on the naturalecosystems of the region. Dams and reservoirshave altered free-flowing streams, numerous plantand animal species have been eliminated orreduced to low numbers,and agriculture andranching have transformed lowland ecosystems. Bysome estimates,90% of California’s wetlands havedisappeared (Wilkinson and Rounds,1998b). All ofthis alteration has made natural ecosystems vulner-able to invasion by hundreds of non-native species.

229


California contains more threatened and endangeredspecies (257) than any of the other lower-48 states(US Fish and Wildlife Service,1999) and is secondhighest in rate of species extinction (The NatureConservancy, 1999). Myers et al.(2000) considerthe California Floristic Province as one of the 25“hotspots”in the world that have exceptional diver-sity of species and are experiencing exceptional lossof habitat.

The rise in population has resulted in more urbandevelopment and development into wooded areaswhich among other things has exposed human set-tlements to wildfires (see box on fires). Fire is a nat-ural part of the ecology of the West. However, firesuppression has resulted in an unnatural increase inthe density of vegetation,thereby making the land-scape more susceptible to severe fires. In addition,some invasive species,such as cheatgrass,haveincreased fire frequency, while species such as starthistle and Tamarix have reduced water suppliesand increased flooding (Chapin et al.,2000).

Figure 8: Currently the West has a large diversity of ecosystems.Under the two climate change scenarios, the area in arid and grass-land ecosystems would decrease and the area in forest ecosystemswould increase. See Color Plate Appendix.

Ecosystem Models

The Diverse Ecosystems of the West

Current Ecosystems

Climate change is projected to causemajor changes in vegetation distribu-tion during the 21st century. Overall,the model scenarios project increas-es in grasslands, woodlands, andforests in the West, and a loss ofdesert vegetation. The upper mapshows potential vegetation types inthe West (the vegetation that wouldnaturally flourish in the absence ofhuman activity given today’s climate),while the two maps below showmodel-projected scenarios for futurevegetation shifts in the face of cli -mate change.



230


Vegetation

Under both the Canadian and Hadley climate scenar-ios,using the VEMAP biogeography and biochem-istry models,biomass is projected to increase andvegetation to shift from deserts and grasslands towoodlands and forests in many parts of the region.Forests are projected to expand in California,Utah,and Colorado,mostly in the mountains. Nevada,

northern Arizona,and western New Mexico are pro-jected to see a shift toward shrub woodland andsavanna woodland,while southwestern Arizona andsoutheastern California are projected to shift fromarid lands to grasslands (see Chapter 2: Vegetationand Biogeochemical Scenarios:Future Vegetation).Across the West,a wetter climate is likely to increaseforest productivity, including shifting some coniferforests to broadleaf forests,although there could still

Fire in the West

The rise in population in the West has resulted in more development into wooded areas and increased expo-sure to fire risk,which was already high (see Figure 9). For example,there were major fires in recent years inurban areas,including Oakland,Santa Barbara,Malibu,and Los Alamos. The Oakland fire destroyed or damagedabout six thousand structures. In addition, fire suppression,which has resulted in dense growth and invasionof non-native species such as cheat grass,have made many Western forests more vulnerable to major fires.Continued development into forested areas,along with continued suppression of fires and spread of non-native species,is likely to increase risks of severe fires.

Studies suggest there is a good chance that climate change will increase the risk of fire frequency, whetherprecipitation increases or decreases in the region. Lower precipitation renders montane forests more fire-prone. These forests are already at risk because of the massive fuel buildup and predisposition to uncontrol-lable crown fires. Torn et al.(1998) found that warmer and drier conditions could lead to a “dramatic”increase in land area burned and potentially catastrophic fires in California. Higher precipitation increasesthe fuel loads of sparse vegetation in arid areas. If interannual variability of precipitation does not decrease,wet periods will be followed by dry periods and there is a good chance fires would increase. Modeling witha dynamic global vegetation model (MC1) found that fires across the West could increase under such condi-tions. As temperatures continue to rise,so would evapotranspiration,which can lead to more drying andmore fires (Neilson and Drapek,1998). Under the Hadley and Canadian scenarios,the fire severity rating inthe West increases 10%.

Increased fire could reduce the indigenous vegetation in some cases and promote conversion to nonnativeweeds. More fire could degrade water quality because of increased runoff of sediments. Fires also add to airpollution. Should fire increase,there could be increased risks for human settlements within or close to forestsand grasslands.

The risk of fire in urban areas and in heavilyforested areas could be reduced through anumber of measures. Restrictions can beplaced on development in fire-prone areas.Building and landscape design criteria havebeen developed for fire-prone areas.Construction with nonflammable materials andinstallation of “firescape”landscape designs arealso being used in high-risk areas. Controlledburns may also need to be used as part of avegetation management strategy in urbanareas. Many of these adaptations have beenimplemented in response to urban fires such asthose in Oakland. Fires in natural areas shouldnot be suppressed to the degree that a largeamount of fuel buildup is allowed. These adap-tations should be implemented in anticipationof climate change.

Figure 9: Relative fire severity across the United States in July,1994. All of the states with high fire severity were in the West.Source: Liverman, 1998. (see http://udallcenter.arizona.edu/publica-tions/pdfs/swclimatereport-final.pdf, page 22.)

Fire Severity - July 1994

be a net increase in conifer forest cover (Neilsonand Drapek,1998). The higher temperatures,how-ever, are likely to result in many alpine areas virtual-ly disappearing from the West and being replaced bytemperate forests (see Chapter 2:Vegetation andBiogeochemical Scenarios).Note that the projectedchanges do not show steady increases in biomass inall places at all times. Under the Hadley model, veg-etation productivity declines in New Mexico andArizona by the 2030s. One model result shows thatin Colorado, forests first decrease in area by 2030,but expand by 2095 to cover an area larger thantoday.

There are a number of reasons for caution aboutthese projections. First the CO2 fertilization effecton plant growth and water use efficiency may notbe as positive as assumed in the models (Walker andSteffen,1997). Under the Canadian model,assumingno CO2 fertilization effect,biomass is projected todecline in some parts of the West (Aber et.al.,2001). Modeling conducted for this Assessment andother studies discussed in the box on fire show anincreased risk of fire in the West. Climate changecould also make conditions more favorable for pestoutbreaks and introduction and spread of invasivealien species (Dale et al.,2001). Should high levelsof air pollution continue and wind storms increase,these would be additional stresses on forests. It isalso uncertain whether transitions from one type ofecosystem to another would be smooth or involvedisruptions.

Furthermore,as climate continues to change,theCO2 fertilization effect (which increases growth andwater use efficiency) becomes saturated anddeclines,and higher temperatures would imposemore moisture stress on vegetation.

If conditions become drier, productivity of vegeta-tion is likely to decrease (Neilson and Drapek,1998). There could be a shift from forests, wood-lands,and shrublands,to grasslands and deserts.

Biodiversity

As noted above,development has resulted in frag-mentation of habitats,creation of barriers to migra-tion,such as urban areas and dams,and introductionof invasive species. This,in combination with thecomplex topography and varied climate of theregion,is likely to make it difficult for many speciesto adapt to climate change through migration. It isalso likely that development would favor the spreadof invasive and non-indigenous species becauseinvasive species are generally better suited to chang-

ing conditions. Without development,the adverseimpacts of climate change on biodiversity wouldlikely be substantially reduced.

While the mountains of the West can serve as barri-er to species migration,they also provide higher alti-tude and northern routes for migration as well asmany microclimates that can create refugia for somespecies. But,migration upslope also means migrat-ing to smaller and smaller areas of habitat,whichwould only support smaller and smaller popula-tions. As climate change continues,species migrat-ing upslope are very likely to be threatened as theirhabitats figuratively disappear off the tops of moun-tains.

The faster the rate of climate change,the greater thestress will be on many species and populations.

Terrestrial SpeciesHansen et al.(2001) found there is a slight chancethat Quaking Aspen and Engleman Spruce will notsurvive under projected climate change (however,this study did not account for the positive effects ofCO2 fertilization). Interestingly, paper birch is pro-jected to expand southward in the RockyMountains. Hansen et al.(2001) also found that ani-mal populations could change. It is possible thathigher temperatures lead to a decrease in bird andmammal populations that are currently found in theregion because they cannot tolerate higher tempera-tures. It is possible that higher temperatures couldincrease reptiles and amphibians in the southernRocky Mountains because of their greater tolerancefor heat.

Murphy and Weiss (1992) projected that a 5˚F (3˚C)warming would result in a substantial reduction inthe area of the Great Basin suitable for borealspecies. They estimated that plant species would bereduced from 305 to 254, four of nine mammalswould be lost,and 23 to 30% of butterflies living inboreal areas in the Great Basin would becomeextinct. On the other hand,there is some chancethat higher temperatures would enable some south-western desert plants to invade the Great Basin(Neilson and Drapek,1998),although such a large-scale change could take thousands of years to berealized.7

231


7In warm periods in the past,some species migrated to new locations,while others remained in the same general location (Tausch et al.,1995).

California and show signs of stress in the warmeryears (Wilkinson and Rounds,1998a). Drier condi-tions are likely to result in the loss of many smallwater bodies and aquatic ecosystems (Grimm et al.,1997).

In addition,the change in seasonality of runoff islikely to have adverse effects on many species. It isdifficult to anticipate exactly how these changes inflow magnitude and timing would affect particularspecies or flow-dependent habitats. However, somegeneral predictions can be made based on knowl-edge of species life history strategies in relation tohydrology. In general, climate-related hydrologicchanges are very likely to favor some species morethan others, resulting in decreased species diversityand altered composition of native biological com-munities. For example,it is possible that alterationsto the timing and magnitude of spring flows willfavor non-native riparian plants that would other-wise be suppressed by high runoff in spring(Kattelmann and Embury, 1996). Modified flowregimes are also very likely to affect populations ofnative fish species. For example,the distributionand abundance of the four seasonal runs of chinooksalmon native to the Sacramento River drainage thatare already in jeopardy are likely to be furtheraltered by seasonal changes in the availability ofspawning flows (Yoshiyama et al.,1996).

Observed Effects on SpeciesThe effects of climate change on species are alreadybeing observed. Parmesan (1996) found that themean location of populations of Edith’s Checkerspotbutterfly shifted northward and upward in elevationsince the beginning of the century (Figure 10). Shefound that the southern boundary moved north-ward but was unable to determine if the northernboundary moved further northward (CamilleParmesan,University of Texas,personal communica-tion.) These butterflies do not migrate;in fact,it istheir relatively sedentary nature that makes them agood choice for tracking long term trends in wildliferange shifts in response to climatic warming. Arange shift northward is a process which takesdecades. In theory, as climate change makes themost southern regions less suitable and the farnorthern regions more suitable,populations at thesouthern end of the range go extinct while newpopulations are established northward of the previ-ous boundary. However fragmentation of habitatand barriers to migration are likely to impede north-ward migration of many species, resulting indecreases in their total range.

Aquatic SpeciesAquatic and riparian ecosystems in the West are alsovulnerable to changing precipitation and runoffregimes. Wetlands may have some resiliency to cli -mate change because they currently cope with high-ly variable climate conditions (Grimm et al.,1997).While wetter conditions are likely to alleviate someexisting stresses,higher temperatures are likely toexceed the thermal tolerances of many fish speciesand lead to increased fragmentation of many coldwater fish habitats particularly in mountains (Meyeret al.,1999). It is probable that some alpine andcold water fish species will not survive in the region(Grimm et al.,1997). In addition,higher tempera-tures are likely to allow for invasions by non-nativefish species (Wagner and Barron,1998). It is alsopossible that higher water temperatures would be aproblem for salmonid populations,since these fishare near the southern end of their range now in

232


Figure 10: On this map of studied sites, the lighter triangles repre-sent extinct populations of Edith’s Checkerspot butterfly, while thedarker triangles represent present populations. The mean locationof populations of this butterfly has shifted northward by 57 miles(92 kilometers) and upward in altitude by 407 feet (124 meters)since 1900. This is an indication that climate change is alreadyhaving an affect on the some species ranges. Source: Parmesan,1996. See Color Plate Appendix.

Observed Shift in Range of Edith's CheckerspotButterfly: 1900 to 1990s

233


Sagarin et al.(1999) found that in the past 50 years,the southern invertebrates have become more com-mon and northern invertebrates have become lesscommon in the rocky intertidal community inPacific Grove,California. Both of these changesappear to be the result of higher temperatures.

Adaptation Options A number of steps could be taken to at least helpreduce some of the pressures of development onecosystems and biodiversity and even anticipate theneed for species to migrate in response to climatechange. Urban development could be managed tobetter protect riparian areas and reduce habitat frag-mentation. There could be concerted efforts to linkhabitats and even create migration corridors forspecies to migrate northward or upslope inresponse to climate change. The current trendtoward reduced land for agriculture could presentsome opportunities if abandoned lands are used forhabitat. Reducing offstream water use will also helpimprove aquatic habitats. These measures wouldneed to be implemented in anticipation of climatechange. It is not clear how effective many of thesemeasures,particularly migration corridors, would bein averting negative effects of a warmer and wetterclimate on natural systems. In addition,implement-ing these measures may be challenging,while con-tinued urban and suburban development couldresult in increased stress on ecosystems and speciesdiversity.

Table 2. Relative Share of Crop and Livestock Output in the West

State Output Percentage of Combined Crop and Animal Output

Arizona Crop 60.94% Livestock and Dairy 39.06%

California Crop 79.02% Livestock and Dairy 20.98%

Colorado Crop 33.14% Livestock and Dairy 66.86%

Nevada Crop 42.66% Livestock and Dairy 57.34%

New Mexico Crop 30.22% Livestock and Dairy 39.78%

Utah Crop 29.75% Livestock and Dairy 70.25%

Region Crop 67.74% Livestock and Dairy 32.26%

Figure 11: For most of the states in the West, the majority of value-added agriculture production comes from livestock and dairy produc-tion. However, because California’s agricultural production is dominat-ed by crops (75% of total agricultural output for the state), and becauseCalifornia dominates regional agricultural output (84% of regional cropproduction, 51% of regional livestock and dairy production), the majori-ty of the region’s total agricultural production comes from crops. Thisdifference between the dominant types of agricultural production on astate level and on a regional level highlights the heterogeneity of agri-culture in the West. Source: USDA Economic Research Service StateFarm Sector Value-Added Data; (http://www.econ.ag.gov/briefing/fbe/fi/fivadmu.htm). August 30, 1999. See Color Plate Appendix.

Relative Share of Crop and Livestock Output in the West.

rain can cause molds, ruining the grapes. Higher airtemperatures and humidity can increase risk of dis-eases that can harm vineyards. However, highertemperatures in the Sonoma and Napa Valleys since1951,which is mainly the result of nighttime warm-ing,improved the quality and yield of wines(Nemani et al.,2001).Cotton yields can also bereduced by rain at critical stages of growth.

Should the climate become hotter and drier, agricul-ture would be at particular risk. It is probable thatthe amount of water available for irrigation will bereduced substantially (Hurd et al.,1999b). Thus,agriculture could be squeezed between an increasedneed for water and less available water. If additionalirrigation water is applied,there would be increasedsalinity in soils and rivers. Rural communities wouldbe sensitive to declines in agriculture or ranching.

Estimated changes in irrigated crop yields using sce-narios derived from the Hadley and Canadian cli-mate models in the 2030s and 2090s for the“Pacific”and “Mountain” regions are displayed inTable 3. The Pacific region includes California,Oregon,and Washington,and the Mountain regioncontains all of the Rocky Mountain states. The 2030results assume a CO2 concentration of 445 parts permillion (ppm),and the 2095 results assume CO2 lev-els of 660 ppm (Francesco Tubiello,GoddardInstitute for Space Studies,personal communication,1999). The specific numerical results should betreated with caution since they include states out-side the West as it is defined here and include opti-mistic assumptions about the CO2 fertilization effectwhile not considering other effects such as pestsand disease. The results show increases in yields formany crops,but decreases for some crops such astomatoes in the Pacific and hard red spring wheat inthe mountain states. Although not shown,theresults tend to show small changes in demand forirrigation water for the major western crops,but ina few cases,significant decreases in demand. Cropproduction in the Pacific and Mountain states is pro-jected to increase (see Chapter 13:Agriculture).

Adaptation Options One strategy to adapt to the effects of climatechange is to maintain and increase the diversity ofcrop types and varieties,because diversity increasesthe likelihood of having some crops that fare wellunder variable climate conditions. For example,inCalifornia,the artichoke crop was good in 1998,butthe orange crop was devastated by freezing condi-tions. Farmers may also plant low-chill varieties ofcertain tree crops in anticipation of higher averagetemperatures. This adaptation is already under way

3. Agriculture

The total value of crop and livestock productionin the West in 1997 was $32 billion (US BEA,1999a;see Figure 11. About two-thirds of thevalue of western agriculture is from crops,withthe rest from livestock (Figure 11 and Table 2).Fruits,tree nuts,and vegetables comprise abouttwo-thirds of the value of crop production,whileseven-eighths of livestock production is frommeat animals and dairy products. The West pro-duces 17% of the nation’s agricultural output,butthree-fifths of the country’s fruits and tree nuts,almost half of the vegetables,and almost one-quar-ter of the dairy products.

Higher CO2 concentrations are likely to helpincrease crop yields and decrease water demand,although higher temperatures are also likely tohasten phenological development of crops (result-ing in reduced yields) and increase demand forwater. Higher precipitation can increase yieldsbut can also cause flooding and waterlogging ofcrops.The net effect on yield will depend on rela-tive changes in CO2 concentrations,temperature,and precipitation.

Milder winter temperatures are likely to lengthenthe growing season and result in a northwardshift of where some crops are planted,assumingthe land and infrastructure are available for suchgeographic shifts.In addition,there is somechance that frost-sensitive plants once grown pri-marily in areas such as the Imperial Valley ofCalifornia will be grown in the state’s CentralValley.

Conversely, it is possible that crops that prefercold winters such as winter wheat and potatoescould be limited to more northern areas(although other wheat varieties could be grown).It is very likely to be more difficult to relocateperennial crops such as vineyards,fruits and nuts,than to relocate annual crops,because perennialscan take many years to decades to get established.In addition, warmer temperatures can inhibitgrowth of certain fruit and nut crops that requirechilling during the winter. It is also possible thatwarmer temperatures will increase heat stress,weeds,pests,and pathogens that affect plants,ani-mals,and farm workers.

Changes in the seasonality of precipitation couldcause some problems. There is some chance thatvineyards, for example,could experience losses ifrains increase near harvest time — unseasonable234


Table 3: Estimated Changes in Crop Production in the West

Estimated Percent Changes in Dryland Crop Production for the Mountain Region from the Canadian andHadley Models (%)

2030s 2090s Crop Canadian Hadley Canadian Hadley Cotton 4.86 16.73 50.41 38.55 Hard Red Spring Wheat 12.92 16.90 -10.54 27.47 Hay 9.57 11.14 16.77 30.50 Tomatoes (processed) 21.92 23.59 -22.99 35.19 Oranges (processed) 66.90 69.90 114.60 111.60 Pasture 20.90 19.50 51.49 49.27

Estimated Percent Changes in Irrigated Crop Production for the Mountain Region from the Canadian andHadley Models (%)

2030s 2090s Crop Canadian Hadley Canadian Hadley Cotton 74.22 92.11 188.24 170.36 Hard Red Spring Wheat -16.98 -1.22 -29.62 -1.41 Hay 17.29 30.58 16.32 33.00 Tomatoes (processed) 21.92 23.59 -22.99 35.19 Oranges (processed) 66.90 69.90 114.60 111.60

Estimated Percent Changes in Dryland Crop Production for the Pacific Region from the Canadian andHadley Models (%)

2030s 2090s Crop Canadian Hadley Canadian Hadley Cotton 6.58 22.63 68.22 52.17 Hard Red Spring Wheat 16.25 65.75 137.90 131.10 Rice 6.49 6.27 1.76 5.77 Hay 26.76 28.38 62.24 50.29 Tomatoes (processed) -19.95 -9.14 -7.62 -19.54 Oranges (processed) 36.87 42.77 77.90 73.03 Pasture 47.53 58.83 102.12 92.55

Estimated Percent Changes in Irrigated Crop Production for the Pacific Region from the Canadian and Hadley Models (%)

2030s 2090s Crop Canadian Hadley Canadian Hadley Cotton 41.66 51.70 105.66 95.62 Hard Red Spring Wheat 0.25 4.60 4.80 11.75 Rice 6.49 6.27 1.76 5.77 Hay 38.26 61.06 52.94 70.33 Tomatoes (processed) -19.95 -9.14 -7.62 -19.54 Oranges (processed) 36.87 42.77 77.90 73.03

235


tions that breed in the muck. However, should con-ditions become generally wetter, it is likely vegeta-tion will get more dense,which may reduce wintermud.

Ranching is extremely vulnerable to drought(Liverman,1998) and should the climate becomedrier, vegetation productivity, water supplies,andthe carrying capacity of land and,hence,livestockproduction, would be reduced. In addition,highertemperatures can increase livestock diseases andcalving problems (Wagner and Baron 1998). Theeconomic impact would be felt most strongly in therural and intermountain areas.

Adaptation Options Stakeholders identified improvement in weatherforecasting to be the most important adaptation forranching. The timing of cattle sales and breeding,and the range of management strategies that ranch-ers employ, depend on knowledge of anticipatedand observed range conditions and long-term wateravailability. Consideration may be given to raisingdifferent species or breeds more suitable for hotterconditions (Wagner and Barron,1998).Management practices should be adjusted tochanges in conditions to reduce stress on ecosys-tems when appropriate.

5. Coastal Resources

Although a large portion of California’s coast ismade up of cliffs,many of the state’s most populouscoastal areas are vulnerable to sea-level rise,includ-ing the San Francisco Bay area and the coast southof Santa Barbara. If no protective measures aretaken,sea-level rise will inundate hundreds ofsquare miles of low-lying land in California (Gleick,1988). Unless protected,coastal structures fromharbors to houses could succumb to the ocean,asnumerous California beachfront homes did inFebruary 1998. Also,beaches will be flooded unlessdefensive actions are taken. Agricultural lands inthe Sacramento-San Joaquin delta,some already asmuch as 25 feet below sea level,are threatenedwith inundation. As the ocean encroaches,someaquifers near the coast will become contaminatedby saltwater intrusion. Rising sea level could inun-date many coastal wetlands and unprotected devel-opment (see Figure 12). Should sea walls be usedto protect coastal areas downslope, wetlands arelikely to be blocked from migrating inland with thesea and could thus be lost.

A study of the costs of protecting the margins ofSan Francisco Bay from a 3.3-foot (1-meter) sea-level

and can be enhanced in response to climatechange. Breeding crops better suited to takeadvantage of higher CO2 levels and more heat mayalso make sense.

Development of drought- and heat-resistant cropswill help reduce the vulnerability of the agricul-ture sector. Bioengineering could be helpful inthis regard,but this is a complicated issue withadvantages and disadvantages.

There is substantial potential to reduce currentand future water use through less water demand-ing technologies and better water managementpractices. Agriculture could switch from highwater use crops such as irrigated pasture,alfalfa,cotton,and rice,to less water demanding cropssuch as soybeans,wheat,barley, corn for grain,andsorghum (USDA,1997;Gleick et al.,1995). Water-intensive crops grown in desert areas could possi-bly become uneconomic if water prices increase.More efficient irrigation technologies such assprinklers or drip irrigation can reduce waterdemand. Crops may need to be planted earlier totake better advantage of earlier runoff (higher tem-peratures may also favor earlier planting of crops).

4. Ranching

Ranching is quite sensitive to climate variability.The cattle industry in Arizona reduced herd size byabout 80,000 head during the 1994 to 1996drought,but an increase in precipitation in NewMexico in the same period resulted in an increaseof 100,000 head of cattle (McClaren and Patterson,1998)

It is possible that an increase in temperature andprecipitation could have the benefit of increasingforage production in many locations,and lengthen-ing the growing and grazing season on nativerangelands. Moreover, increased water suppliesand longer growing seasons would make it possi-ble to harvest more alfalfa crops per year (nowtypically two to three),increase hay supplies,andreduce prices.

A warmer and wetter climate can pose problemsfor dairy cattle. There is some chance that flood-ing could wash out holding ponds. If wintersbecome wetter, it is possible dairy cattle will suf-fer. In the Chino,California area,which produces25% of the state’s milk,some 6,500 head of cattledied during El Niño conditions in February 1998.Cows and calves became mired in mud and weak-ened by the cold,succumbing to bacterial infec-236


8Gleick and Maurer (1990) also noted that many costs were not,or

could not be,quantified.

rise concluded that more than $1 billion (1990$)would be needed for new or upgraded levees toprotect existing industrial and commercial develop-ments,with an additional annual maintenance costexceeding $100 million (Gleick and Maurer, 1990).8

Adaptations Options Strategies for protecting developed coastal areasinclude defending with engineered fortifications anyassets of high economic value such as cities,air-ports,ports,and delta levees (for water supplysecurity); relocating vital assets to higherground (or engineering alternative solutions);and, for less economically valued areas of thedeveloped coast (housing on coastal bluffs), retreat-ing. Building coastal defenses can block inlandmigration of wetlands and result in loss of beaches.Advance planning can prevent new developmentsfrom being built in areas likely to be at risk in thefuture. Avoiding new construction is likely to provefar less costly than trying to protect such develop-ment in the future. For new development of anykind,local government agencies such as the CoastalCommission could be authorized to consider “risk ofharm”from impacts of climate change. After consid-eration of risk of harm,developments would beapproved only with no assured warranty of safety orloss,and private insurance would underwrite therisk or self-insurance would bear any costs or losses.

6. Tourism and Recreation

The spectacular scenery, favorable climate,and largeamounts of public land,especially in national parks,have made the West a major destination for touristsfrom around the world. Billions of dollars havebeen invested in ski resorts in all of the region’sstates,with Colorado,Utah,and California havingparticularly extensive facilities which attract manyvisitors. Tourist expenditures in the West are grow-ing. Hotels,lodging,amusement,and recreation pro-vided $32 billion in revenues in 1996 and are pro-jected to provide $52 billion in 2045 (US BEA,1999a).

Since the tourism industry in the West is so stronglyoutdoors oriented,it is particularly sensitive to cli-mate. The period for winter activities is likely toshrink,while the period for summer activities islikely to increase. Natural vegetation provides partof the aesthetic attraction,and significant climate-change effects on western ecosystems are very like-ly to change the distribution and abundance of vege-tation and animals. Much of the attraction fortourists is associated with water:its inherent aes-thetic appeal,and the growing water-related sports

of fishing,whitewater rafting,kayaking,and canoe-ing. Some of this recreation is on the many artificiallakes such as Lake Powell and Lake Mead. Increasesin runoff could possibly enhance these sports whiledecreases could possibly reduce their attractiveness.

The skiing industry is at particular risk from highertemperatures. With rising temperatures,snowpackseasons are very likely to shorten. Moreover, snow-

237


Figure 12: This figure shows the spatial extent and distribution ofcurrent and projected wetland habitat types in southern SanFrancisco Bay (derived from US Fish and Wildlife, NationalWetlands Inventory data) following sea-level rise as calculatedusing the Sea Level Affecting Marshes Model (SLAMM4) (Galbraithet al., In prep.). The sea-level rise scenarios use historic rates thatinclude local subsidence (obtained from tide gages at or close toeach of the sites), superimposed on the median estimate of the like-ly rate of sea-level change due to climate change (Titus andNarayanan, 1996). The historic rate of sea-level rise in the southernpart of San Francisco Bay is estimated to be 3.0 feet (0.9 meter) by2050 and 5.3 feet (1.6 meter) by 2100. This could be due to tectonicmovements resulting in land subsidence and/or crustal subsidencedue to the depletion of subterranean aquifers. When combined withthe projected median estimate of 13.4 inches (34 cm) eustatic (glob-al) sea-level rise by 2100 from climate change, sea-level rise is esti-mated to be 3.3 feet (1.0 meter) by 2050 and 6.1 feet (1.9 meters) by2100. The numbers shown in parenthesis on the figure indicatethat approximately 57.7% of tidal flat habitat will be lost by 2050and 62.1% by 2100, compared to the current condition. Using onlythe historic rate of local sea level rise, approximately 58.9% (2050)and 61.1% (2100) of tidal flat habit. See Color Plate Appendix.

Current and Projected Wetlands in South SanFrancisco Bay

Air Quality Air quality is a significant problem in many parts ofthe West. For example,with 17 million inhabitantsoccupying a basin subject to many temperatureinversions,the greater Los Angeles area has a partic-ularly serious problem with ground-level ozone lev-els and particulate matter. In addition,SanFrancisco,Las Vegas,Phoenix,Reno,and Salt LakeCity have problems meeting federal governmentstandards for ozone levels,and many western citieshave particulate matter concentrations close to orexceeding federal standards.

If precursors are not reduced and temperaturesincrease,it is possible that ozone levels,which are attheir peak in the summer, will increase. Higher tem-peratures increase ozone formation when precur -sors are available. Should wetter conditionsincrease biomass,which emits ozone precursors,airquality could further decline. Fine particulate mat-ter concentrations could also increase. This couldlead to more health problems. On the other hand,increased El Niño conditions,which would result inmore storms and precipitation in the winter, wouldbe likely to reduce levels of winter air pollutants,such as carbon monoxide and particulates.Reducing emissions of air pollutants,which is need-ed anyway in many Western cities,may be evenmore necessary because of climate change.

Health Effects Since the West is generally dry, it is likely to be atlower risk of increase in vector-borne infectious dis-eases than more humid regions. Should the Westbecome warmer and significantly wetter, there issome chance that there could be an increase in thepotential presence of disease vectors. In recentyears, wetter conditions contributed to the outbreakof cases of Hantavirus in the region,particularly inthe Four Corners area (Engelthaler et al.,1999). It ispossible that wetter conditions would increase thepotential for a Hantavirus outbreak and other cli-mate sensitive diseases such as plague (Parmenter etal.,1999),assuming other control measures are nottaken. But,because of the capability of the publichealth system,it is unlikely that there will be largeoutbreaks of infectious diseases in the West. It ismore likely that if climate gets warmer and wetter,the potential for small outbreaks from people carry-ing the diseases from other countries into theregion would increase.To keep health risks low, it iscritical that the public health system be maintained.

The region currently has lower risk of heat stressmortality than Midwest and Northeastern cities.Kalkstein and Greene,1997 found in San Francisco

line elevations will rise. Lower-elevation and moresouthern ski areas are likely to be at greatest risk.

On the other hand, rising temperatures are likely toresult in a longer summer season for warm weatherrecreation activities. Backpacking,biking,mountainclimbing,and rock climbing have been growing inpopularity. For example,the number of backpack-ers in the Canyonlands of Utah rose sevenfold fromthe early 1970s to the mid-1990s (Riebsame,1997).But there is some chance that increased precipita-tion could decrease the number of days desirablefor summer recreation activities. Whether warmerand wetter conditions would result in a net increaseor decrease in summer recreation is unclear.

Adaptations Options Adaptations for tourism and recreation generallyinvolve diversification of income sources. The larg-er, better-capitalized resorts such as Aspen and Vailhave already adapted their facilities to serve as sum-mer destination resorts with a range of warm-seasonrecreational activities,conference facilities,andmusic and dance programs;those with private landhave extensive,high-priced real estate development.The smaller areas may not be sufficiently capitalizedor have the private land to achieve these forms ofdiversification. This strategy can be taken inresponse to climate change and can be done inanticipation of climate change only to the extentthat current recreation patterns support it.

ADDITIONAL ISSUESMining The mining industry is quite sensitive to climatevariability and change because of the importance ofwater to its production processes,and the fact thatenvironmental laws hold mines liable for the qualityof effluent water. Water is needed for the concen-tration step of processing. In addition,a typical min-ing operation is required to collect and use orprocess all precipitation that falls within the limitsof the facility or otherwise comes in contact withunnaturally exposed material. There is some chancethat increased precipitation can result in morerunoff of pollutants,while decreased precipitationcould result in reductions in water supplies for pro-cessing. The mining industry is likely to adapt to cli-mate variability by relying on short-term forecasts ofprecipitation in day-to-day operations,interannualforecasts of precipitation for temporary enhance -ment of water treatment facilities,and long-term cli-mate outlooks to decide on capital improvements inwater holding areas,mechanical pumps,and watertreatment facilities.238


and Los Angeles,winter mortality would decrease,while in Los Angeles summer mortality wouldincrease.The estimated net change in mortalityacross the nine large western cities studied is closeto zero.

ADAPTATION STRATEGIESFor managed systems in the West,there appears tobe significant potential to reduce negative conse-quences of climate change and take advantage ofpositive impacts. For example,wise water manage-ment can reduce the risks from droughts andfloods. The potential for adaptation appears to behigh in many of the other potentially affected sec-tors of the economy. And many of the measuresmentioned above would have significant benefitsregardless of climate change. Clearly though,theseadaptations will involve costs,are not necessarilyeasy to implement,and can result in both winnersand losers. The costs and feasibility of these adapta-tions were not assessed. Should there be sudden orextreme climate changes,it is not clear how effec-tive adaptations would be in ameliorating adverseimpacts.

Risks from climate change are likely to be greatestfor those affected sectors or subsectors that lack theresources or capacity to adapt. For example,it isuncertain how effective the adaptations discussedabove would be in reducing the vulnerability of nat-ural ecosystems and biodiversity to climate change.Reducing current stresses on natural systems mayhelp,but adverse impacts are still likely to occur.Poor or immobile people are likely to bear particu-lar risks from climate change. In addition,activitiesthat are fixed in place,such as national parks andIndian reservations,are at particular risk becausethey are unable to relocate in response to climatechange. The development of adaptation strategiesmay need to pay particular attention to these typesof situations.

Many development trends can increase vulnerabilityto climate change. But the development of the Westalso presents many opportunities to prepare for andthereby reduce the risks of climate variability andchange in development plans and projects. Forexample,development can attempt to minimizewater use and degradation of water and air quality.Coastal structures can be designed to minimize therisks of sea-level rise and harm to natural ecosys-tems. Development in flood plains can be reduced.The tourist industry can further diversify into bothwinter and summer recreation. The public health

system can be maintained and improved. Riparianareas can be protected,fragmentation of ecosystemsreduced,and migration corridors developed or main-tained. The capability of the poor and immobile toadapt can be enhanced. The effectiveness of thesestrategies in reducing the risks of climate change hasnot been assessed.

One strategy that should help virtually all affectedsectors is improved forecasting of climate. In partic-ular, improved seasonal and annual forecasting of cli-mate would help water supply managers, farmers,ranchers,miners,health care professionals,and oth-ers plan for wet or dry seasons and extreme heatand cold episodes. Improved multidecadal forecastsof climate change would help infrastructure design-ers,land use planners,and others in identifyingfuture directions of climate change.

CRUCIAL UNKNOWNS ANDRESEARCH NEEDS Clearly there are many uncertainties about how cli-mate in the West will change and what the impactsof such changes will be,and there are many researchneeds that should be addressed to help resolveuncertainties. Improved research is a coping strate-gy itself, and many of the research areas will helpimprove the effectiveness of adaptations identifiedabove. A number of general research needs cutacross all sectors sensitive to climate change:

• Improve climate forecasts for the West:improvepredictions of the sign,magnitude,and seasonali-ty of change of important climate variables suchas precipitation,and improve the estimation ofprobabilities.

• Seek a better understanding of the interrelation-ships between climate impacts and the institu-tional structures that facilitate or constrain effec-tive action.

• Improve methods for involving the public inresearch and communicating research results tothe public and decision makers.

• Conduct more research on adaptation,specificallyto improve understanding of the potential effec-tiveness,costs,and impediments to adaptations.

Water• Develop a better understanding of the human

and ecological impacts of climate variability andchange on water resources,particularly at thelocal and regional levels.

• Analyze all water resource options,including full239


• Examine the impact of climate change on thecompetitiveness of ranching with other regionsin the US and globally.

Coastal Issues• Develop a statewide (California) map identifying

the extent of sea-level rise. Certain areas havebeen mapped using a simple 1-meter demarca-tion,but the maps have not been based on thebest available mapping technology, such as thatused by NOAA and NASA.

• Analyze the impacts of sea-level rise and acceler-ated cliff erosion on buildings,energy, transporta-tion,coastal infrastructure,and other features.The impacts of altered sediment flows along thecoast may also have important implications forharbors and navigation.

Ecosystem Management• Conduct extensive interdisciplinary ecosystem

research,monitoring,and modeling in the regionto provide an understanding of ecosystem struc-ture and function on which sound land-manage-ment practices can be based.

• Improve understanding of CO2 fertilization onnatural ecosystems.

• Improve understanding of the effectiveness ofpossible adaptations for preserving biodiversity.

Fire• Improve modeling and predictive capacity to

allow fire personnel to deploy resources as need-ed.

• Link the remote sensing and GIS-based imagesbeing used with models to better understand firerisk and the dynamics of fire to increase ground-truthing. Additional work on the dynamics of fireand ecological communities would improve themodeling efforts.

Health• Improve understanding of the vulnerability of

the region to the spread of infectious diseasesand heat waves.

• Improve understanding of the relationshipsbetween emissions of air pollutants, climatechange,and resulting air pollution.

Landscape Processes• Conduct more research on how climatic change

will affect the land surface,in terms of erosionby wind and water, sediment discharge,and land-slide potential.

efficiency potential in all sectors, water transferoptions,impacts of pricing changes on all sectors(including the impacts of dif ferent water pricelevels on the types of crops grown in differentlocations).

• Develop methodologies,analytical tools,anddesign criteria for incorporating increased climat-ic variability and change into hydraulic designand water resources planning and management.

• Develop effective long-term strategies for conser-vation.

• Study improvements in flood forecasting andresponse,improvements in reservoir manage-ment,and enhancement of other infrastructurethat may be vulnerable to climate impacts.

• Improve understanding of groundwaterresources in terms of amounts,locations, waterquality, relationship to surface water, and poten-tial for recharge,including effects of climate vari-ability and altered precipitation. Develop anaccurate and complete inventory of groundwa-ter, ascertain the rates of use and potential fornatural recharge, examine the extent to which itcan be recharged by technology, and understandhow all of these parameters would be affectedby an increase or decrease in precipitation.

• Examine how to ef fectively transfer knowledgeand technology from the research community tothe public,particularly with regard to improvinglong-term planning and developing more realisticsupply/demand water budgets.

AgricultureMany of the research topics that apply to waterresources are critical for agriculture. Additionalresearch topics include the following:

• Improve understanding of the effect of climatechange on plant yield and health.

• Enhance knowledge of how climate change andvariability may affect pest and disease problems.

• Improve understanding of the effects of ENSOon agriculture.

• Examine the impact of climate change on thecompetitiveness of agriculture with otherregions in the US and globally.

• Analyze the institutional obstacles to adaptationto climate change in agriculture (water laws,endangered species,etc.)

Ranching• Examine how ranchers cope with climate vari-

ability and how their experience can be used toenhance their ability to adapt to climate change.Examine the interactions between urban devel-opment, climate change,and loss of land forranching.240


LITERATURE CITEDAber, J.,R.Neilson,S.McNulty, J. Lenihan, D. Bachelet,and R.Drapek, Forest processes and global environ-mental change: Predicting the effects of individual andmultiple stressors,BioScience,in press,2001.

Bales,R.C.,and D. M.Liverman,Climate patterns andtrends in the Southwest,in Climate Variability andChange in the Southwest: Final Report of theSouthwest Regional Climate Change Symposium andWorkshop, edited by R.Merideth, D. Liverman,R.Bales,and M. Patterson,The University of Arizona, Tucson,Arizona,1998.

Brown,T. C.,Projecting US freshwater withdrawals,Journal of Water Resources Research, 36, 769-780.

California Department of Finance,California’s popula-tion tops 33 million,press release,May 1998.

California Trade and Commerce Agency, California: Aneconomic profile,September 1997.(http://commerce.ca.gov/california/economy).

Chapin, F. S.,et al.,Consequences of change biodiversi-ty, Nature 405, 234-242,2000.

CLIMAS,Climate variability in the Southwest region,(http://geo.ispe.arizona.edu/swclimate/water%con-text.htm.),1998.

Dale,V. H.,et al., Forest disturbances and climatechange, BioScience, in press,2000.

Dettinger, M. D.,and D. R.Cayan,Large-scale atmospher-ic forcing of recent trends toward early snowmeltrunoff in California, Journal of Climate, 8, 606-623,1995.

Diaz,H. F.,and C.A.Anderson,Precipitation trends andwater consumption related to population in the south-western United States: A reassessment, WaterResources Research, 31,713-720,1995.

Engelthaler, D. M.,et al.,Climatic and environmentalpatterns associated with Hantavirus pulmonary syn-drome, Four Corners region,United States, EmergingInfectious Diseases 5(1),87-94,1999.

Galbraith,H., D. Park,R. Jones, J. Clough,B.Harrington,S.Herrod-Julius,and G. Page, Potential impacts of sealevel rise due to global climate change on migratoryshorebird populations at coastal sites in North America,Report to US Environmental Protection Agency, StratusConsulting Inc.,Boulder, Colorado. Draft.April 19,2000.

Gleick, P.,Regional hydrologic consequences ofincreases in atmospheric CO2 and other trace gases,Climate Change, 10, 137-161,1987.

Gleick, P.,Climate change and California: Past,present,and future vulnerabilities, Societal Responses toRegional Climate Change: Forecasting By Analogy,edited by M.H.Glantz,Westview Press,Boulder,Colorado,1988.

Gleick, P. H.,Vulnerability of water systems,in ClimateChange and US Water Resources, edited by P. E.Waggoner, John Wiley & Sons,New York,pp.223-240,1990.

Gleick, P. H.,and E. P. Maurer,Assessing the costs ofadapting to sea level rise: A case study of SanFrancisco Bay, The Pacific Institute for Studies inDevelopment,Environment,and Security and theStockholm Environment Institute,Stockholm,Sweden,1990.

Gleick, P. H.,and L.Nash,The societal and environmen-tal costs of the continuing California drought, PacificInstitute for Studies in Development,Environment,andSecurity, Berkeley, California,66 pp.,1991.

Gleick, P. H., P. Loh,S.V. Gomez,and J. Morrison,“California Water 2020: A Sustainable Vision,” PacificInstitute for Studies in Development,Environment,andSecurity, Oakland,California,113 pp.,1995.

Gleick, P. H.,and E.L.Chalecki,The impacts of climatechanges for water resources of the Colorado andSacramento-San Joaquin River basins, Journal of theAmerican Water Resources Association, 35(6),1429-1441,1999.

Grimm, N. B.,A.Chacon,C. N. Dahm,S.W. Hostetler, O.T. Lind, P. L.Starkweather, and W.W.Wurtsbaugh,Sensitivity of aquatic ecosystems to climatic andanthropogenic changes: The basin and range,AmericanSouthwest and Mexico, Hydrologic Processes, 11,1023-1041,1997.

Hansen,A. J.,R. P. Neilson,V. Dale,C.Flather, L.Iverson,D. J. Currie,S.Shafer, R.Cook,and P. J. Bartlein,Globalchange in forests: Responses of species,communities,and biomes, BioScience, in press,2000.

Hay, L.E.,R.L.Wilby, and G. H.Leavesley,A comparisonof delta change and downscaled GCM scenarios forthree mountainous basins in the United States,Journal of the American Water Resources Association,36(2),2000.

241


Krieger, D. J.,Saving open spaces: Public support forfarmland protection,Center for Agriculture in theEnvironment,(http://farm.fic.niu.edu/cae/wp/99-1/wp99-1.html) December 29,1999.

Lettenmaier, D. P.,and D. P. Sheer, Climatic sensitivity ofCalifornia water resources, Journal of Water ResourcesPlanning and Management, 117, 108-125,1991.

Liverman, D.,Southwest overview, in Climate Variabilityand Change in the Southwest: Final Report of theSouthwest Regional Climate Change Symposium andWorkshop, edited by R.Merideth, D. Liverman,R.Bales,and M. Patterson,University of Arizona,Tucson,Arizona,1998.

McCabe, G. J.,and D. M.Wolock,General-Circulation-Model simulations of future snowpack in the westernUnited States, Journal of the American Water ResourcesAssociation, 35(6),1473-1484,1999.

McClaren,M.,and M. Patterson,Ranching,in ClimateVariability and Change in the Southwest: Final Reportof the Southwest Regional Climate ChangeSymposium and Workshop, edited by R.Merideth, D.Liverman,R.Bales,and M. Patterson,University ofArizona,Tucson,Arizona,1998.

Meyer, J. L.,M. J. Sale, P. J. Mulholland,and N. L. Poff,Impacts of climate change on aquatic ecosystem func-tioning and health, Journal of the American WaterResources Association, 35(6),1373-1386,1999.

Miller, K.A.,S.L.Rhodes,and L.V. MacDonnell,Water allo-cation in a changing climate: Institutions and adapta-tion, Climatic Change, 35,157-177,1997.

Miller, N. L., J. Kim,R.K.Hartman,and J. Farrara,Downscaled climate and streamflow study of the south-western United States, Journal of the American WaterResources Association, 35(6),1525-1537,1999a.

Miller, N. L., J. Kim,and M. D. Dettinger, California streamflow evaluation based on a dynamically downscaledeight year hindcast,observations,and physically basedhydrologic models, Eos, Transactions, AGU, 80,F406,1999b.

Miller, N. L., J. Kim,and M. D. Dettinger, Climate changesensitivity study for two California river basins, U.S.Department of Energy Workshop on the Climate ChangePrediction Program,Sponsored by the U.S.Departmentof Energy, Bethesda,Maryland,March 27-29,2000.

Houghton, J.T.,L. G. Meira Filho,B.A.Callander, N.Harris,A.Kattenberg,and K.Maskell,(Eds.), ClimateChange 1995: The Science of Climate Change,Contribution of Working Group I to the SecondAssessment Report of the Intergovernmental Panel onClimate Change, Cambridge University Press,Cambridge,England,1996.

Hurd,B.H., N. Leary, R. Jones,and J. B.Smith,Relativeregional vulnerability of water resources to climatechange, Journal of the American Water ResourcesAssociation, 35(6),1399-1410,1999a.

Hurd,B.H., J. M.Callaway, J. B.Smith,and P. Kirshen,Economic effects of climate change on US waterresources,in The Economic Impacts of ClimateChange on the US Economy, edited by R.Mendelsohnand J. E.Neumann,Cambridge University Press,Cambridge,England,133-177,1999b.

Jeton,A.E.,M. D. Dettinger, and J. L.Smith, Potentialeffects of climate change on streamflow, eastern andwestern slopes of the Sierra Nevada,California,andNevada, Water-Resources Investigations Report 95-4260, US Geological Survey, Sacramento,California,1996.

Kalkstein,L.S.,and J. S.Greene,An evaluation of cli-mate/mortality relationships in large US cities and thepossible impacts of a climate change, EnvironmentalHealth Perspectives, 105(1),2-11,1997.

Karl,T.R.,C.N.Williams,Jr., F.T. Quinlan,and T.A.Boden,1990:United States Historical Climatology Network(HCN) Serial Temperature and Precipitation Data,Environmental Science Division,Publication No.3404,Carbon Dioxide Information and Analysis Center, OakRidge National Laboratory, Oak Ridge,TN, 389 pp.

Karl,T. R.,and R.W. Knight,Secular trends of precipita-tion amount,frequency, and intensity in the UnitedStates, Bulletin of the American MeteorologicalSociety, 79, 231-241,1998.

Kattlemann,R.,and M.Embury, Riparian areas and wet-lands, Sierra Nevada Ecosystem Project: Final Reportto Congress, Volume III,Assessments, CommissionedReports, and Background Information, Centers forWater and Wildland Resources,University of California,Davis,1996.

Knox, J. B.,Global climate change: Impacts onCalifornia,an introduction and overview, in GlobalClimate Change and California, edited by J. B.Knox,University of California Press,Berkeley, California,1991.

242


Murphy, D. D.,and S.B.Weiss,Effects of climate changeon biological diversity in western North America:Species losses and mechanisms,in Global Warmingand Biological Diversity, edited by R.L. Peters and T. E.Lovejoy,Yale University Press,New Haven,Connecticut,pp.355-368,1992.

Myers, N.,R.A.Mittermeier, C. G. Mittermeier, G.A.B.daFonseca,and J. Kent,Biodiversity hotspots for conserva-tion priorities, Nature, 403, 853-858,2000.

NCAR (National Center for Atmospheric Research),VEMAP tables of means and variances, Also availableon Web site(http://www.cgd.ucar.edu/naco/vemap/vemtab.html),October 11,1999a.

NCAR (National Center for Atmospheric Research),Soil Moisture [Changes Projected by GCMs],Also avail-able on Web site(http://www.cgd.ucar.edu/naco/gcm/sm.html),August31,1999b.

NCAR (National Center for Atmospheric Research),Foundation document figures,Also available on Website (www.cgd.ucar.edu/naco/found/figs.html),November 5,1999c.

Neilson,R. P.,and R. J. Drapek, Potentially complex bios-phere responses to transient global warming, GlobalChange Biology, 4, 505-521,1998.

Nemani,R.R.,M.A.White, D. R.Cayan, G.V. Jones,S.W.Running,and J. C.Coughlan,Asymmetric climate warm-ing improves California vintages, Climate Research, inpress,2001.

Neumann, J. E., G.Yohe,R.Nicholls,and M.Manion,Sea-level rise and global climate change: A review ofimpacts to the US coasts, Pew Center on GlobalClimate Change,Arlington, Virginia,2000.

NPA Data Services,Inc., Demographic Databases;Three Growth Projections 1967-2025, NPA DataServices,Inc.,Washington,DC,1999.

NPS,National Park Service,(http://www.nps.gov/grca/),March 17,1999.

NYT, New rules sought on tapping the Colorado River,The New York Times, May 23,1999.

Parmenter, R.R.,E. P.Yadav, C.A. Parmenter, P. Ettestad,and K.L.Gage, Incidence of plague associate withincreased winter-spring precipitation in New Mexico,American Journal of Tropical Medical Hygiene, 61,814-821,1999.

Parmesan,C.,Climate and species’ range, Nature, 382,765-766,1996.

Riebsame,W. E.(Ed.), Atlas of the New West, W.W.Norton & Company, New York,1997.

Sagarin,R. D., J. P. Barry, S.E.Gilman,and C.H.Baxter,Climate related changes in an intertidal communityover short and long time scales, EcologicalMonographs, 69, 465-490.

Solley,W. B.,R.R.Pierce,and H.A. Perlman,Estimateduse of water in the United States in 1995,US GeologicalSurvey Circular 1200,US Government Printing Office,Denver, Colorado,1998.

Tausch,R. J.,C.L.Nowak,and R.S.Nowak,Climatechange and plant species responses over theQuaternary: Implications for ecosystems management,in Interior West Global Change Workshop, edited by R.W.Tinus,General Technical Report RM-GTR-262, USDAForest Service, Fort Collins,Colorado,1995.

Templin,W. E.,California — Continually the Nation’sleader in water use,(http://ca.water.usgs.gov/wuse/awra/),accessed May13,1999.

The Nature Conservancy,(http://www.consci.tnc.org/library/pubs/rptcard/map.html),accessed March 17,1999.

Titus, J. G.,and V. Narayanan,The risk of sea level rise,Climatic Change, 33,151-212,1996.

Torn,M.,E.Mills,and J. Fried,Will climate change sparkmore wildfire damages?,Lawrence Berkeley Laboratory,Berkeley, California,(http://eande.lbl.gov/CBS/EMills/wild.html),1998.

(US BEA) US Bureau of Economic Analysis,Gross StateProduct by component and industry 1977-1997.” 1999a( h t t p : / / w w w. b e a . d o c . gov / b e a / re gi o n a l / g s p / g s p l i s t . h t m l ) .

(US BEA),US Bureau of Economic Analysis,RegionalEconomic Information System (http://fisher.lib.vir-ginia.edu/reis/),1999b.

US Bureau of Reclamation, Updating the Hoover Damdocuments, Denver, Colorado,1978.

US Bureau of Reclamation,Compilation of records inaccordance with Article V of the Decree of the SupremeCourt of the United States in Arizona versus Californiadated March 9,1964, for calendar years 1989-1997.

243


Willman,M.L.,State may consider mudslide insurancefor homeowners, Los Angeles Times, February 14,1998.

Wilkinson,R.,and T. Rounds, Potential impacts of cli-mate change and variability for California: Californiaregional workshop report,National Center forEcological Analysis and Synthesis,University ofCalifornia at Santa Barbara,1998a.

Wilkinson,R.,and T. Rounds,Climate change and vari-ability in California: White paper for the CaliforniaRegional Assessment,National Center for EcologicalAnalysis and Synthesis,University of California at SantaBarbara,1998b.

Wolock, D. M.,and G. J. McCabe,Simulated effects of cli-mate change on mean annual runoff in the contermi-nous United States, Journal of the American WaterResources Association, 35(6),1341-1350,1999.

Wong,A.K.,L.Owens-Viani,A.Steding, P. H.Gleick, D.Haasz,R.Wilkinson,M. Fidell,and S.Gomez,Sustainableuse of water: California success stories, PacificInstitute for Studies in Development,Environment,andSecurity, Oakland,California,1999.

Yoshiyama,R.M.,E.R.Gerstung, F.W. Fisher, and P. B.Moyle, Historical and present distribution of Chinooksalmon in the Central Valley drainage of California,Sierra Nevada Ecosystem Project: Final Report toCongress, Volume III,Assessments, CommissionedReports, and Background Information, Centers forWater and Wildland Resources,University of California,Davis,1996.

US Bureau of Reclamation,Estimate of 1999 ColoradoRiver Use,Boulder City, Nevada, US Bureau ofReclamation Memorandum to All Interested Persons,May 26,1999.

US Census Bureau, Statistical Abstract of the UnitedStates: 1998 (118th edition),US Department ofCommerce,Washington,DC,1998.

US Department of Agriculture (USDA),Agriculturalresources and environmental indicators,1996-1997,Agricultural Handbook No. 712, USDA,EconomicResearch Service, Washington,DC,1997.

US Environmental Protection Agency, Climate changeand New Mexico,EPA 236-F-98-007p,US EPA,Washington,DC,1998a.

US Environmental Protection Agency, Climate changeand Utah,EPA 236-F-98-007z,US EPA,Washington,DC,1998b.

US Fish and Wildlife Service,Listed species by state andterritory,(http://www.fws.gov/r9endspp/listmap.html),AccessedMarch 17,1999.

US Geological Survey (USGS),Seasonal land coverregions:scale:1:7,500,000,” map,US Geological Survey,Reston,Virginia,1993.

US Geological Survey (USGS),(http://goechange.er.usgs.gov/sw/changes/anthro-pogenic/cropland/lu4592_lyt.gif),1999.

US Historical Climatology Network (USHCN),1999.(http://www.ncdc.noaa.gov/ol/climate/research/ushcn/ushcn.html)

Wagner, F. H.,and J. Barron,Rocky Mountain/GreatBasin regional climate-change workshop,Utah StateUniversity, Logan,Utah,1998.

Walker, B.,and W. Steffen,An overview of the implica-tions of global change for natural and managed terres-trial ecosystems, Conservation Ecology Vol 1;Issue 2.(online)(http://www.consecol.org/Journal/vol1/iss2/art2/ ),1997.

Western Water Policy Review Advisory Commission,Water in the West: Challenge for the Next Century,National Technical Information Service,Springfield,Virginia,1998.

244


ACKNOWLEDGMENTSMany of the materials for this chapter are based oncontributions from participants on and those work-ing with the:

California Workshop and Assessment SteeringCommitteesRobert Wilkinson,University of California-Santa

Barbara,co-chairJeff Dozier, University of California-Santa Barbara,co-

chairRichard Berk,University of California,Los AngelesDan Cayan,Scripps Institution of Oceanography,

University of California,San DiegoKeith Clarke,University of California,Santa BarbaraFrank Davis,University of California,Santa BarbaraJames Dehlsen,Dehlsen AssociatesPeter Gleick, Pacific Institute for Studies in

Development,Environment,and SecurityMichael Goodchild,University of California,Santa

BarbaraNicholas Graham,Scripps Institution of

Oceanography / University of California,SanDiego

William J. Keese,California Energy CommissionCharles Kolstad,University of California,Santa

BarbaraMichael MacCracken,USGCRP and Lawrence

Livermore National LaboratoryJim McWilliams,University of California,Los AngelesJohn Melack,University of California,Santa BarbaraNorman Miller, Lawrence Berkeley National

Laboratory / University of California,BerkeleyHarold A.Mooney, Stanford UniversityPeter Moyle,University of California,DavisWalter Oechel,San Diego State UniversityLarry Papay, Bechtel GroupClaude Poncelet, Pacific Gas and Electric CompanyThomas Suchanek,NIGEC / University of California,

DavisHenry Vaux,University of California Office of the

PresidentJames R.Young,Southern California Edison

Rocky Mountain/Great Basin Workshop andAssessment TeamsFrederic Wagner, Utah State University, co-chairThomas Stohlgren,US Geological Survey, co-chairConnely Baldwin,Utah State UniversityJill Baron,US Geological Survey, Fort Collins,COHope Bragg,Utah State UniversityBarbara Curti,Nevada Farm Bureau,Reno,NV

Martha Hahn, U.S.Bureau of Land Management,Boise,ID

Sherm Janke,Sierra Club,Bozeman,MTUpmanu Lall,Utah State UniversityLinda Mearns,National Center for Atmospheric

Research,Boulder, COHardy Redd,Private Rancher, Lasal,UTGray Reynolds,Sinclair Corporation,Salt Lake City,

UTDavid Roberts,Utah State UniversityLisa Schell,Colorado State UniversitySusan Selby, Las Vegas Valley Water DistrictCarol Simmons,Colorado State UniversityDale Toweill,Idaho Dept.of Fish and Game,Boise,IDBooth Wallentine,Utah Farm Bureau Federation,Salt

Lake City, UTTodd Wilkinson, Journalist,Bozeman,MT

Southwest/Colorado River Basin Workshop andAssessment TeamsWilliam A.Sprigg,University of Arizona,co-chairTodd Hinkley, US Geological Survey, co-chairDiane Austin,University of ArizonaRoger C.Bales,University of ArizonaDavid Brookshire,University of New MexicoStephen P. Brown, Federal Reserve Bank of DallasJanie Chermak,University of New MexicoAndrew Comrie,University of ArizonaPrabhu Dayal,Tucson Electric Power CompanyHallie Eakin,University of ArizonaDavid C.Goodrich,US Department of AgricultureHoward P. Hanson,Los Alamos National LaboratoryLaura Huenneke,New Mexico State UniversityWilliam Karsell,WAPAKorine Kolivras,University of ArizonaDiana Liverman,University of ArizonaRachel A.Loehman,Sandia National LaboratoriesJan Matusak,Metropolitan Water District of Southern

CaliforniaLinda Mearns,National Center for Atmospheric

ResearchRobert Merideth,University of ArizonaKathleen Miller, National Center for Atmospheric

ResearchDavid R.Minke,ASARCOBarbara Morehouse,University of ArizonaDan Muhs,US Geological SurveyWilson Orr, Prescott CollegeThomas Pagano,University of ArizonaMark Patterson,University of ArizonaKelly T. Redmond,Desert Research InstitutePaul R.Sheppard,University of ArizonaVerna Teller, Isleta PuebloJames R.Young,Southern California Edison

245


247

CHAPTER 9

POTENTIAL CONSEQUENCES OF CLIMATEVARIABILITY AND CHANGE FOR THEPACIFIC NORTHWESTEdward A. Parson1,2, with contributions from members of the Pacific NorthwestAssessment Team: Philip W. Mote3,Alan Hamlet4, Nathan Mantua5,Amy Snover6,William Keeton7, Ed Miles8, Douglas Canning9, Kristyn Gray Ideker10


Chapter Summary



Ecological Context


Key Issues

Freshwater

Salmon

Forests

Coasts

Additional Issue

Agriculture


Literature Cited

Acknowledgments

1John F. Kennedy School of Government,Harvard University, 2Coordinating author for the National AssessmentSynthesis Team, 3University of Washington (UW),Chair, Pacific Northwest Regional Assessment Team, 4Dept of Civiland Environmental Engineering,UW, 5Climate Impacts Group,UW, 6Climate Impacts Group,UW, 7College of ForestResources,UW, 8Climate Impacts Group,UW, 9Washington State Dept of Ecology, Olympia, 10Ross and Associates(work completed while at UW)


• Over the 20th century, annual-average tempera-ture in the Northwest rose 1 to 3°F (0.6 to1.7°C) over most of the region,with nearly equalwarming in summer and winter.

• Annual precipitation increased nearly every-where in the region, by 11% on average,with thelargest relative increases about 50% in northeast-ern Washington and southwestern Montana.

• Year-to-year variations in the region’s climateshow a clear correlation with two large-scale pat-terns of climate variation over the Pacific,the ElNiño/Southern Oscillation (ENSO),and PacificDecadal Oscillation (PDO). The region-wide pat-tern associated with these phenomena is thatwarm years tend to be relatively dry with lowstreamflow and light snowpack,while cool onestend to be relatively wet with high streamflowand heavy snowpack. This has clear ef fects onimportant regional resources: warmer drier yearstend to have summer water shortages,less abun-dant salmon,and increased risk of forest fires.


• Regional warming is projected to continue at anincreased rate in the 21st century, in both sum-mer and winter. Average warming over theregion is projected to reach about 3°F (1.7°C) bythe 2020s and 5°F (2.8°C) by the 2050s.

• Annual precipitation changes projected through2050 over the region range from a small decrease(-7% or 2”) to a slightly larger increase (+13% or4”).

• Projected precipitation increases are concentrat-ed in winter, with decreases or smaller increasesin summer. Because of this seasonal pattern,even the projections that show increases inannual precipitation show decreases in wateravailability.

CHAPTER SUMMARY

Regional Context

The Northwest,which includes the states ofWashington,Oregon,and Idaho,has a great diversityof resources and ecosystems,including spectacularforests containing some of the world’s largest trees;abrupt topography that generates sharp changes inclimate and ecosystems over short distances;moun-tain and marine environments in close proximity,making for strong reciprocal influences betweenterrestrial and aquatic environments;and nearly allthe volcanoes and glaciers in the contiguous US.The region has seen several decades of rapid popu-lation and economic growth,with population nearlydoubling since 1970,a growth rate almost twice thenational average. The same environmental attrac-tions that draw people and investment to the regionare increasingly stressed by the region’s rapid devel-opment. The consequences include loss of old-growth forests, wetlands,and native grass andsteppe communities,increasing urban air pollution,extreme reduction of salmon runs,and increasingnumbers of threatened and endangered species.Climate change and its impacts will interact withthese existing stresses in the region.

248


Key Findings

• Projected warmer wetter winters are highly like-ly to increase flooding risk in rainfed rivers,while projected year-round warming and driersummers are highly likely to increase risk of sum-mer shortages in both rainfed and snowfedrivers,because of smaller snowpack and earliermelt.

• Salmon are likely to be harmed by increased win-ter flooding, reduced summer and fall flows,andrising stream and estuary temperatures. It is alsopossible that earlier snowmelt and peak stream-flow will deliver juveniles to the ocean beforethere is enough food for them. Climate change isconsequently likely to hamper efforts to restoredepleted stocks,and to stress presently healthystocks.

• The coniferous forests that dominate much ofthe Northwest landscape are sensitive to summermoisture stress. Their extent,species mix,andproductivity are likely to change under projected21st century climate change,but the specifics ofthese changes are not yet known.

• Sea-level rise will likely require substantial invest-ment to avoid coastal inundation,especially inlow-lying communities of southern Puget Soundwhere the coast is subsiding. Projected heavierwinter rainfall is likely to increase soil saturation,landsliding,and winter flooding.

• El Niño events increase erosion both by raisingsea level for several months and by changing thedirection of winds and waves from westerly tosouthwesterly. Climate change is projected tobring similar changes and associated impacts,including severe storm surge and coastal erosion.

249

Chapter 9 / The Pacific Northwest

250


PHYSICAL SETTING ANDUNIQUE ATTRIBUTESThe Pacific Northwest region includes the states ofWashington,Oregon,and Idaho,and for assessingimpacts in the Columbia River basin,some areas inadjoining states and the Canadian province ofBritish Columbia. The region has a great diversity ofresources and ecosystems,including spectacularforests containing some of the world’s largest trees;abrupt topography that generates sharp changes inclimate and ecosystems over short distances;moun-tain and marine environments in close proximity,making for strong reciprocal influences betweenterrestrial and aquatic environments;and nearly allthe volcanoes and glaciers in the contiguous US.

The region is divided climatically, ecologically, eco-nomically, and culturally by the Cascade Mountains.The low-lying areas west of the Cascades hold threequarters of the region’s population,concentrated inthe metropolitan areas of Tacoma-Seattle-Everettalong the Puget Sound coast,and Portland in theWillamette Valley. Here,once-dominant forestry, fish-ing and agriculture have been overtaken by aero-

space,computer software and hardware,trade andservices,although the relatively declining resourcesectors remain economically and culturally impor-tant. The Northwest still provides about a quarter ofthe nation’s softwood lumber and plywood (Hayneset al.,1995,tables 18 and 27). Agriculture is nowmuch more important in the east of the region.Thanks in part to massive water-management andirrigation projects,the fertile lowlands of easternWashington are the “Fruit Bowl”of the nation,pro-ducing 60% of the nation’s apples and large frac-tions of its other tree fruit,while Idaho producesabout a quarter of the nation’s potatoes (USDA,2000).

SOCIOECONOMIC CONTEXTThe region has seen several decades of rapid popu-lation and economic growth. Population has nearlydoubled since 1970,a growth rate almost twice thenational average. Growth has been strongly concen-trated in the major western metropolitan areas andin the smaller but fast-growing inland cities of Boiseand Spokane (Jackson and Kimmerling,1993).Federal lands comprise roughly half the region’sland area.1 The region’s environment presents agreat variety of outdoor recreational opportunities,and its moderate climate and quality of life con-tribute to its continuing attraction to so many new-comers. The region is projected to continue grow-ing faster than the national average,its populationincreasing from the present 10.5 million to 19 mil-lion by 2050 (with a range of 14.5 million to 23 mil-lion) (Terleckyj 1999a,1999b;US Census Bureau,2000). Both recent and projected growth rates aresimilar east and west of the Cascades (very slightlyhigher on the west),so the west side is projected tocontinue to contain nearly three quarters of theregion’s population.

The same environmental attractions that draw peo-ple and investment to the region are increasinglystressed by the region’s rapid development. Thepredominant current stresses arise from direct

POTENTIAL CONSEQUENCES OF CLIMATEVARIABILITY AND CHANGE FOR THE PACIFICNORTHWEST

Figure 1: The Puget Sound area has experienced rapid growthover the past two decades. Source: P. Mote, University ofWashington

1 Federal lands are 30% of Washington,48% of Oregon,and 64% ofIdaho,or 48% of the region overall,with an additional 5% state-owned.(Jackson and Kimmerling 1993,p.32).

251


Idaho has been converted to agriculture (Mac et al.,1998,p.649;Noss et al.,1995). Grazing has trans-formed nearly all of the remaining grassland andsagebrush steppe,leading to large-scale replacementof native perennials with invasive annuals such asCheatgrass,Medusahead and Yellow Starthistle,andto expansion of Juniper woodlands into formerrangeland (Miller and Rose,1995;Miller and Wigand,1994;West and Hassan,1985).

human interventions in the landscape,through suchactivities as dam building, forestry (includingreplacement of natural forests by plantations),andland-use conversion from the original forests, wet-lands, grasslands,and sagebrush to expansion ofmetropolitan areas,intensively managed forests, agri-culture,and grazing. The consequences include lossof old-growth forests, wetlands,and native grass andsteppe communities,increasing urban air pollution,extreme reduction of salmon runs,and increasingnumbers of threatened and endangered species.

ECOLOGICAL CONTEXTThe Northwest has a great diversity of landscapesand ecosystems, reflecting the region’s varied cli -mate and topography. Dense,tall moist coniferousforests cover about 80% of western Washington andOregon,with Douglas fir, western red cedar andwestern hemlock at most low-elevation locations,western hemlock and Pacific silver fir at middle ele-vations,and mountain hemlock at high elevations(Franklin and Dyrness,1973). A century of commer-cial logging has cut nearly all this forest at sometime, greatly altering its species and age distribution.Only 10 to 20% of the original extent remains asold-growth forest (Marcot et al.,1991; Kellogg,1992). The west also includes oak forests and grass-lands in low-lying river valleys,coastal salt marshesand freshwater wetlands,and in the KlamathMountains of southern Oregon,a mixed forest ofdrought-resistant conifers and hardwoods (Mac etal.,1998,p.646). East of the Cascade crest,the drierclimate and more frequent fires generate an open,park-like forest of ponderosa pine and Douglas fir,with sub-alpine fir, Engelmann spruce and patchesof alpine larch at higher elevations and whitebarkpine especially prominent near the upper tree line.In the Rocky Mountains and the east slope of theCascades, forest gives way at high elevations toalpine meadows,and at lower elevations to juniperwoodlands,sagebrush steppe,and grasslands,as wellas high desert and lava fields in Idaho.

As on the west side of the Cascades,human influ-ence has greatly altered east-side ecosystems. Firesuppression, grazing,and selective cutting havetransformed all but a few percent of the originalponderosa pine forest into overstocked mixed-species forests that are highly susceptible to fire,insects,and disease (Henjum et al.,1994). Morethan 99% of the prairie grasslands near the meetingpoint of Idaho,Oregon,and Washington have beenconverted to crops,mostly wheat,while about 90%of the sagebrush steppe on the Snake River plain in

Figure 2: Old-Growth Douglas Fir Forest in the Cascades.Source: T.B. Thomas, US Forest Service

Figure 3: The Columbia is one of the most intensively developedriver systems in the world. Source: ©P. Grabhorn

endangered or threatened under the EndangeredSpecies Act (ESA),including the Puget SoundChinook,the first ESA listing to affect a major metro-politan area (NMFS,2000). Of roughly 450 speciesof birds identified in five sub-regions of theNorthwest by the Breeding Bird Survey, from 10 to35 species per sub-region show decreased numberssince the 1960s,while 3 to 25 show increases,withthe largest net decreases in the coastal forests ofsouthern Oregon (Carter and Barker, 1993). Amongmammals,seven carnivores — the Grizzly Bear, GrayWolf, Lynx,Wolverine, Fisher, Marten,and Kit Fox —have small and threatened regional populations,principally due to disturbance,loss of forest habitat,and the secondary effects of logging road construc-tion (Weaver et al.,1996).

CLIMATE VARIABILITY ANDCHANGEWest of the Cascades the climate of the Northwestis maritime,with abundant winter rains,dry sum-mers,and mild temperatures year-round — usuallyabove freezing in winter, so snow seldom stays onthe ground more than a few days. Most places westof the Cascades receive more than 30 inches (75cm) of precipitation annually, while some westwardmountain slopes of the Olympics and Cascadesreceive more than 200 inches (500 cm). Although amild maritime climate has prevailed in the regionfor several centuries,thousand-year records showsubstantial fluctuations. For example,from about4,000 to 8,000 years ago in the Puget Sound area,dominance of dry vegetation types such asCalifornia chaparral suggests the region’s climatewas much warmer and drier, resembling the presentclimate of California’s northern Central Valley(Detling,1953,1968).

East of the Cascade crest,the climate shifts sharplyfrom abundant rainfall to abundant sunshine,withannual precipitation generally less than 20 inches(50 cm),as little as 7 inches (20 cm) in some places.These precipitation differences are most pro-nounced in winter:summer precipitation in thewest is only slightly higher than in the east,whilewinter precipitation is four to five times higher.Figures 5 and 6 illustrate the large differences inannual and seasonal precipitation across the region.Even the inland mountain ranges receive much lessprecipitation than the western Cascades orOlympics. Though average temperatures are similareast and west,the east has larger daily and annualranges,with hotter summers and colder winters.

An additional stress on inland forests is the devasta-tion of whitebark pine (Pinus albicaulis), a domi-nant species near the upper tree-line in the Rockiesand Cascades, by the introduced fungus,white pineblister rust. Throughout the species’ range in north-ern Idaho more than half of the trees are dead,while infection rates of living trees are above 50%throughout Washington and Idaho,and above 20%in Oregon (Keane and Arno,1993; Kendall,1995).

The region is high in biodiversity across major taxa.In the wet west-side forests,more than 150 speciesof terrestrial snails and slugs have been identified,and 527 species of fungi,of which 234 are rare andoccur nowhere else (FEMAT, 1993). It is estimatedthat these forests may support 50,000 to 70,000species of arthropods,although only preliminarysurveys of arthropods have been conducted. A sur-vey of one experimental forest in the OregonCascades found more than 3,400 (Parsons et al.,1991). Oregon contains between 3,000 and 4,000identified species of vascular plants,Idaho andWashington between 2,400 and 3,000,puttingOregon in the top six states for plant diversity andWashington and Idaho in the top 15. Of these plantspecies, about 8-12 are rare in Oregon,5-8 inWashington and Idaho.2 The 33 species of amphib-ians in the region include 17 that are endemic,butonly one candidate for federal listing,the Oregonspotted frog (Bury, 1994). Although 67 populationsof fish in the region are on either federal or statesensitive species lists, much more is known aboutsalmon and trout,the most highly valued species inthe region,than about other species. Of 58 distinctsalmonid stocks in the region,26 are now listed as

252


Figure 4: Major ecological regions of the Pacific NorthwestSource: United States National Atlas – See Color Figure Appendix

2"Rare" means an international endangerment rank of G1 to G3 (Morseet al.,1995).

Observed Climate Trends

Over the 20th century, the Northwest has grownwarmer and wetter. Annual-average temperaturerose 1 to 3°F (0.6 to 1.7°C) over most of the region,with nearly equal warming in summer and winter.Annual precipitation also increased nearly every-where in the region, by 11% on average,with thelargest relative increases about 50% in northeasternWashington and southwestern Montana.3

In addition to this trend toward a warmer, wetter cli-mate,the Northwest’s climate also shows significantrecurrent patterns of multi-year variability. Theseyear-to-year variations tend to be consistent over theentire region,and are evident in both winter andsummer. The predominant pattern is that warmyears tend to be relatively dry with low streamflowand light snowpack,while cool ones tend to be rela-tively wet with high streamflow and heavy snow-pack. Although the differences in temperature andprecipitation are relatively small (differences inmonthly-average temperature of up to 2 to 4ºF or1.1 to 2.2ºC in winter),they have clearly discernibleeffects on important regional resources. Warmerdrier years tend to have summer water shortages,less abundant salmon,and increased risk of forestfires (dell’Arciprete et al.,1996;Mantua et al.,1997;Hulme et al.,1999).

These year-to-year variations in the region’s climateshow a clear correlation with two large-scale pat-terns of climate variation over the Pacific,one moreand one less well known. The El Niño/ SouthernOscillation (ENSO) is an irregular oscillation with aperiod of 2 to 7 years,which is widely known andintensively studied. ENSO’s positive El Niño phasewarms sea-surface temperature in the equatorialPacific and cools it in the central North Pacific,deepening the winter low-pressure system off theAleutians and bringing substantial changes in mid-latitude atmospheric circulation (Trenberth,1997).A more recently identified pattern of longer-termvariability is the Pacific Decadal Oscillation (PDO),defined in terms of changes in Pacific sea-surfacetemperature north of 20 degrees latitude. Like thewarm El Niño phase of ENSO, the warm or positivephase of PDO warms the Pacific near the equatorand cools it at northern mid-latitudes. But unlikeENSO,PDO’s effects are stronger in the central andnorthern Pacific than near the equator, and its irreg-ular period is several decades,tending to stay in one

phase or the other for 20 to 30 years at a time. PDOis also much less well understood than ENSO, inpart because its period is so long relative to the his-tory of reliable records that only two completeoscillations have been observed. The PDO was in itscool,or negative phase from the first sea-surfacetemperature records in 1900 (and possibly before)until 1925,then in warm or positive phase until1945,cool phase again until 1977,and warm phaseuntil the 1990s (Miller et al.,1994;1998). Evidenceis beginning to mount that another change to the 253


Figure 5: The Cascade mountains divide the wetter west from thedrier east. Source: Mapping by C. Daly, graphic by G. Taylor and J.Aiken, copyright © 2000, Oregon State University. – See ColorFigure Appendix

Average Annual Precipitation, Pacific Northwest ,1961-1990

Average Monthly Precipitation in the Pacific Northwest

Figure 6: West of the Cascades is wetter than east, but nearly allthe difference occurs in winter. Source: Mote et al. (1999), Figure 3,pg. 5.

3Analyses of Historical Climate Network data by NCDC. A similaranalysis of historical trends by UW JISAO (in Mote et al.,1999) usingslightly different re-weighting algorithms and regional boundariesfound a 14% average precipitation increase over the 20th century. Thedifference between these is not significant.

254


cool phase of PDO likely occurred in the mid-1990s,but it is too early to tell with confidence. The warmphase of PDO, like El Niño,strengthens the Aleutianlow, bringing warmer winter temperatures overwestern North America and warmer ocean tempera-tures along the coast. In these winters,the mid-lati-

tude storm track tends to split,with one branch car-rying storms south to California,the other north toAlaska. These winters consequently tend to be drierthan normal in the Pacific Northwest and wetterthan normal along the coasts to both the south andnorth. In contrast, years during the cool phase ofPDO and during the cool,La Niña phase of ENSOare associated with a weaker Aleutian low,whichtends to bring winters that are cooler and wetterthan normal in the Pacific Northwest. The majorexception to this cool-wet versus warm-dry patternoccurs during the strongest El Niño events,such asthat of 1998. During these events,the Aleutian lowis very strong and is also shifted to the Southeast,making winters on the Northwest coast warmer andwetter – i.e.,while moderate El Niños tend to makeNorthwest winters warmer and drier, the strongestEl Niños reverse the effect on precipitation andmake the region warmer but with near normal pre-cipitation.

Scenarios of Future Climate

Projections of climate change in the Northwestwere conducted through 2100 using the Canadianand Hadley models (Boer et al.,1984,1999a,b;McFarlane et al.,1992;Flato et al.,1999;Mitchell etal.,1995;Mitchell and Johns,1997; Johns et al.,1997),and through 2050 with five additional generalcirculation models (GCMs),two of 1998 vintage and

Figure 7: The red line shows annual-average temperature in theNorthwest in the 20 th century, observed from 113 weather sta -tions with long records. The blue line shows the historicalNorthwest average temperature calculated by the Canadianmodel from 1900 to 2000, and projected forward to 2100.Source: Mote et al (1999), Summary (p. 6). – See Color FigureAppendix

Northwest Average Temperature, Observed and Modeled

Temperature Change - 20th & 21st Centuries

Precipitation Change - 20th & 21st Centuries

Observed 20th Canadian Model 21stBoth climate models projectcontinued precipitationincreases, with the largestincreases in the southernpart of the region.

Precipitationhas increasedover most ofthe PacificNorthwestsince 1900.

100%

75%

50%

25%

0

-25%

-50%

-75%

-100%

100%80%60%40%20%0-20%-40%-60%-80%-100%

Hadley Model 21st

Observed 20th Canadian Model 21stBy 2100, both models proj-ect warming near 5ºF westof the Cascades, with muchlarger warming further eastin the Canadian model.

Warmingsince 1900 inthe PacificNorthwestranges from0 to 4ºF.

Hadley Model 21st+15ºF

+10ºF

+5ºF

0º

-5ºF

+15ºF

+10ºF

+5ºF

0º

-5ºF

Figure 8: Temperature change observed in the 20th and projected for the 21st centuries. – See Color Figure Appendix

Temperature Change 20th and 21st Centuries

255


that is too “maritime”—milder in both winter andsummer, with precipitation more evenly distributedacross the region. One study comparing theCanadian model to a finer-scale regional climatemodel for western Canada (where GCMs have thesame bias) suggested that these biases are moreacute for precipitation than for temperature (Lapriseet al.,1998). In projecting future climate,eachmodel’s bias relative to the present climate isremoved. Despite their overall maritime bias for theNorthwest,the Canadian and GFDL models(although not the Hadley model) do reproduce theNorthwest region’s observed 20th-century climatetrends fairly well,particularly for temperature.Figure 7 shows the annual-average PacificNorthwest temperature calculated by the Canadianmodel through the 20th and 21st centuries,with acomparison to the observed record for the 20th cen-tury. The model matches the observed trend of the20th century closely, and projects substantially morerapid warming through the 21st century. Figure 8shows the changes in temperatue and precipitationover the region projected by the Canadian andHadley models for the 21st century, and comparesthese to the observed pattern of changes over the20th century.

These projections all show regional warming con-tinuing at an increased rate in the next century, inboth summer and winter. Average warming over theregion is projected to reach about 3°F (1.7°C) bythe 2020s and 5°F (2.8°C) by the 2050s. Annual pre-cipitation changes projected through 2050 over theregion range from a small decrease (-7% or 2”) to aslightly larger increase (+13% or 4”) (Hamlet andLettenmaier, 1999),but these precipitation changes,

three of 1995 vintage. In addition,one analysislinked another GCM to a regional atmosphericmodel,to provide a climate projection with the finespatial resolution (1.5 kilometers, about 1 mile) nec-essary for hydrological studies (Leung and Ghan1999a,1999b). In addition to the Canadian andHadley models,the 1998 models included thosedeveloped by the Max-Planck Institut furMeteorologie (MPI) and the National Oceanic andAtmospheric Administration’s Geophysical FluidDynamics Laboratory (GFDL). These four modelswere run using the standard emission scenario,a 1%annual increase in equivalent atmospheric CO2 con-centration with trends in sulfate aerosol loadingfrom the IPCC IS92a scenario. The 1995 modelsincluded earlier versions from the Hadley Center,MPI,and GFDL. Compared to the 1998 group,thesemodels included simpler representations of sea,ice,and land surfaces. It is important to note that theseearlier model runs also used a different emissionscenario,which did not include aerosols. The finer-scale analysis used the Community Climate Model(CCM3) of the National Center for AtmosphericResearch (NCAR) driving the Regional ClimateModel of the Pacific Northwest National Laboratory(PNNL-RCM). This analysis used the same scenarioas the 1995 models,1% annual CO2 –equivalentincrease with no aerosols.

The coarse spatial resolution of GCMs is particularlytroublesome for replicating the spatial character ofclimate in the Northwest,which is strongly shapedby the region’s abrupt topography. Instead of thesharp Cascade crest,the models show a relativelysmooth rise from the Pacific to the Rockies.Consequently, they simulate a climate for the region

Table 1: Model Projections of Northwest Regional Climate

2020s 2050s

Model Precip change,inches Precip change,inches Temp change Apr-Sep Oct-Mar Temp change Apr-Sep Oct-Mar

Canadian 3.5ºF +0.2 +3.0 5.9ºF +0.3 +4.1 Hadley 3.2ºF +1.3 +3.5 4.8ºF +0.8 +2.5 MPI 3.7ºF -0.3 +0.6 5.3ºF -0.8 -0.4 GFDL 4.5ºF -0.2 +0.5 7.3ºF +0.4 +1.7

MPI 95 2.2ºF -2.5 +0.8 4.6ºF -1.8 +0.7 Hadley 95 2.8ºF -1.7 +2.5 5.4ºF -1.1 +2.5 GFDL 95 3.3ºF +0.4 +2.7 6.1ºF -0.8 +2.8 Average 3.1ºF -0.3 +1.6 5.3ºF -0.5 +1.4

Note:Results from 1995-vintage climate models were based on an emission scenario that did not include aerosols.Source:Table 3 (pg.21),Mote etal.,(1999b).

unlike projected temperature changes,lie within the20th century range of year-to-year variability.Projected precipitation increases are concentratedin winter, with decreases or smaller increases insummer. Because of this seasonal pattern, even theprojections that show increases in annual precipita-tion show decreases in water availability. The 1995and 1998 models show no systematic differences intheir projections. These results are shown in Table1,and are compared to 20th century climate variabil-ity in Figure 9. Most models by 2020,and all modelsby 2050,project a regional climate that is warmerthan even the warmest years of the 20th century.The simulation that linked a GCM to a regionalmodel showed similar results,with average annualwarming of about 4°F (2.2°C) by 2050, wetter win-ters and drier summers,but showed larger seasonaldifferences – about 6°F (3.3°C) warming in winterand 2°F (1.1°C) in summer – due to its finer-scalerepresentation of snow-albedo feedback (Leung andGhan,1999a,1999b).

After 2050,the projected trend to a warmer, wetterregional climate continues in both the Hadley and256


Canadian models,with substantially more warmingin winter than in summer in both models. By the2090s,projected average summer temperatures riseby 7.3°F (4.1°C) in the Canadian model and 8.3°F(4.6°C) in the Hadley model,while winter tempera-tures rise 8.5°F (4.7°C) in the Hadley model and10.6°F (5.9°C) in the Canadian model. Projectedprecipitation increases over the region range from afew percent to 20% (with a regional average of 10%)in the Hadley model,and from 0 to 50% (with aregional average of 30%) in the Canadian model.

These projected changes are associated with large-scale shifts in atmospheric circulation over thePacific,especially in winter, which resemble thechanges that occur during the strongest El Niñoevents. The Aleutian low is both strengthened andmoved to the southeast,displacing the mid-latitudestorm track southward and making the averagewinds on the Oregon and Washington coast strongerand more northward. Consequently, winters alongthe coast are warmer and wetter – both in total pre-cipitation,and in the amount of rainfall in heavystorms,because warmer temperatures increase thequantity of water vapor the atmosphere can hold.

KEY ISSUESOf the many potentially significant areas of climateimpact in the region,the Northwest regional studyselected four critical issues to examine:

Freshwater. Wetter winters are highly likely toincrease flooding risk in rainfed and mixedrain/snow rivers,while year-round warming anddrier summers are highly likely to increase risk ofsummer water shortages in rainfed,mixed,andsnowfed rivers,including the Columbia. In theColumbia system,allocation conflicts are alreadyacute,while a cumbersome network of overlappingauthorities limit the system’s adaptability, making itquite vulnerable to shortages.

Salmon. While non-climatic stresses on Northwestsalmon presently overwhelm climatic ones,salmonabundances have shown a clear correlation with20th century climate variations. Climate models can-not yet project the most important oceanic condi-tions for salmon,but the likely effects on their fresh-water habitat,such as warmer water and reducedsummer streamflow, are all highly likely to be unfa-vorable. Climate change is consequently likely toimpede efforts to restore already depleted stocks,and to stress presently healthy stocks.

Projected Northwest Climate Change, Compared to 20th Century Variability

Figure 9: Climate change by the 2020s and 2050s over theNorthwest Region from seven climate model scenarios. Any pointon the graph shows a particular combination of regional annual-average temperature and total annual precipitation. The asteriskand arrow through it show the average climate over the 20th centuryand its trend, warming about 1.5°F (0.8°C) with a 2.5" (6 cm) precip-itation increase. The oval illustrates how much the region’s cli-mate varied over the 20 th century, enclosing all combinations oftemperature and precipitation that were more than 5% likely tooccur. Each letter shows one model’s projection of the region’saverage climate, either in the 2020s or the 2050s. The models proj-ect that regional precipitation changes will lie within the range of20th century variability, but projected temperature changes lie out-side it. By the 2050s, all models project a climate so much warmerin the Northwest that it lies well outside the range of 20th centuryvariability ( *=1995-vintage model; H=Hadley; M=Max-Planck;G=GFDL; C=Canadian). Source: Regional report, Mote et al. (1999),fig. 12, pg. 19. – See Color Figure Appendix

Forests. Northwest forests have been profoundlyaltered by timber harvesting and land-use conver-sion,both east and west of the Cascades. WhetherNorthwest forests will expand or contract underprojected climate change,and what the effects onspecies mix and productivity will be,are highly like-ly to depend on assumptions related to plantresponse of water use efficiency to CO2 enrichment,on which present evidence is incomplete. Modelsproject several decades of forest expansion east ofthe Cascades and contraction to the west,withpreliminary indications of larger forest contractionin the longer-term,as increased moisture stressoverwhelms CO2-induced increases in water-useefficiency.

Coasts.. Sea level rise is likely to require substantialinvestments to avoid coastal inundation,and aban-donment of some property, especially in low-lyingcommunities of southern Puget Sound where theland is subsiding 0.3 to 0.8 inches per decade.Other likely effects include increased risk of winterlandslides on bluffs around Puget Sound,andincreased erosion on sandy stretches of the PacificCoast.

1. Freshwater

Freshwater is a crucial resource in the Northwest,and climatic effects on water resources stronglyinfluence and couple with many other domains ofimpact. Despite the region’s reputation as a wetplace,this only applies annually to the west slopes,and even they are dry in summer. Most of theregion receives less than 20 inches (50 cm) of pre-cipitation a year, and dry summers make freshwatera limiting resource for many ecosystems and humanactivities. Water supply, availability, and quality arealready stressed by multiple growing demands.

The Cascades largely divide rivers partly or entirelycontrolled by rainfall,whose flow peaks in winter,from those controlled by snowmelt,whose flowpeaks in late spring. The Columbia,a snowmelt-dominated river, is one of the nation’s largest,drain-ing roughly three-quarters of the region and carry-ing 55-65% of its total runoff. The Columbia is theregion’s primary source of energy and irrigationwater, and is managed by multiple agencies for mul-tiple,often conflicting values,including electricity,flood control, fish migration,habitat protection,water supply, irrigation,navigation,and recreation.Agriculture takes the largest share of present with-drawals,but other demands are growing,in particu-lar, the recent demand for in-stream flow require-

257


ments to protect salmon. With more than 250 reser-voirs and 100 hydroelectric projects,the Columbiasystem is among the most developed in the worldand has little room for further expansion, even asregional population growth and changing allocationpriorities are intensifying competition for water.Because its watershed is so large,the Columbia’sflow reflects an averaging of weather conditionsover large areas and seasonal time-scales.Consequently, climatic effects on its flow can bedetected and projected with more confidence thanfor smaller river systems,which respond moststrongly to shorter-term and more local events.

Columbia basin hydrology shows a strong signal ofboth ENSO and PDO. Because the warm phases ofthese oscillations tend to make winters both warmand dry, their effects on snowpack and streamflow,and hence on regional water supply, are strongerthan their effects on either temperature or precipita-tion. Warm-phase years accumulate less snowpack,and shift from snow accumulation to melting earlierin the season. In the Columbia,the warm phases ofENSO and PDO each reduce average annual flow byroughly 10% relative to the long-term mean,withlarger reduction of peak spring flow. The effects ofthe two oscillations are nearly additive,so years withboth in their warm phase have brought the lowestsnowpack and streamflow, and the highest incidenceof droughts. Five of the six extreme multi-yeardroughts since 1900 occurred during the warmphase of PDO (Mote et al.,1999b,p.32). Each oscil-lation’s cool phase has the opposite effect,bringingaverage stream flows about 10% higher than themean and the highest incidence of flooding. Four orfive of the five highest-flow years recorded occurredwhen PDO was in its cool phase,three of themwhen ENSO was also in its cool phase.4 When thetwo cycles are out of phase,streamflow tends to benear its long-term mean. A study of historical flood-ing in five smaller river basins found parallel results,a higher probability of flooding in both cool ENSOand cool PDO years,although the pattern was weak-er than for the Columbia and not uniform.Snowmelt-dominated rivers,whose floods reflectseason-long snow accumulation,showed it morethan rivers with strong rainfall components,whosefloods more typically reflect less predictable,individ-ual extreme precipitation events. Moreover, the pat-tern was present for the likelihood but not the inten-sity of flooding. Still,the pattern suggests a limitedability to predict seasonal flood risk,which mostmanagement agencies in the region are not presently

4The fifth of these years was 1997,the second-highest flow year. Thisyear’s status is ambiguous,because its extreme flow is one piece of evi-dence for the not yet resolved claim that PDO shifted back to coolphase in the mid 1990s.

258


exploiting (Mote et al.,1999b,p.31 and Table 4;Jones and Grant,1996).

Understanding the effects of projected warmer tem-peratures, wetter winters,and drier summers on theregion’s hydrology requires finer-scale analysis thanGCMs alone can provide. To obtain this,two analy-ses were conducted linking a GCM projection tofiner-scale models. In an analysis of snowpack forthe entire Northwest region using the PNNL-RCMmodel described above,projected snowpackdeclined about 30% over the Northern Rockies and50% over the Cascades by roughly 2050,the timethat CO2 concentration had doubled (Leung andGhan,1999a,1999b). Projected rise of the snowlineand earlier spring melt are corroborated by broadlysimilar results from earlier GCM-driven studies offine-scale hydrology models elsewhere in theAmerican West (Giorgi and Bates,1989;Giorgi et al.,1994;Matthews and Hovland,1997). In a separatestudy of the Columbia basin alone,several climatemodels were used to drive detailed models ofhydrology, and dam and reservoir operating rules,allowing an integrated examination of interactionsbetween natural and managed responses to climatechange. A striking result of these simulations was

widespread loss of moderate-elevation snowpack,asFigure 10 shows. Typical snowcover on March 1 inthe 2090s is projected to be about equal to presentsnowcover on in mid-May, although deep snowcoverin the upper basin still persists through June.

On snowmelt-dominated rivers like the Columbia,the very likely ef fect of these linked changes in tem-perature,precipitation,and snowpack will be toincrease winter flow and decrease summer flow.Both precipitation and temperature matter. Winterflow increases both because there is more winterprecipitation and because more of it falls as rain;summer flow decreases both because there is lesssnowpack and because it melts earlier in the spring.When changes in temperature and precipitation areconsidered separately, temperature — which climatemodels project with greater confidence than precip-itation — has the larger effect on streamflows (Moteet al.,1999b, Figure 28;Nijssen et al.,1997).

While all the climate models studied agree on theseseasonal effects on Columbia streamflow, they differin the relative size of winter and summer changes,and consequently in whether total annual flowincreases or decreases. Projections for annual flowin the 2020s in four models range from a 22%increase to a 6% decrease,with a mean 5% increase;for the 2050s,projected changes range from a 10%increase to a 19% decrease,with a mean 3%decrease. Figure 11 illustrates the range of project-ed seasonal flow shifts for all seven models in the2050s.

To assess the socioeconomic effects of thesechanges in stream flow, a reservoir operations modelwas used to project how the reliability of differentwater-management objectives (i.e.,the probability ofmeeting the objective in any year) changes underclimate change. The model was first used to exam-ine how present climate variability affects reliabilityof five uses,under two different sets of system oper-ation rules:present rules,under which the practiceis to grant highest priority to ensuring availability ofhydroelectric energy sold on “firm”contracts,5 and aset of alternative,“fish-first” rules that would givehighest priority to maintaining minimum flows toprotect fish. The effect of alternative operationrules for the system is to distribute risks of shortageamong uses. The results showed that reliability ishigh for all objectives in favorable,high-flow years(i.e.,cool PDO/La Niña years) and that one top-pri-ority objective can be maintained at or near 100%

Figure 10: By the 2090s, projected Columbia Basin snowpack onMarch 1 will be only slightly greater than present snowpack onJune 1. Simulations use the VIC hydrology model under the Hadleyscenario. Units in millimeters. Source: Hamlet and Lettenmaier,1999. – See Color Figure Appendix

Projected Reduction in Columbia Basin Snowpack

MAR APR MAY JUN

Base

5Despite recent policy changes to protect salmon, firm energy con-tracts still receive highest priority in practice and insufficient reservoircapacity is available to increase late-summer flows for fish.

2020s

2090s

reliability even in unfavorable,low-flow years (i.e.,warm PDO/El Niño years),but that other uses sufferlarge reliability losses in unfavorable years. For exam-ple,the reliability of both the fish-flow objectiveunder present rules,and the firm energy objectiveunder “fish-first” rules, fall to about 75% in warmPDO/El Niño years.(Mote et al.,1999b, figure 32) Ofthe five objectives considered,the most sensitivewere fish flows and recreational demand for full sum-mer reservoirs,which both drop below 85% reliabili-ty when annual flow is only 0.25 standard deviationsbelow its mean,a condition that could occur as oftenas four years out of ten.(Mote et al.,1999b,section2.4.2,pp.39-41.)

Though projected changes in total annual flows arerelatively small,the projected seasonal shifts are like-ly to bring large effects. Smaller, rainfed,and mixedrain/snow basins on the west slope are already sus-ceptible to winter flooding,and have significantlyincreased risk of flooding in the wetter winters of LaNiña years (Mote et al.,1999b,Table 4). The histori-cal severity of flooding,which is more influenced bysingle extreme events and less by season-long precip-itation,shows no such climate signal. Projectedwarmer, wetter winters are likely to bring furtherincreases in winter flooding risk in these basins,while continuing growth in coastal population willvery likely increase the property vulnerable to suchflooding. Impacts on human health are also possible,particularly where urban storm-sewer systems areinadequate to handle the increased runoff. No sys-tematic assessment of changes in westside winterflooding risk under climate change,and potentialconsequences for property damage and humanhealth,has yet been conducted.

In contrast,in large,snowmelt-dominated systemslike the Columbia, even large increases in winter

flows pose little increased risk of flooding,sinceexisting management systems are adequate torespond to floods and even the highest projectedwinter flows remain well below present springpeaks. But the system is much less robust to lowflows than high. Reduced summer flows are like-ly to bring substantial reductions in bothhydropower and freshwater availability by mid-century, exacerbating already-sharp allocationconflicts driven by population growth, expansionof irrigated farmland,and increasing priority formaintaining salmon habitat (Cohen et al.,1999).

Projections of changes in reliability for six objec-tives under present operational rules,using twoclimate models for the 2020s,and one for the2090s,are shown in Table 2 (Hamlet and

259


Projected Seasonal Shift in Columbia River Flow

Figure 11: While only small changes are projected in annualColumbia flow, seasonal flow shifts markedly toward larger win-ter and spring flows, and smaller summer and fall flows. Theblue band shows the range of projected monthly flows in the2050s under the Hadley and Canadian scenarios and the twoother 1998-vintage climate models used in the Northwestassessment (MPI and GFDL). Source: Mote et al. (1999),Summary, Figure 7. – See Color Figure Appendix

Table 2: Changes in Reliability of Various Columbia ManagementObjectives, Assuming Present Operating Rules.

Source:Mote et al.(1999b),Table 6.

2020s 2090s

Objective Base Case Hadley Max-Planck Hadley

Flood Control 98% 92% 96% 93% Firm Energy 100% 100% 98% 99% Non-firm Energy 94% 98% 87% 90% Snake River Irrigation 81% 88% 76% 75% Lake Roosevelt Recreation 90% 88% 79% 78% McNary Fish Flow 84% 85% 79% 75%

projected long-term climate trend would very likelyreduce reliability even more,but these interactionshave not yet been quantified.

Increasing stresses on the system are highly likely tocoincide with increased water demand,principallyfrom regional growth but also induced by climatechange itself. For example,an analysis of Portland’smunicipal water demand for the 2050s projected

Lettenmaier, 1999). Under present rules, reliabilityof firm energy is projected to remain near 100%,while other uses suf fer reliability losses up to 10%,similar to the ef fect of PDO. The effects of rulechanges,which will interact with both climatechange and variability, are likely to be even larger.For example,“fish-first” rules would reduce firmpower reliability by 10% even under present cli -mate,and by 17% in warm-PDO years. Adding the

Yakima Valley Irrigated Agriculture: Rigidity and Vulnerability

The Yakima Valley in south central Washington is one of the driest places in the United States,and containssome of its most fertile farmland. With annual precipitation of only about seven inches,the valley producesannual agricultural revenue of $2.5 billion,with the largest share from tree fruit. Fully 80% of the farmed areaof 578,000 acres is irrigated. The basin’s hydrology is strongly snowmelt-dominated with its main reservoirsin the mountains. With reservoir capacity equal to about half of annual demand,the region can tolerate amoderate one-year drought. Most farmers also have wells and pump groundwater in times of shortage,sub-ject to state permits.

Drought in the valley, agricultural expansion,and water policy have all followed the PDO cycle. Water short-ages have occurred eight times since 1945,all but once in warm-PDO years. Moreover, during the last coolphase (1945-1976) expectations of continued abundance of water led to sharp expansion of agriculture. Inthe subsequent warm phase farmers experienced substantial hardships,but nearly no contraction occurred.The first major decision allocating water among users was made in 1945,at the end of a warm phase period.Thereafter, no further controls were enacted until after 1979,when the warm phase had returned (Glantz,1982). Shortage years since then have seen many hundreds of new wells dug,under emergency state permitsthat typically allow pumping only for the duration of the drought.

Several economic and institutional factors systematically increase the valley’s vulnerability to drought.Division of water-right holders into “senior”users (who receive their full allocation every year) and “junior”users (who bear all risk of shortages) imposes large losses on junior users in shortage years – e.g.,a 63% pro-ration of water allocation and $140 million of losses in 1994 — and gives strong incentives for unsustainablepumping,depleting the region’s groundwater and perpetuating the myth that the region has enough water. Aprogressive shift from annual crops to more lucrative perennials,such as tree fruit, grapes,and hops,has fur-

ther exacerbated vulnerability:the perennials con-sume more water, reduce farmers’ flexibility to altertheir planting for projected dry years,and put manyyears’investment at risk from a single drought,fur-ther increasing incentives to pump groundwater.Some promising initiatives to improve managementof the region’s water are underway, including state-subsidized investments to increase efficiency of jun-ior users’irrigation systems,and a novel partnershipbetween one junior and one senior irrigation dis-trict,but the valley still lacks a coherent basin-widedrought strategy. Moreover, if PDO is now re-enter-ing its cool phase,making continuance of the amplesupplies of the past two years likely, the region islikely to face renewed pressure to expand irrigatedcropland. As happened during previous cool-PDOperiods,such expansion would further increase theregion’s vulnerability to future recurrence of dryconditions (Gray, 1999).

260


Figure 12: Irrigated agriculture in the Yakima Valley, where annualprecipitation is about seven inches. Source: P. Mote, University ofWashington

that climate change would impose an additional 5-8% increase in total summer demand (5% - 10% inpeak day demand) on top of a 50% increase in sum-mer demand from population growth (Mote et al.,1999b,section 2.4.3,p.43). Such climate-relateddemand increase,which is also highly likely forother demands such as electric generation,irriga-tion,and salmon habitat preservation, would com-pound the climate-related supply decreases dis-cussed above. Further assessment is required to

quantify risks of shortfall under the interaction ofregional growth, climate-driven demand increase,and climate-driven shifts in seasonal supply.

Adaptation Options While the Columbia’s infrastructure and institutionalauthority are able to manage high flows,at least upto some fairly high thresholds (Hamlet andLettenmaier, 1999;Miles et al.,2000),the means fordealing with low flows are rigid and inadequate.

261


Seattle Public Utilities: Learning From Water Shortages

Seattle Public Utilities (SPU) experienced summer droughts and potential shortages in 1987,1992,and 1998.Its responses to the three events illustrate institutional flexibility and learning. Summer 1987 began with fullreservoirs,but a hot dry summer and a late return of autumn rainscreated a serious shortage in which water quality declined,inade-quate flows were maintained for fish,and the main reservoir fell solow that an emergency pumping station had to be installed. Inresponse,the City developed a plan to manage anticipated shortageswith four progressive levels of response:advising the public ofpotential shortages and monitoring use; requesting voluntary usereductions;prohibiting inessential,high-consumption uses such aswatering lawns and washing cars;and rationing.

Another drought came in 1992, following a winter with low snow-pack but during which SPU had spilled water from its reservoirs tocomply with flood-control rules. With a small snowmelt, reservoirswere low by the spring,and SPU invoked mandatory restrictions dur-ing the hot dry summer that followed. The resultant low demandcaused water quality in the distribution system to decline. Lowreservoir levels also caused a decline in the quality of source water,prompting a decision to build a costly ozone-purification plant.

Their regrettable spilling of early 1992 alerted SPU to the risks inher-ent in following rigid reservoir rule curves. Since then,they have used a model that includes both ENSO andPDO to generate probabilistic projections of supply and demand six to twelve months in advance. Using thismodel during the strong El Niño of 1997-1998,they undertook conservation measures early in the year,including both weekly public announcements of supply conditions and allowing higher than normal winterreservoir fill. When 1998 brought a small snowmelt and a hot dry summer, these measures allowed thedrought to pass with the public experiencing no shortage.

In integrating seasonal forecasts into its operations,SPU is an uncommonly adaptable resource-managementagency. But it still has a long way to go in adapting to longer-term climate variability and change. SPUpresently projects that new conservation measures will keep demand at or below present levels until at least2010,while conservation measures and planned system expansion (including a connection with a neighbor-ing system) will maintain adequate supply until at least 2030 (A.Chinn,Seattle Public Utilities,personal com-munication,2000). Over this period, climate change is likely to have significant effects on both supply anddemand,but is not yet included in planning. The warmer drier summers projected under climate change arelikely to stress both supply and demand, requiring earlier capacity expansion and triggering the more restric-tive conservation provisions more often (Gray, 1999). Moreover, it is possible that the recently suggested shiftto cool PDO could mask this effect for a couple of decades, risking sudden appearance of shortages whenPDO next shifts back to its warm phase.

Figure 13: Since 1992, Seattle Public Utilities has usedinformation on year-to-year climate variability to help guideits seasonal forecasting and reservoir operations." Source: Seattle Public Utilities.

Moreover, long-term climate change is not yet usedin planning decisions, even for investment in infra-structure with expected lifetimes of many decades.

2. Salmon

Salmon are anadromous fish,meaning that theyswim upstream to spawn after spending most oftheir adult lives at sea. After hatching, young salmonremain in the stream for a few weeks to severalyears,depending on the species,then swim down-stream to the ocean. Most species make this trip inspring or early summer, taking advantage of the highstreamflows that accompany peak spring melting(in snowfed rivers) and ar riving at roughly the onsetof coastal and estuarine upwelling that fuels marinefood-chain productivity. In the ocean the salmongrow to adulthood and live for several months to sixyears before returning to their spawning grounds.Most die in their natal streams after spawning,there-by delivering marine-derived nutrients that are nowrecognized as important inputs to stream and ripari-an ecosystems.

Northwest salmon stocks have been highly stressedfor decades by intense fishing pressure and threatsto their stream habitats including urbanization,sedi-mentation and pollution of streams, wetland drain-ing,and dam building. Construction of the GrandCoulee and Hell’s Canyon dams eradicated allsalmon stocks above these points on the Columbiaand Snake Rivers respectively. Fish ladders onother dams are only partly ef fective at allowing fishto pass,and dams also degrade salmon habitat bychanging free-running rivers into chains of lakes,warming in-stream temperatures, reducing dis-solved oxygen,and altering sediment loads.

These factors have brought regional salmon stocksto widespread decline (Myers et al.,1998),despitemassive efforts to restore habitat and to supple-ment wild stocks with hatcheries,which reflectsalmon’s status not just as a commercially impor-tant fish but as a regional cultural icon. Over thepast century, Pacific salmon have disappeared fromabout 40% of their historical breeding range inWashington,Idaho,Oregon,and California,and manyremaining populations are severely depressed. Thedecline is not universal:populations from coastalrather than interior streams,from more northerlyranges,and with relatively short freshwater rearingperiods have fared better than others. In manycases,the populations that have not declined arenow composed largely or entirely of hatchery fish(National Research Council,1996;Beechie et al.,1994;Bottom,1995).

The Pacific Northwest Coordinating Agreement,charged with allocating water under scarcity on thebasis of defined priorities,is a weak and fragmentedbody with no one clearly in charge – Idaho is noteven a member. Moreover, the persistence of theprior appropriation doctrine under western waterlaw rigidly maintains large allocations to low-valueuses,hindering attempts to rationalize use underpresent and increasing scarcity.

Managing under projected future scarcity is highlylikely to require some combination of reducingdemand,increasing supply, and reforming institu-tions to increase flexibility and regional problem-solving capacity. Demand for water can be reducedthrough various technical means (e.g.,more effi-cient irrigation methods, changes in agricultural landmanagement,or high-efficiency plumbing fixtures innew construction) and through various policyapproaches (e.g.,tax incentives for conservationinvestments, revision of rate structures). The mostpromising approaches to encourage conservation,however, would be development of institutions toallow reallocation of water to higher-value uses,par-ticularly in times of shortage. Although the barriersto such a major shift of ownership rights are formi-dable,the Northwest could gain insights from priorwestern experience with water banks and contin-gent marketing (Huffaker,Whittlesey, andWandschneider, 1993;Miller, 1996;Miller, Rhodes,and MacDonnell,1997). Supply can likewise beincreased through various technical and policymeans,such as developing groundwater sources orusing groundwater recharge for water storage;improving system management by using seasonalclimate forecasts (Callahan et al.,1999);promotingoptimal use of existing lower-quality supplies,e.g.,by delivering reclaimed non-potable water for someuses;or developing non-hydro electric generatingcapacity;or if permitted by both governments,nego-tiating water purchase from Canada.

Increasing institutional flexibility is an essentialcomponent of response,but is rendered difficult bythe fragmentation of the current system. In a surveyof water managers on their use of climate forecastsin planning,most stated that they had limited flexi-bility to use even ENSO forecasts to take advantageof predictably higher-flow years,and most werecompletely unaware of PDO, whose effect on flowsappears to be as large as that of ENSO (Mote et al.,1999b,Section 2.5.3,pp.47-49.). Indeed,long-termassessment of institutional responses suggests thatin at least some cases, responses have served toincrease, rather than decrease,vulnerability to inter-decadal variability (see Yakima Valley box).

262


suggesting that the mechanism involves persistentwarm-water conditions over several years. The PDOsignal is much weaker for Puget Sound salmon thanfor stocks that exit directly from rivers into theopen ocean,suggesting that the gradual increase ofsalinity experienced by juveniles passing through anestuarine environment may increase their resilience(Pinnix,1998).

Climate models presently lack the detail to projectchanges in many specific factors in the marine envi-ronment that are most important for salmon,such asthe timing of seasonal coastal upwelling, variationsin coastal currents,and vertical stability of the watercolumn. But where climate models are moreinformative,in salmon’s inshore and estuarine habi-tats,their projections are uniformly unfavorable.Increased winter flooding, reduced summer and fallflows,and warmer stream and estuary temperaturesare all harmful for salmon (Baker et al.,1995;Bottom,1995. Earlier snowmelt and peak springstreamflow will likely deliver juveniles to the oceanbefore there is adequate food for them,unless theonset of summer northerly winds is also advancedunder climate change,but climate models cannotyet address this question. One study has suggestedthat oceanic warming from even a CO2 doublingmay push the range of some salmon north out ofthe Pacific entirely (Welch et al.,1998),althoughrecent studies suggest the notion of direct ocean

In March 1999,eight new salmon stocks were listedunder the Endangered Species Act (ESA) as threat-ened and one as endangered,bringing the numberof salmonid stocks listed to 26. The new listingsincluded Puget Sound Chinook,the first-ever ESAlisting of a species inhabiting a highly urbanizedarea. Just as prior listings of Columbia and coastalOregon stocks severely constrained forest and watermanagement,it is possible that the impacts of thisnew listing on the Seattle economy will be large,butit will be some years before these impacts areknown (Klahn,1999).

Salmon are sensitive to various climate-related condi-tions,both inshore and offshore,at various times oftheir life cycle. Eggs are vulnerable to stream scour-ing from floods,and migrating juveniles must makethe physiological transition from fresh to salt waterand require food immediately on reaching theocean. According to a long-standing but uncon-firmed hypothesis,their fate is keenly sensitive tothe timing of their arrival relative to the onset ofsummer northerly winds and the resultant upwellingof nutrient-rich deepwater and spring phytoplank-ton bloom:either too early or too late,and their sur-vival is imperiled from insufficient food (Pearcy,1992). More recently, it has been suggested thattheir survival is predominantly controlled by the bal-ance between predator and baitfish populations atthe time of their arrival,which determines predationpressure on salmon (Emmet and Brodeur, 2000).

Although the relative contributions of and interac-tions between climate and non-climate factors,andinshore and offshore conditions are highly uncer-tain,salmon stocks throughout the North Pacificshow a strong association with PDO (Figure 14)(Hare et al.,1999;Mantua et al.,1997; Francis,1997;Francis et al.,1998;Reeves et al.,1989).Salmon inthe Northwest are more abundant in the cool PDOphase,less abundant in the warm phase,whileAlaska salmon show the opposite pattern. When thePDO shifted from cold to warm in 1977,catches inthe Northwest dropped sharply while Alaskan catch-es soared.6 The mechanisms for this observed cli-mate effect on stocks are poorly known,and proba-bly include some ef fects of both freshwater andmarine changes. It is speculated that coastal watersoff Washington and Oregon during the warm phaseare warmer and more thermally stratified,and conse-quently poorer in nutrients. Although ENSO andPDO have similar effects on ocean and terrestrialenvironments in the Northwest,the signal of PDO insalmon stocks is much stronger than that of ENSO,

263


Figure 14: 20 th century catches of Northwest and Alaska salmonstocks show clear influence, in opposite directions, of the PacificDecadal Oscillation. Source: Mote et al. (1999), Figure 36, pp. 56.– See Color Figure Appendix

Salmon Catches and Inter-decadal Climate Variability

PDOIndex

6Catches are a good indicator of stocks,because catch variation sincethe 1930s is almost entirely due to stock fluctuations,not variation offishing effort (Beamish and Bouillon,1993)

Adaptation Options At present,the only climatic effects on salmon thatare sufficiently understood to allow an informedresponse involve warming of streams (McCullough,1999). Maintaining forest buffers for shade alongbanks and operational changes on managed riverscan reduce present warming (about 4.5º F, largelydue to dams) (Quinn and Adams,1996),and couldpotentially slow future warming under climatechange,although such measures would ultimately beoverwhelmed by continued climate warming.Though other mechanisms of climatic ef fect are lessknown,observed fluctuations of stocks with PDOsuggest applying a conservative bias to allowable-catch decisions during warm-PDO years. Measures toreduce general stress on stocks,such as changingdam operations to provide adequate late-summerstreamflows,might increase salmon’s resilience toother stresses,including climate,although maintain-ing such flows is highly likely to become increasinglydifficult under regional warming that shifts peakstreamflows to earlier in the year. More forcefulmeasures would include removing existing dams,asnow proposed for four dams on the lower SnakeRiver, and accepting the resultant reduced ability tomanage summer water shortages. Salmon haveevolved a great diversity of life histories and behav-iors to thrive in a highly variable and uncertain envi-ronment. Maintaining that diversity, which is highlylikely to require even greater efforts to preservehealthy, complex freshwater and estuarine habitat,islikely among the most effective options to enhancesalmon’s resilience to climate change.

3. Forests

Evergreen coniferous forests dominate the landscapeof much of the Northwest. In the west,coniferousforests cover about 80% of the land,including someof the world’s largest trees and most productiveforests,and about half the world’s temperate rainfor-est. Belying their lush appearance,on most sitesthese forests are constrained by moisture deficit dur-ing the warm dry summers,which limits seedlingestablishment and summer photosynthesis,and cre-ates favorable conditions for insect outbreaks andfires (Agee,1993; Franklin,1988). Forests of the dryinterior operate under an even more severe summersoil-moisture deficit (Law et al.,1999),which is con-sequently the dominant factor controlling the speciesdistribution and productivity of forests throughoutthe region (Zobel et al.,1976;Grier and Running,1977;Waring and Franklin,1979;Gholz,1982).

thermal limits to salmon survival is too simplistic(Emmet and Schiewe,1997;Bisbal and McConnaha,1998;Hargreaves,1997). Data from temperature-recording tags show that salmon move hourly anddaily through wide temperature ranges,presumablyby moving between surface and deep waters. Thisresult,and high-seas sampling of salmon,suggeststhat the effect of ocean temperature on salmon isnot primarily direct,but operates through changesin food supply (Walker et al.,1999; Pearcy et al.,1999).

Salmon are already beset with a long list of man-made problems,to which climate change is a poten-tially important addition. At sea, fishing pressure hasrecently been sharply curtailed,but wild salmonface intense competition from hatchery fish,someof which are being released at ten times the rate ofnatural smolt migrations. Onshore,the effects ofpresent climate variability on salmon in streams areswamped by clear-cutting, road building,and dams.The effects of future climate trends,and their poten-tial to interact with other stresses,are not known.Neither is the extent to which current and pro-posed measures to protect salmon by restoringstream habitat and changing dam operations willrestore depleted stocks and increase their resilienceto climatic stresses. If endangered-species listingsand public concern prompt a strong restorationresponse,it would likely take the form of far-reach-ing restrictions on land use near rivers and streams.Such measures would very likely have far-reachingsocial and economic impacts,which could greatlyexceed the direct effects of decreases in salmonabundance.

264


Figure 15: Old-growth Ponderosa Pine forest with open, grassyunderstory near Bend, Oregon. Source: S. Garrett and the McKayCollection, Bend, OR.

Forests throughout the Northwest have been pro-foundly altered by human intervention over the past150 years. West of the Cascades, forests have beencleared for conversion to other land uses or clear-cut for timber and replanted, replacing massive old-growth forests with young, even-aged managedforests. By various estimates,75 to 95% of originalold-growth forest has been logged,and much ofwhat remains is in small fragmented stands (Mac etal.,1998,p.646). The transition is estimated to havereleased to the atmosphere 2 billion metric tons ofcarbon over the century (Harmon et al.,1990). Eastof the Cascades,decades of intensive grazing, firesuppression,and selective harvesting of maturetrees have transformed the former open,park-likeforest of ponderosa pine,Douglas fir, and westernlarch into a dense mixed forest overstocked withshade-tolerant pines and firs. The new forest mix ishighly susceptible to insect outbreaks,disease andcatastrophic fire (Mason and Wickman,1988;Lehmkuhl et al.,1994). Of the ponderosa pine for-est that formerly comprised three quarters of east-side forests,more than 90% has been logged or lost(Figure 15). While the former open forest structurewas maintained by frequent,low-intensity fires (incontrast with western forests,whose natural regimewas of catastrophic fires at intervals of several cen-turies), fire suppression in the east has allowed largeaccumulation of fuel. These high fuel loads increasethe risk of extreme fires that replace stands overlarge areas (Quigley et al.,1996). The effects ofthese human activities overwhelmed climatic effectson Northwest forests during the 20th century, andcontinued forest management is highly likely tointeract strongly with climate effects during the 21st

century.

Tree growth can show a clear effect of climate vari-ability (which is why tree rings can provide a usefulrecord of past climate),most pronounced in standsnear their climatic limits,e.g.at the upper (cold) orlower (hot and dry) timberline. Figure 16 showsthe effect of 20 th century climate variability on threeNorthwest conifer populations,one at high and twoat low elevation (Peterson and Peterson,2000).Near the upper timberline,where trees are notmoisture-constrained, growth shows strong positivecorrelation with PDO. This is because positive PDOperiods tend to have lighter snowpack and conse-quently an earlier start to the high-elevation grow-ing season,promoting tree growth and upwardexpansion of forests to colonize sub-alpine mead-ows (Franklin et al.,1971;Rochefort et al.,1994).Near the lower timberline,the opposite relationshipis present. Growth is negatively correlated withPDO, because the warm dry winters and light snow-

pack of positive-PDO periods increase summerdrought stress,which is the limiting factor at theseelevations (Little et al.,1994; Peterson,1998). Inother regions,such as intermediate elevation standsin the interior, and the western hemlock and pacificsilver fir zones west of the Cascade crest,presentclimate variations have little discernible influenceon the structure and composition of most maturestands,which – once established – have substantialability to buffer themselves against climate variation.In such stands,competition and other factorsobscure present climate signals in individual trees(Brubaker, 1986;Dale and Franklin,1989).

The principal effect of climate on these forests hascome not through direct effects but indirectly,through changes in disturbance by fire,insect infes-tation,and disease (Overpeck et al.,1990; Fosberg etal.,1992;Ryan,1991). Major disturbances can resetforests to their establishment stage,when trees arethe most sensitive to adverse environmental condi-tions (Gardner et al.,1996). For insect and disease

265


Figure 16: Trees near their climatic limits show strong signals ofinter-decadal climate variability. Those near the upper treelinegrow best in warm-PDO years because snowpack is lighter, whilethose near the dry lower treeline grow worst in warm-PDO years,because of summer moisture deficit. Source: Peterson andPeterson, 2000.

Tree Growth and Inter-Decadal Climate Variabilit y

mortality, the available 20th-century data are inade-quate to quantify the region-wide effects of climatevariability, but smaller-scale studies have shownstrong correlations of both bark beetle and defolia-tor outbreaks with severe drought conditions(Swetnam and Lynch,1993). For fire,total

Figure 18: Under the Hadley scenario, the MAPSS (top row) andMC1 (bottom row) models project expansion of forests east of theCascades and contractions to the west, assuming increased water-use efficiency under elevated atmospheric CO 2 (right column).When no such increase is assumed (left column), projections arenearly unchanged in the MC1 model, but change to a large contrac-tion region-wide in the MAPSS model. Source: Bachelet et al(2000). – See Color Figure Appendix

Projected Northwest Vegetation Changes under twoEcosystem Models, 2100

Northwest forest area burned shows a significantregion-wide association with PDO and with thePalmer Drought Severity Index,especially before theintroduction of widespread fire suppression. Forexample,as Figure 17 shows,the warm-PDO periodof 1925-1945 had much more area burned than thecool-PDO periods immediately preceding and fol-lowing it (Mote et al.,1999a). The effect of ENSOon fire is less clear. In contrast with the strongENSO signal observed in wildfire in the SouthwestUS (Swetnam and Betancourt,1990),no region-wideeffect is evident in the Northwest. At smaller scales,however, one study of historical fire data and ongo-ing tree-ring studies suggest a significant ENSOeffect (Heyerdahl 1997;T. Swetnam,personal com-munication,1999).

Under projected future climate change,the effect onNorthwest forests is highly likely to reflect complexinteractions between several factors,some of whichvary with particular sites. The direct effect of pro-jected warmer summers without substantialincrease in rainfall would be to increase summersoil moisture deficit, resulting in reduced net photo-synthesis and tree growth,increased stress and treemortality, and decreased seedling survival. Reducedsnowpack has different effects at different sites:itextends the growing season and facilitates seedlingestablishment where snowpack is presently heavy,but reduces growing-season moisture availabilityand consequently increases drought stress in dryareas (Peterson,1998). One early empirical studyused observed associations between existing forestcommunities and local climate to project impacts offuture climate change on Northwest forests. Thisstudy projected that forested area in the Northwestwould contract,principally through forest diebackand sagebrush-steppe expansion at the dry lowertreelines east of the Cascades (Franklin et al.,1991).

It is likely, however, that the effects of increasedsummer drought will be offset to some degree bywetter winters,or by the direct effects of elevatedatmospheric CO2 concentration. Though droughtstress occurs in late summer, its severity dependsprincipally on winter and spring precipitation,because forests rely on moisture stored seasonally indeep soil layers to offset summer water deficits.Under climate model projections,the forest growingseason begins several weeks earlier in the springthan at present,making some increased winter pre-cipitation immediately available to forests. Beyondthis water put to use immediately, wetter winterscould mitigate summer drought stress still further, ifthe soil has enough storage capacity to hold theadditional water until it is needed in the late sum-

266


Figure 17: Prior to modern fire suppression, annual Northwest areaburned in forest fires showed a clear association with inter-decadalclimate variability. Dashed lines show PDO regime shifts. Source:Mote et al. (1999), p. 65.

Annual Northwest Area Burned in Forest Fires overthe 20th Century

MAPSS: Percent Change in Leaf AreaConstant CO2

MC1: Percent Change in Vegetation Carbon

Elevated CO2

Elevated CO2Constant CO2

mer. Although there are some indications thatNorthwest forest soils below the snow line arealready fully recharged under present winter rains,so precipitation increases would be lost as runoff(Harr, 1977; Jones and Grant,1999; Perkins,1997),the advance of growing season makes the signifi-cance of these results for seasonal water storageunder climate change ambiguous (Bachelet et al.,1998). Elevated CO2 concentration may also possi-bly mitigate productivity losses from summerdrought stress, by increasing photosynthetic effi-ciency or – likely more important in water-limitedNorthwest forests – by increasing trees’ water-useefficiency through reduced stomatal conductance.Laboratory and field studies have shown increasesin both net carbon uptake and water-use efficiencyin many plants (Bazzaz et al.,1996),but most suchstudies have examined agricultural crops, grasses,ortree seedlings. The evidence available on maturetrees is quite limited.One experimental study in ayoung pine plantation in North Carolina found thatelevated CO2 brought increased carbon assimilationwith no change in water use - an increase in water-use efficiency, but not through the expected mecha-nism of increased stomatal resistance (Ellsworth,1999).The applicability of this result to projectingresponses of mature connifer forests in the moremoisture-stressed Northwest remains unknown.Cool coniferous forest systems appear to be amongthe least responsive in aggregate to elevated CO2

(Ellsworth,1999;Mooney et al.,1999).

Process-based models are needed to represent andquantify these effects,because empirical studies canonly observe forests’ responses to the present rangeof climatic conditions with present CO2 concentra-tion (VEMAP Members,1995). With assumptions ofincreased water-use efficiency, ecological models

driven by the Hadley,Canadian,and other climate-change scenarios project opposite effects east andwest of the Cascade Crest. Cool coniferous foreststo the west are projected to contract,with reduc-tions in vegetation carbon or leaf area exceeding50% in some areas and replacement by mixed tem-perate forests over substantial areas. Dry forests tothe east are projected to expand,with increases invegetation carbon or leaf area also exceeding 50% insome locations (Neilson and Drapek,1998;Daly etal.,2000;Neilson et al.,1998). The increases in theeast are relative to a smaller present biomass,and soare smaller in absolute terms. While these modelresults reflect substantial recent progress,they havesignificant remaining weaknesses and uncertainties.For example,none adequately represents fog,whichcan comprise as much as half the water input toforests on some coastal and hillside sites. Moreover,results of different models in the Northwest differstrongly in their sensitivity to the water-use efficien-cy assumption. When no increase in efficiency isassumed,model results range from a very similarpattern of expansion in the east and contraction inthe west,to a region-wide contraction that resem-bles the projection of the earlier empirical analysis(Bachelet et al.,2000).

The magnitude and consequences of future changesin water-use efficiency represent the most impor-tant uncertainties in projecting the climate responseof Northwest forests over the next century. Whilethe balance of preliminary evidence suggests atmost a small water-use efficiency increase inNorthwest coniferous forests due to enhanced CO2,model results differ substantially in whether theeffect of this unknown on the extent,density, andspecies distribution of Northwest forests is large orsmall. Over the longer term,there are preliminary

267


Figure 19: The MC1 ecosystem model projects substantial increases in the amount of forest biomass burnt annually in the Northwest underboth the Hadley and Canadian scenarios. Source: Bachelet et al. (2000)

Northwest Forest Fire Projections

Hadley Scenario Canadian Scenario

cant increases in forest area and biomass burned inthe Northwest interior by 2100. Changes in otherdisturbances,such as wind,insects,and disease arealso highly likely under climate change and biomassburned in the Northwest interior by 2100 (e.g.,results from the MC1 model in Figure 19). Thepotential character of these disturbances under cli-mate change is not yet adequately understood.Some likely effects have been suggested for individ-ual disturbance processes,based on present distur-bance patterns and tree-ring studies. For example,general warming is likely to encourage northwardexpansion of southern insects,while longer growingseasons are highly likely to allow more insect gener-ations in a season. Forests that are moisture stressedare more susceptible to attack by sap-suckinginsects such as bark beetles. It is also possible thatdrought stress encourages attack by foliage eaterssuch as spruce budworm,although the evidence forthis is divided. Some studies show attack by foliageeaters increasing when foliage is lush,others whenit is stressed (Swetnam and Betancourt,1998;Thomson et al.,1984; Kemp et al.,1985;Swetnamand Lynch,1993;Larsson,1989;Price,1991). Verylittle is understood,however, about crucial questionsof the interactions between multiple disturbances,e.g.,between insect attack and fire,under projectedclimate change and the changes in forest characterthat follow (Neilson et al.,1994). Finally, it is crucialto note that ecosystem models only project poten-tial vegetation,the vegetation that would be presenton a site in the absence of human intervention. Inthe Northwest, forest management and land-usechange are presently, and are likely to remain,pre-dominant factors shaping the structure,species mix,and extent of forest ecosystems (Franklin andForman,1987). Interactions between these human-driven factors and the multiple direct and indirectpathways of climate influence on Northwest forestsare essentially unexamined,and are key areas forresearch (Sohngen et al.,1998).

Adaptation Options In contrast to water and salmon,where manage-ment already reflects at least limited awareness ofclimate variability, a survey of Northwest forestmanagers suggests they regard climate as unimpor-tant,because mature stands are resilient to wideclimatic ranges and because 40-70 year timber har-vest rotations average out the ef fects of shorter-term climate variation (Mote et al.,1999b,Section4.5,pp.73). Long-term climate trends are highlylikely to make these assumptions invalid. Trees arelikely to mature in a climate substantially differentthan when they were planted,possibly requiringchanges to many dimensions of forest management

indications that,increased evapotranspiration underwarmer temperatures could possibly overwhelmeven the largest assumptions of increased water-useefficiency, leading to substantial forest dieback.

The largest effects of future climate variability orchange on Northwest forests are likely to arise fromchanges in disturbances (McKenzie et al,1996;Tornand Fried,1992). Two dynamic vegetation modelsnow incorporate fire models,which project signifi-

268


Figure 20: This chart compares possible Northwest impacts fromclimate change by the 2050s with the effects of natural climate vari-ations during the 20 th century. The orange bars show the effects ofthe warm phase of the Pacific Decadal Oscillation (PDO), relative toaverage 20th century values. During warm-PDO years, theNorthwest is warmer, there is less rain and snow, stream flow andsalmon catch are reduced, and forest fires increase. The blue barsshow the corresponding effects of cool-phase years of the PDO,during which opposite tendencies occurred.

The pink bars show projected impacts expected by the 2050s,based on the Hadley and Canadian scenarios. Projected regionalwarming by this time is much larger than variations experienced inthe 20th century. This warming is projected to be associated with asmall increase in precipitation, a sharp reduction in snowpack, areduction in streamflow, and an increase in area burned by forestfires. Although quite uncertain, large reductions in salmon abun-dance ranging from 25 to 50%, are judged to be possible based onprojected changes in temperature and streamflow. Source: basedon Mote et al., 1999, pp. 27. – See Color Figure Appendix

Climate Change Projected for 2050 vsObserved 20th Century Variability

Temperature Change in annual average regional temperature (ºF)Precipitation Change in annual average regional precipitation (%)Snow depth Change in average winter snow depth at Snoqualmie Pass, WA (%)Streamflow Change in annual streamflow at The Dalles on the Columbia River

(corrected for changing effects of dams) (%)Salmon Change in annual catch of Washington Coho salmon (%)Forest fires Change in annual area burned by forest fires in WA and OR (%)

Cool PDOYearsWarm PDO Years Regional impacts of climate change in 2050s

(1900-1924, 1946-76)(1925-1945, 1977-95)

even beyond those recently adopted due to endan-gered species concerns. Required adaptations mightinclude:planting species best adapted to projectedrather than present climate (e.g.,planting Douglas firon suitable sites in the silver fir zone),or withknown broad climatic resilience;further measures torestore and maintain complexity of forest structureand composition within intensively managed areas;managing forest density for reduced susceptibility todrought stress;and using precommercial thinning,prescribed burning,and other means to reduce therisk of large,high-intensity disturbances and to facili-tate adaptation to changed climate regimes.

Managing forests effectively under these conditionsis very likely to require increased capacity for long-term monitoring and planning. In addition,theimportance of seasonal and interannual climate vari-ability is highly likely to increase when forests arealso stressed by long-term trends. Increased under-standing of climate variability and predictive skill canallow projected periods of drought stress and firerisk to be factored into short-term forest-manage-ment decisions such as timing and species of plant-ing,and use and timing of prescribed burning.

Maintaining forest ecosystem services and biologicaldiversity is also likely to grow increasingly challeng-ing under climate change. Options for doing sowould include establishment of further protectedareas,incorporating the maximum possible diversityof topography and landscape;active measures to pro-mote species migration and maintain diversity, evenin non-commercial forests;and further reduction inthe intensity of commercial harvest.

4. Coasts

The Pacific Northwest has three distinct coasts. Theinland marine waters of Puget Sound and the Straitof Juan de Fuca are bounded by narrow beachesbacked by steep bluffs,and contain large areas ofintensive shoreline development. The open Pacificcoast has rocky bluffs and headlands punctuated bysmall pocket beaches,with wide beaches and sand-spits in southern Washington and coastal sand dunesin central Oregon. This coast has generally low-inten-sity development,with no major cities and manystretches in parks,Indian reservations,and largeundeveloped parcels,but contains a few pockets ofincreasing development. Finally, the coastal estuariesof Oregon and southern Washington are principallybordered by farmlands and small towns at river-mouths,and support extensive shellfish aquacultureand harvesting in their shallow waters and on theirbroad mudflats.

Present stresses in coastal regions include: blufflandsliding from heavy winter rains,principally onthe steep hillsides around Puget Sound (Tubbs,1974);erosion of beaches,barrier islands,sandspitsand dunes on the open Pacific coast,principally dueto winter storm waves;coastal flooding near river-mouths,particularly in areas that have neitherupstream protection nor sufficient height abovehigh tide;loss of wetlands to development or ero-sion;and invasion by exotic species,particularly inthe coastal estuaries. Extensive development oncoastal bluffs and near beaches,mainly in PugetSound but increasingly also along the Pacific coast,has placed considerable valuable property at riskfrom erosion and landslides. Coastline near majorriver-mouths is sensitive both to ocean erosion andto variation in the rivers’sediment loads,whichhave increased on some rivers from adjacent clear-cutting,and decreased on others (including theColumbia) from damming (Canning,1991;Canningand Shipman,1994; Field and Hershman,1997; Parket al.,1993).

Little long-term data are available on coastal effectsof climate variability, but a few effects are evident.Severe storm surges and erosion events on the opencoast occur on average every five years. Theseappear to be associated with El Niño events,whichincrease erosion both by raising sea level for severalmonths and by changing the direction of winds andwaves from westerly to southwesterly (Komar andEnfield,1987). The 1997-1998 El Niño, for example,brought rapid erosion to a 1,000-foot built up seg-ment of the sandspit on which Ocean Shores,Washington is situated, reversing several centuriesof slow growth and requiring emergency construc-tion of an armored beach fill. One Oregon studyfound that construction of shore protection meas-ures follows an ENSO cycle,increasing sharply inthe years immediately following a strong El Niño(Good,1994). As well as causing extreme erosionevents,the elevated sea level during El Niño eventsalso increases inundation risk in the low-lying areasof southern Puget Sound such as the city ofOlympia,where it reaches 2-4 inches (5-10 cm). Incontrast,La Niña events bring reduced erosion onthe open coastline,but their heavier than normalwinter rainfalls increase soil saturation and landslid-ing risk on coastal bluffs (Gerstel et al.,1997). Forexample,the four years with highest landslide inci-dence in the Seattle area were all La Niña winters(1933-4,1985-6,1996-7,1998-9).

Suggestive indications of a PDO signal in coastalphenomena have also been noted,although theseremain speculative. For example,the warm decade

269


Adaptation Options The most effective adaptation strategies for coastalclimate impacts involve conserving remaining natu-ral coastal areas and placing less property at risk inlow-lying or flood-prone areas,on beaches,or onand below unstable slopes. Although these generaladaptation strategies are well known,a series ofinterviews with coastal-zone managers about howthey use climate information suggested that thecoastal management system is not particularlyadaptable even to current climate variability and itsassociated risks (Mote et al.,1999b,section 5.5,pp.80-81). Most managers reported that they are seri-ously constrained in their ability to incorporate cli-mate in planning,indeed that any climate considera-tions are overwhelmed in their planning by thepresent and potential endangered listings of varioussalmonid species. Many managers did not evenview the long-range threat of flooding or inundationas a significant risk to the resources they managed.As for restricting development in vulnerable loca-tions,there appears to be little inclination to movein that direction. A weaker and perhaps more feasi-ble alternative would be to assign more of the riskof living in a coastal zone to property-owners,through incorporating geological assessment intoproperty-insurance rates.

ADDITIONAL ISSUE

Agriculture

Due to limited time and resources in this firstAssessment,the Northwest regional study wasrestricted to the four critical issues discussed above.The choice of these four reflected the concerns ofstakeholders in the regional workshops,but wasemphatically not intended to imply that these arethe only important areas of impact in theNorthwest. Other areas of potentially significantimpact not covered in this Assessment mightinclude, for example,human health,urban quality oflife, recreation,and agriculture. Agriculture,becauseof its importance in many parts of the region,is aparticularly conspicuous omission,and an obviouspriority for subsequent assessment. A very prelimi-nary discussion is provided here,based on workconducted by the agriculture sector team of theNational Assessment.

The Agriculture Sector Assessment included model-ing of climatic effects on several crops at five loca-tions in the Northwest:Boise,Idaho;Medford andPendleton,Oregon;and Yakima and Spokane,Washington. The studies examined dryland and irri-

of the 1980s, following the shift to warm-phase PDOin the late 1970s, was marked by two striking shiftsin the ecosystem of Willapa Bay in southernWashington. The exotic cordgrass (Spartina),whichwas introduced to the bay nearly 100 years earlier,began a rapid expansion that threatened localspecies for the first time (Feist and Simenstad,2000).After several productive decades,the condition ofcommercially important oysters in Willapa Bay alsobegan a substantial decline in the late 1970s,thoughother factors such as pollution in the bay could alsobe responsible (Mote et al.,1999b,section 5.2,pp.77-78).

Several coastal effects of future climate change havebeen identified,although detailed assessment ofthese remains to be done. Future climate warming islikely to raise mean sea level 10 to 35 inches in the21st century, as opposed to the 4 to 8 inch rise of the20th century. The apparent rise will differ from placeto place,partly due to regional differences in oceancirculation and heat content — for example,theHadley model projects a larger sea-level rise on thePacific than the Atlantic coast of North America —and partly due to local variation in the rate of upliftor subsidence of the land surface. In the PacificNorthwest, regions of uplift are centered at themouth of the Strait of Juan de Fuca, rising at 0.1 inch(2.5 mm) per year, and the mouth of the ColumbiaRiver, rising by 0.06 inch (1.7 mm) per year, whilesouthern Puget Sound is subsiding at up to 0.08 inch(2 mm) per year (Shipman,1989). Consequently,risks of sea-level rise are greatest in southern PugetSound. Here,low-lying settlements are already at riskof inundation and existing shoreline protection isinadequate even for the high end of projected meansea level rise,when the far greater risk is from meanrise combined with storm surge (Craig,1993).

Higher mean sea level is also likely to increase sedi-ment erosion and redistribution on the open coast,which possibly may be further amplified by project-ed shifts in prevailing wind direction to resemblesustained El Niño conditions. In addition,projectedheavier winter rainfall is likely to increase soil satura-tion,landsliding,and winter flooding. All thesechanges would increase the risk to property andinfrastructure on bluffs and beachfronts,and besiderivers. Climate change could also bring continuedchanges in coastal and estuarine ecosystems throughchanges in runoff and warmer water temperatures,with possible increased risk of exotic species intro-duction or health risks from shellfish contamination.

270


gated yields,and water use for irrigation,under sev-eral climate scenarios and using several crop mod-els. Under all scenarios,dryland yields for mostcrops were projected to increase through the 21st

century. The exception was potatoes,whose dry-land yields by 2090 declined by as much as 30-35%,with the largest declines in Idaho. Changes in irri-gated yields and irrigation water requirements weremore mixed. Wheat and potatoes showed largereductions in irrigated yields (7-25% and 35-40%,respectively by 2090),while hay increased 50-70%and tomatoes showed a mixed trend,increasing 15-20% by 2030,then declining to roughly presentyields by 2090. Irrigation water needs declined by20-40% for wheat,and increased by 10-40% for hay,potatoes and tomatoes. To the extent that aggregateirrigation water needs increase under the combinedeffects of climate change,CO2 enrichment,and agri-cultural response,this would exacerbate the sum-mer water shortages and allocation conflicts dis-cussed above. Interactions between climate andrelated impacts on agriculture and forest lands,andlinkages between them as mediated by human land-use conversion and management,are little under-stood and remain important knowledge needs (Aliget al.,1998).

CRUCIAL UNKNOWNS ANDRESEARCH NEEDSAlthough this Assessment has produced and synthe-sized significant advances in our understanding ofclimate impacts in the four key areas examined indetail, clear needs for additional knowledge are evi-dent for each of them,in order to understandimpacts and vulnerabilities more thoroughly andassess potential responses.

For freshwater, the analysis here has concentratedprimarily on the response of the Columbia River sys-tem to climate variability and change,and on theconsequences of projected summer shortages onthe reliability with which various present manage-ment objectives could be met under potential futureflow regimes. Further analysis is required to:projectfuture shifts in demand for multiple uses,and toexamine the joint effects of climate variability andchange on seasonal supply and demand;to identifythe degree of vulnerability to scarcity, assess theeffects of alternative operational rules and allocationschemes,and identify and evaluate specific responseoptions. In addition,this Assessment has made onlypreliminary investigation of other basins than theColumbia. In particular, further examination is need-ed to characterize the effects of future climate vari-

ability and change on winter flooding risk in thelow-elevation rainfed and mixed basins west of theCascades. In view of the rapid development ofwest-side metropolitan areas,assessment of these cli-mate risks,of how alternative future developmentpatterns may compound or mitigate it,and of poten-tial responses,are of high priority.

For salmon, research priorities include further char-acterization of historical climate influences onsalmon and identification of the mechanisms bywhich they operate. In particular, very little isknown of the effect of climate variability andchange on the open-ocean phase of salmon’s lifecycle. For forests,the top priority is further workon the effects of climate and atmospheric CO2 onNorthwest forests’ water use efficiency,which isnecessary to understand even the direction of futureclimate effects on forests. Other priorities includeinteractions between climate and forest distur-bances,including fire,insects,and disease,as well asinteractions between these disturbances. Finally, lit-tle work has been done quantifying the effects ofclimate variability on the region’s coasts.

Assessment of other sectors and issues than the fourexamined in detail here will be required. Particularlyimportant areas for investigation in subsequentNorthwest climate assessments will be agriculture,energy, and urban issues (e.g.,infrastructure,haz-ards,and quality of life).

More broadly, there are several areas of climateresearch that are important for better understandingNorthwest impacts. This assessment has been basedon a single run each of two primary climate models,each using the same emissions scenario,with someadditional results drawn from single runs of othermodels. A more reliable regional assessment wouldrequire controlled regional-level comparison of sev-eral state-of-the-art models,each with a statisticalensemble of multiple similar runs under each of sev-eral emissions scenarios. Ensembles of multipleruns for each model are necessary to allow examina-tion of patterns of climate variability projected byeach model. For example,several results in thisassessment were strongly influenced by the fact thatin the particular Hadley model run employed,the2020s were an unusually wet decade. Studies ofsuch variability over ensembles of multiple modelruns are necessary to interpret such excursions.Comparisons of multiple models with several emis-sion scenarios would allow useful model compari-son and explanation of significant regional-scale dis-parities,and examination of how major impacts varywith higher or lower future emissions.

271


methods to couple socioeconomic projections toclimatic,ecological,and hydrological models toallow more precise examination of impacts andpotential responses. These methods should permitthe examination of alternative sets of technologicaland institutional assumptions,and should supportcomprehensive examination of uncertainties acrosssocioeconomic, climatic,and ecological domains.

Furthermore,more insight is needed into the feasi-bility and likely consequences of various adaptationstrategies,addressing sectors both singly and jointly.This Assessment has made only a preliminary identi -fication of potential adaptation measures for eachsector, and has not attempted to assess their costs,benefits,efficacy, or ancillary ef fects. Systematicexamination of technological,managerial,and insti-tutional adaptation options should be conducted,based on partnerships between researchers,stake-holders,and experienced resource managers. Suchpartnership will be necessary to identify, develop,and assess adaptation options that are effective,lowin cost,and are practical to implement in the con-text of present management practices,which strong-ly shape climate impacts. For example,future forest-management and agricultural practices are very like-ly to strongly mediate climate effects on these sec-tors,and may also contribute to mitigation of cli-mate change through adjustments to managementthat increase sequestration of carbon in forests andsoils. Where the present assessment has identifiedthat climate variability and change are inadequatelyconsidered even in long-lived investment, resourcemanagement and infrastructure design decisions,research is needed to identify the cause and poten-tial approaches to increase the time-horizon of plan-ning. Two aspects of adaptation will be particularlyimportant:investigation of interactions with impactsand responses in Canada,since many of the affectedresources are shared;and linkages,whether conflict-ing or complementary, between adaptation meas-ures in multiple sectors. For example,increasingwater storage to manage increased summer droughtsuggests maintaining present dams or buildingmore,while restoring salmon habitat to reduce theirvulnerability to climate change would suggest theopposite. The most useful approach to assessmentwould examine impacts and adaptation measurestogether with emission scenarios and mitigationmeasures,to identify the most cost-effective strategyfor dealing with climate-change in aggregate.

Understanding the dynamics of longer-term variabili-ty, in particular whether and how presentlyobserved patterns would shift under a global green-house trend, requires further development of cli-mate models. Models are now beginning to repre-sent ENSO, but cannot yet reproduce interdecadalvariability of the size and character observed,eitherin the PDO or elsewhere in the world.

More accurate fine-scale modeling of climate-topog-raphy interactions in the Northwest is also required,and their effects on the region’s rivers,estuaries andcoasts,to understand several processes that stronglyaffect multiple areas of regional impacts. Theseprocesses include, for example,coastal upwelling,the interaction of fresh and salt water in estuaries,windstorms,and rain-on-snow events.

Because interannual and interdecadal climate var i-ability exert such strong influences on theNorthwest,better understanding of their dynamicsand their relationship to long-term climate trends isalso required. Their effects on summer climate arepotentially important and have been little studied.An immediate uncertainty is whether the PDO hasrecently re-entered its cool phase,and indeed,whether the PDO will continue to exhibit more orless regular phase changes. If the PDO continues tobehave for the next few decades as it has over the20th century, and it has re-entered a cool phase,theresultant cooler wetter regional climate would belikely to partly offset greenhouse warming over thenext two or three decades,until the PDO reversesagain. While this could delay the onset of significantimpacts from climate change,it could also obscurethe need to undertake long-lead adaptation meas-ures when they can be done with the least disrup-tion.

Finally, a large but crucial area for further researchconcerns interactions between climate changes,theecological and hydrological impacts discussed here,and human responses. Land-use change has beenthe dominant source of environmental stresses inthe Northwest over the 20th century, and furtherpopulation and economic growth are highly likelyto bring more pressure for conversion,with com-plex interactions between metropolitan develop-ment, forestry, and other land uses. Coherent sce-narios of socioeconomic futures in the Northwestthat elaborate more climate-relevant aspects,espe-cially land use and development,are needed. So are

272


LITERATURE CITEDAgee, J. K., Fire Ecology of Pacific Northwest Forests,Island Press,Washington,DC,1993.

Alig,R. J., D. M.Adams,B.A.McCarl,Impacts of incorpo-rating land exchanges between forestry and agriculturein sector models, Journal of Agricultural and AppliedEconomics, 30, 2 (December),389-401,1998.

Bachelet, D.,M.Brugnack,and R. P. Neilson,Sensitivityof a biogeography model to soil properties, EcologicalModeling, 109, 77-98,1998.

Bachelet, D.,R. P. Neilson, J. M.Lenihan,and R. J. Drapek,Climate change effects on vegetation distribution andcarbon budget in the US, Ecosystems, in review, 2000.

Baker, P. F.,T. P. Speed,and F. K.Ligon,Estimating theinfluence of temperature on the survival of chinooksalmon smolts migrating through the Sacramento SanJoaquin River delta of California, Canadian Journal ofFisheries and Aquatic Sciences, 52, 855-863,1995.

Baker,W. L.,Long-term response of disturbance land-scapes to human intervention and global change,Landscape Ecology, 10,143-159,1989.

Bazzaz, F.A.,S.L.Bassow, G. M.Berntson,and S.C.Thomas,Elevated CO2 and terrestrial vegetation:Implications for and beyond the global carbon budget,in Global Change and Terrestrial Ecosystems, editedby B.Walker and W. Steffen, pp.43-76,CambridgeUniversity Press,New York,NY, 1996.

Beamish,R. J.,and D. R.Bouillon, Pacific salmon produc-tion trends in relation to climate, Canadian Journal ofFisheries and Aquatic Sciences, 50, 1002-1016,1993.

Beechie,T.,E.Beamer, and L.Wasserman,EstimatingCoho salmon rearing habitat and smolt production loss-es in a large river basin,and implications for habitatrestoration, North American Journal of FisheriesManagement, 14, 797-811,1994.

Bisbal, G.A.,and W. E.McConnaha,Consideration ofocean conditions in the management of salmon,Canadian Journal of Fisheries and Aquatic Sciences,55, 2178-2186,1998.

Boer, G. J., N.A.McFarlane,R.Laprise, J. D. Henderson,and J.-P. Blanchet,The Canadian Climate Centre spectralatmospheric General Circulation Model, Atmosphere-Ocean, 22(4),397-429,1984.

Boer, G. J., G. M.Flato,M.C.Reader, and D. Ramsden,Atransient climate change simulation with historical andprojected greenhouse gas and aerosol forcing:Experimental design and comparison with the instru-mental record for the 20th century, Climate Dynamics,16, 405-426,1999a.

Boer, G. J., G. M.Flato,and D. Ramsden,A transient cli -mate change simulation with historical and projectedgreenhouse gas and aerosol forcing:projected climatefor the 21st century, Climate Dynamics, 16, 427-450,1999b.

Bottom, D. L.,Restoring salmon ecosystems: Myth andreality, Restoration and Management Notes, 13, 162-170,1995.

Brubaker, L.B., Responses of tree populations to cli -matic change, Vegetation, 67, 119,1986.

Bury, R.B.,Vertebrates in the Pacific Northwest:Species richness,endemism,and dependency on old-growth forests,in Biological Diversity: Problems andChallenges, edited by S.K.Majumdar, F. J. Brenner, J. E.Lovich, J. F. Schalles,and E.W. Miller, pp.392-404,Pennsylvania Academy of Science,Easton, Pennsylvania,1994.

Callahan,B.,E.Miles,and D. Fluharty. Policy implica-tions of climate forecasts for water resources manage-ment in the Pacific Northwest. Policy Sciences 32 :269-293,1999.

Canning, D. J.,Sea level rise in Washington State: State-of-the-knowledge,impacts,and potential policy issues,Washington Department of Ecology, Olympia,Washington,1991.

Canning, D. J.,and H.Shipman,Coastal erosion manage-ment studies in Puget Sound, Washington: Volume 1,executive summary, report 94-94,WashingtonDepartment of Ecology, Olympia,Washington,1994.

Carter, M. F.,and K.Barker,An interactive database forsetting conservation priorities for western neotropicalmigrants,in Status and Management of NeotropicalMigratory Birds, edited by D. M. Finch and P.W. Stangel,US Forest Service General Technical Report RM-GTR-229, pp.120-144, Rocky Mountain Forest and RangeExperiment Station,Ft.Collins,Colorado,1993.

Cohen,S. J.,K.A.Miller,A. F. Hamlet,and W.Avis,Climatechange and resource management in the ColumbiaRiver Basin, Water International, in press,2000.

273


littoral flats in Willapa Bay,Washington, Estuaries,23(2),268-275,2000.

FEMAT (Forest Ecosystem Management AssessmentTeam), Forest ecosystem management: An ecological,economic,and social assessment,US Department of theInterior, US Department of Agriculture,US Departmentof Commerce,and US Environmental ProtectionAgency,Washington,DC,729 pp.,1993.

Field, J.,and M.Hershman,Assessing coastal zone sensi-tivity and vulnerability to regional climate variabilityand change in the Pacific Northwest,Climate ImpactsGroup,University of Washington,Seattle,Washington,1997.

Flato, G. M., G. J. Boer,W. G. Lee, N.A.McFarlane, D.Ramsden, M.C.Reader, and A. J.Weaver,The CanadianCentre for Climate Modelling and Analysis global cou-pled model and its climate, Climate Dynamics, inpress,2000.

Fosberg,M.A.,L. O. Mearns,and C.Price,Climatechange: Fire interactions at the global scale:Predictions and limitations of methods,in Fire in theEnvironment, edited by P. J. Crutzen and J. G.Goldammer,Wiley and Sons,New York,1992.

Francis,R.C.,Sustainable use of salmon: Its effects onbiodiversity and ecosystem function,in HarvestingWild Species —- Implications for BiodiversityConservation, edited by C.H. Freeze,pp.626-670,Johns Hopkins University Press,Baltimore,Maryland,1997.

Francis,R.C.,S.R.Hare,A.B.Hollowed,and W. S.Wooster, Effects of interdecadal climate variability onthe oceanic ecosystems of the Northeast Pacific,Fisheries Oceanography, 7, 1-21,1998.

Franklin, J. F., Pacific Northwest forests,in NorthAmerican Terrestrial Vegetation, edited by M. G.Barbour and W. D. Billings,pp.103-130,CambridgeUniversity Press,Cambridge,United Kingdom,1988.

Franklin, J. F.,and C.T. Dyrness, Natural Vegetation ofOregon and Washington, Oregon State UniversityPress,Corvallis,Oregon,1973.

Franklin, J. F.,and R.T.T. Forman,Creating landscape pat-terns by forest cutting: Ecological consequences andprinciples, Landscape Ecology, 1,5-18,1987.

Franklin, J. F.,W. H.Moir, G.W. Douglas,and C.Wiberg,Invasion of subalpine meadows by trees in the CascadeRange, Arctic and Alpine Research, 3, 215-224,1971.

Craig, D.,Preliminary assessment of sea level rise inOlympia,Washington: Technical and policy implica-tions,City of Olympia Public Works Department Policyand Program Development Division,Olympia,Washington,1993.

Dale,V. H.,and J. F. Franklin, Potential effects of climatechange on stand development in the PacificNorthwest, Canadian Journal of Forest Resources, 19,1581,1989.

Daly C, D. Bachelet J. M.Lenihan,R. P. Neilson,W. Partonand D. Ojima,Dynamic simulation of tree-grass interac-tions for global change studies, EcologicalApplications, 10(2),449-469,2000

dell’Arciprete, P., N. Mantua,and R.C. Francis,Theinstrumental record of climate variability in the PacificNorthwest,Progress report for Year 1,JISAO ClimateImpacts Group,University of Washington,Seattle,Washington,1996.

dell’Arciprete, P., D. L. Peterson,R.C. Francis,and N.Mantua,Climate reconstruction through tree growthindices, unpublished manuscript, available from theUniversity of Washington Climate Impacts Group,Seattle,Washington,1998.

Detling,L.E.,Relic islands of xeric flora west of theCascade Mountains in Oregon, Madrono, 12, 39-47,1953.

Detling,L.E.,Historical background of the flora of thePacific Northwest, Bulletin No. 13, Museum of NaturalHistory, University of Oregon,Eugene,Oregon,1968..

Ellsworth, D. S.,CO2 enrichment in a maturing pine for-est: Are CO2 exchange and water status in the canopyaffected?, Plant, Cell, and Environment, 22, 461-472,1999

Emmett,R.L.,and M.H.Schiewe (Eds.),Estuarine andocean survival of Northeast Pacific salmon:Proceedings of the workshop,US Department ofCommerce,NOAA Technical Memo.,NMFS-NWFSC-29,1997.

Emmett,R.L.,and R. D. Brodeur, Recent changes in thepelagic nekton community off Oregon and Washingtonin relation to some physical oceanographic conditions,North Pacific Anadromous Fish Commission,Bull.No.2,in press,2000.

Feist,B.E.,and C.A.Simenstad,Expansion rates andrecruitment frequency of exotic smooth cordgrass,(Spartina alterniflora) (Loisel) colonizing unvegetated

274


Franklin, J. F.,et al., Effects of global climatic change onforests in northwestern North America, NorthwestEnvironmental Journal, 7,233-254,1991.

Gardner, R.H.,W.W. Hargrove,M. G.Turner, and W. H.Romme,Climate change,disturbances,and landscapedynamics,in Global Change and TerrestrialEcosystems, edited by B. Walker and W. Steffen,pp.,149-172,Cambridge University Press,Cambridge,UnitedKingdom,1996.

Gerstel,W. J.,M. J. Brunengo,W. S.Lingley, Jr.,R.L.Logan,H.Shipman,and T. J.Walsh, Puget Sound bluffs: Thewhere,why, and when of landslides following the holi-day 1996-97 storms, Washington Geology, 25(1),17-31,1997.

Gholz,H.L.,Environmental limits on aboveground netprimary production,leaf area,and biomass in vegetationzones of the Pacific Northwest, Ecology, 63, 469,1982.

Giorgi, F.,and G. Bates,On climatological skill of aregional model over complex terrain, Monthly WeatherReview, 117, 2325-2347,1989.

Giorgi, F.,C.S.Brodeur, and G.T. Bates,Regional climatechange scenarios over the United States produced witha nested regional climate model, Journal of Climate, 7,375-399,1994.

Glantz,M.H.,Consequences and responsibilities indrought forecasting: The case of Yakima,1977, WaterResources Research, 18,(1),pp.3-13,1982.

Good, J.W.,Shore protection policy and practices inOregon: An evaluation of implementation successes,Coastal Management, 22, 325-352,1994.

Graumlich,L. J.,Precipitation variation in the PacificNorthwest (1675-1975) as reconstructed from treerings, Annals of the Association of AmericanGeographers, 77, 19-29,1987.

Graumlich,L. J.,Subalpine tree growth, climate,andincreasing CO2: An assessment of recent growth trends,Ecology, 72,1-11,1991.

Gray, K. N.,The impacts of drought on Yakima Valley irri-gated agriculture and Seattle municipal and industrialwater supply, master’s thesis,School of Marine Af fairs,University of Washington,Seattle,Washington,1999.

Grier, C.C.,and S.Running,Leaf area of mature north-western coniferous forests: Relation to site water bal -ance, Ecology, 58, 893,1977.

Hamlet,A. F.,and D. P. Lettenmaier, Effects of climatechange on hydrology and water resources objectives inthe Columbia River Basin, Journal of American WaterResources Association, 35(6),1597-1623,1999.

Hare,S. J., N. J. Mantua,and R.C. Francis,Inverse pro-duction regimes: Alaskan and West Coast Pacificsalmon, Fisheries, 24, 6-14,1999.

Harmon,M.E., W. K. Ferrell,and J. F. Franklin,Effects oncarbon storage of converting old-growth forests toyoung forests, Science, 247, 699-702,1990.

Hargreaves,B. N.,Early ocean survival of salmon offBritish Columbia and impacts of the 1983 and 1991-1995 El Niño events,in Estuarine and OceanSurvival of Northeast Pacific Salmon: Proceedings ofthe Workshop, edited by R.L.Emmett and M.H.Schiewe,pp.,197-211,NOAA Tech.Memo., NMFS-NWFSC-29,Seattle,Washington,1997.

Harr, R. D.,Water flux in soil and subsoil in a steepforested slope, Journal of Hydrology, 33, 37-58,1977.

Haynes,R.W., D. M.Adams,and J. R.Mills,The 1993 RPATimber Assessment update,USDA Forest Service,RockyMountain Forest and Range Experimental Station,General Technical Report RM-GTR. March 1995.

Henjum,M. G., J. B.Karr, D. L.Bottom, D. Perry, J. C.Bednarz,S. G.Wright,S.A.Beckwitt,and E.Beckwitt,Interim protection for late-successional forests, fish-eries,and watersheds: National forests east of theCascade crest,Oregon,and Washington, The WildlifeSociety Technical Review 94-2, Bethesda,Maryland,245pp.,1994.

Heyerdahl,E.K.,Spatial and temporal variation in his-torical fire regimes of the Blue Mountains,Oregon,andWashington: The influence of climate, Ph.D. disserta-tion,College of Forest Resources,University ofWashington,Seattle,Washington,1997.

Huffaker, R., N. K.Whittlesey, and P. R.Wandschneider,Institutional feasibility of contingent water marketingto increase migratory flows for salmon on the UpperSnake River, Natural Resources Journal, 33, 671-696,Summer 1993.

Hulme,M.,E.M.Barrow, N.W.Arnell, P.A.Harrison,T. C.Johns,and T. E.Downing, Relative impacts of human-induced climate change and natural variability, Nature,397, 688-691,1999.

275


Laprise,R., D. Caya,M.Giguere, G. Bergeron,H.Cote, J.-P.Blanchet, G. J. Boer, and N.A.McFarlane,Climate and cli-mate change in western Canada as simulated by theCanadian Regional Climate Model, Atmosphere-Ocean,36, 119-167,1998.

Larsson,S.,Stressful times for the plant stress-insectperformance hypothesis, Oikos, 56, 277-283,1989.

Law, B.E.,R.H. Waring, P. M.Anthoni,and J. D.Abers,Measurements of gross and net ecosystem productivityand water vapor exchange of a Pinus ponderosaecosystem,and an evaluation of two generalized mod-els, Global Change Biology, 5, 1-15,1999.

Lehmkuhl, J. F., P. F. Hessburg,R.L.Everett,M.H.Huff,and R. D. Ottmar, Historical and current forest land-scapes of eastern Oregon and Washington, Part 1:Vegetation pattern and insect and disease hazards,General Technical Report PNW-GTR-328, USDA ForestService,Corvallis,Oregon,1994.

Leung,L.R.,and S. J. Ghan, Pacific Northwest climatesensitivity simulated by a regional climate model drivenby a GCM. Part I: Control simulations, Journal ofClimate, 12(7),2010-2030,1999a.

Leung,L.R.,and S. J. Ghan, Pacific Northwest climatesensitivity simulated by a regional climate model drivenby a GCM. Part II: 2 times CO2 simulations, Journal ofClimate, 12(7),2031-2053,1999b.

Little,R.L., D. L. Peterson,and L.L.Conquest,Regeneration of subalpine fir (Abies lasiocarpa) fol-lowing fire: Effects of climate and other factors,Canadian Journal of Forest Research, 24, 934-944,1994.

Mac,M. J., P.A.Opler, C.E.Puckett Haecker, and P. D.Doran (Eds.),Status and trends of the Nation’s biologi-cal resources, US Geological Survey, US Departmentof the Interior, Biological Resources Division,Reston,Virginia,1998.

Mantua, N. J.,S.R.Hare,Y. Zhang, J. M.Wallace,and R.C.Francis,A Pacific interdecadal climate oscillation withimpacts on salmon production, Bulletin of theAmerican Meteorological Society, 78, 1069—1079,1997.

Mantua, N. J., P. dell’Arciprete,and R.C. Francis, Patternsof climate variability in the Pacific Northwest: Aregional 20th century perspective,in Impacts of cli-mate variability and climate change in the PacificNorthwest: An integrated assessment,edited by E.L.Miles,Climate Impacts Group,University ofWashington,Seattle,Washington,1999.

Jackson, P. L.,and A. J. Kimmerling,(Eds.), Atlas of thePacific Northwest, 8th edition, Oregon State UniversityPress,Corvallis,Oregon,1993.

Johns T. C.,R.E.Carnell, J. F. Crossley, J. M.Gregory, J. F. B.Mitchell,C.A.Senior, S. F. B.Tett,and R.A.Wood,TheSecond Hadley Centre coupled ocean-atmosphereGCM: Model description,spinup,and validation,Climate Dynamics, 13, 103-134,1997.

Jones, J.A.,and G. E.Grant, Peak flow responses toclear-cutting and roads in small and large basins, west-ern Cascades,Oregon, Water Resources Research, 32,959-974,1996.

Jones, J.A.,and G. E.Grant,Hydrologic processes andpeak discharge response to forest harvest, regrowth,and roads in ten small experimental basins, westernCascades,Oregon, Water Resources Research , inreview, 2000.

Keane,R.E.,and S. F.Arno,Rapid decline of whitebarkpine in western Montana: Evidence from 20-yearremeasurements, Western Journal of Applied Forestry,8(2),44-47,1993.

Kellogg,E.,(Ed.),Coastal temperate rain forests:Ecological characteristics,status,and distributionworldwide, Ecotrust Occasional Paper Series 1,Ecotrust and Conservation International, Portland,Oregon,and Washington,DC,64 pp.,1992.

Kemp,W. P., D. O. Everson,and W. G.Wellington,Regional climatic patterns and western spruce bud-worm outbreaks,USDA Forest Service CooperativeState Research Service, Technical Bulletin Number1693, USDA,Corvallis,Oregon,1985.

Kendall,K.C.,Whitebark Pine: Ecosystem in peril, inOur living resources: A report to the Nation on thedistribution, abundance, and health of US plants,animals, and ecosystems, edited by E.T. LaRoe, G. S.Farris,C.E.Puckett, P. D. Doran,M .J. Mac,pp.228-230,US National Biological Service, Washington,DC,1995.

Klahn, J.,This time,city folks will bear cost of restoringspecies,(AP), Seattle Times, pg.1,September 5,1999.

Komar, P. D.,and D. B.Enfield,Short-term sea-levelchanges and coastal erosion,in Sea-level Fluctuationsand Coastal Evolution, Special Publication 41, editedby D. Nummedal, O. Pilkey, and J. D. Howard,pp.17-27,Society of Economic Paleontologists andMineralologists,1987.

276


Marcot,B. G.,R.S.Holthausen, J.Teply, and W. D. Carrier,Old-growth inventories: Status,definitions,and visionsfor the future,in Wildlife and Vegetation inUnmanaged Douglas-Fir Forests, edited by L. F.Ruggiero,K.Aubry, and M.H.Huff, US Forest ServiceGeneral Technical Report PNW-GTR-285, pp.47-60,Pacific Northwest Research Station, Portland,Oregon,1991.

Mason,R.R.,and B.E.Wickman,The Douglas-fir tussockmoth in the interior Pacific Northwest, chapter 10,inDynamics of Forest Insect Populations, edited by A.A.Berryman,Plenum Press,New York,1988

Matthews, D.A.,and T. Hovland,Nested model simula-tions of regional orographic precipitation,GlobalClimate Change Response Program Report,Bureau ofReclamation,Denver, Colorado, 61 pp.,1997.

McCullough, D.A.,A review and synthesis of ef fects ofalterations to the water temperature regime on fresh-water life stages of salmonids,with special reference toChinook Salmon,EPA Region 10 Report 910-R-99-010,July 1999.

McFarlane, N.A., G. J. Boer, J. P. Blanchet,and M.Lazare,The Canadian Climate Centre second-generation gener-al circulation model and its equilibrium climate,Journal of Climate, 5, 1013-1044,1992.

McKenzie, D., D. L. Peterson,and E.Alvarado,Predictingthe effect of fire on large-scale vegetation patterns inNorth America, USDA Forest Service Research PaperPNW-489, Fort Collins,Colorado,1996.

Miles,E.L.,A.Hamlet,A.K.Snover, B.Callahan,and D.Fluharty. 2000. Pacific Northwest regional assessment:The impacts of climate variability and climate changeon the water resources of the Columbia River Basin.Journal of American Water Resources Association,36(2):399-420,2000.

Miller,A. J., D. R.Cayan,T. P. Barnett, N. E.Graham,and J.M.Oberhuber,The 1976-1977 climate shift in thePacific Ocean, Oceanography, 7, 21-26,1994.

Miller,A. J., D. R.Cayan,and W. B.White,A westward-intensified decadal change in the North Pacific thermo-cline and gyre-scale circulation, Journal of Climate, 11,3112-3127,1998.

Miller, K.A., Water banking to manage supply variability,in Advances in the Economics of EnvironmentalResources, vol.1.,edited by D. C.Hall,pp.185-210, JAIPress,Greenwich,Connecticut,1996.

Miller, K.A.,S.L.Rhodes,L. J. MacDonnell,Water alloca-tion in a changing climate: Institutions and adaptation,Climatic Change, 35,157-177,1997.

Miller, R. F.,and J.A.Rose,Western juniper expansion ineastern Oregon, Great Basin Naturalist, 55, 37-45,1995.

Miller, R. F.,and P. E.Wigand,Holocene changes in semi-arid pinyon-juniper woodlands, BioScience, 44(7), 465-474,1994.

Mitchell J. F. B.,T. C. Johns, J. M.Gregory, and S.Tett,Climate response to increasing levels of greenhousegases and sulphate aerosols, Nature, 376, 501-504,1995.

Mitchell J. F. B.,and T. C. Johns,On modification of glob-al warming by sulfate aerosols, Journal of Climate,10(2),245-267,1997.

Mooney, H.A., J. Canadell, F. S.Chapin,III, J. R.Ehleringer, C.H.. Korner, R.E. McMurtrie,W. J. Parton,L.F. Pitelka,and E.-D. Schulze,Ecosystem physiologyresponses to global change,in, The TerrestrialBiosphere and Global Change: Implications forManaged and Unmanaged Ecosystems, edited by B.Walker,W. Steffen, J. Canadell,and J. Ingram,pp.141-189,Cambridge University Press,Cambridge,UK,1999.

Morse,L.E., J.T. Kartesz,and L.S. Kutner, Native vascularplants,in Our living resources: A report to the Nationon the distribution, abundance,and health of US plants,animals,and ecosystems,edited by E.T. LaRoe, G. S.Farris,C.E.Puckett, P. D. Doran,M. J. Mac,US NationalBiological Service, Washington,DC,1995.

Mote, P.W.,W. S. Keeton,and J. F. Franklin,Decadal varia-tions in forest fire activity in the Pacific Northwest,11th Conference on Applied Climatology,AmericanMeteorological Society, Boston,Massachusetts,1999a.

Mote, P.W.,et al.,Impacts of climate variability andchange: Pacific Northwest,A report of the PacificNorthwest Regional Assessment Group for the USGlobal Change Research Program,Climate ImpactsGroup,University of Washington,Seattle,Washington,November 1999b.

Myers, J. M.,et al.,Status review of chinook salmonfrom Washington,Idaho,Oregon,and California,NOAATech.Memo.NMFS-NWFSC-35,1998.

277


Pearcy,W. G., Ocean Ecology of North PacificSalmonids, University of Washington Press,Seattle,Washington,1992.

Pearcy,W. G.,K.Y.Aydin,and R. D. Brodeur,What is thecarrying capacity of the North Pacific Ocean forsalmonids?, Pisces Press, 7, 2,17-22,1999.

Perkins,R.,Climatic and physiographic controls on peak-flow generation in the western Cascades,Oregon, Ph.D.dissertation,Oregon State University, Corvallis,Oregon,1997.

Peterson, D. L.,Climate,limiting factors and environmen-tal change in high-altitude forests of western NorthAmerica,in Climatic Variability and Extremes: TheImpact on Forests, edited by M.Benistion and J. L.Innes,Springer-Verlag,Heidelberg,Germany, 1998.

Peterson, D.W.,and D. L. Peterson,Growth responses ofmountain hemlock (Tsuga mertensiana) to interannualand interdecadal climate variability, Ecology, in review,2000.

Pinnix,W.,Climate and Puget Sound,in JISAO/SMA Year 3report,Climate Impacts Group,University of Washington,Seattle,Washington,1998.

Price, P.W.,The plant vigor hypothesis and herbivoreattack, Oikos, 62, 244-251,1991.

Quigley, R.W.,R.W. Haynes,and R.T.Graham,Integratedscientific assessment for ecosystem management in theinterior Columbia Basin, USDA Forest Service GeneralTechnical Report PNW-382, USDA Forest Service,Corvallis,Oregon,1996.

Quinn,T. P.,and D. J.Adams,Environmental changesaffecting the migratory timing of American shad andsockeye salmon, Ecology, 77, 1151-1162,1996.

Reeves, G. H., F. H.Everest,and T. E.Nickelson,Identification of physical habitats limiting the produc-tion of Coho salmon in western Oregon and Washington,USDA Forest Service General Technical Report PNW-245, US Forest Service,Corvallis,Oregon,1989.

Rochefort,R.M.,R. L.Little,A.Woodward, D. L. Peterson,Changes in subalpine tree distribution in western NorthAmerica: A review of climate and other factors, TheHolocene, 4, 89-100,1994.

Ryan,K.C.,Vegetation and wildland fire: Implications ofglobal climate change, Environment International, 17,169-178,1991.

National Marine Fisheries Service (NMFS),NorthwestRegion,West Coast Salmon and the Endangered SpeciesAct,(http://www.nwr.noaa.gov/1salmon/salmesa/index.htm),2000.

Neilson,R. P., G.A.King,and J. Lenihan,Modeling forestresponse to climatic change: The potential for largeemissions of carbon from dying forests,in CarbonBalance of the World’s Forested Ecosystems: Towardsa Global Assessment, edited by J. Kanninen,pp.150-162, Academy of Finland, Joensuu, Finland,1994.

Neilson,R. P.,and R. J. Drapek, Potentially complexbiosphere responses to transient global warming,Global Change Biology, 4, 101-117,1998.

Neilson R. P.,I.C.Prentice,B.Smith,T. G. F. Kittel,and D.Viner, Simulated changes in vegetation distributionunder global warming,in The Regional Impacts ofClimate Change: An Assessment of Vulnerability, edit-ed by R.T.Watson,M.C.Zinyowera,R.H.Moss,and D. J.Dokken,pp.439-456,Cambridge University Press,Cambridge,United Kingdom,1998.

National Research Council, Upstream, NationalAcademy Press,Washington,DC,1996.

Nijssen,B., D. P. Lettenmaier, X.Liang,S.W.Wetzel,and E.F.Wood,Streamflow simulation for continental-scaleriver basins, Journal of Water Resources Research, 33,4,pp 711ff, 1997.

Noss,R. F.,E.T. LaRoe,III,and J. M.Scott,Endangeredecosystems of the United States: A preliminary assess-ment of loss and degradation, National BiologicalService Biological Report 28, Washington,DC,58 pp.,1995.

Overpeck, J.T., D. Rind,and R.Goldberg,Climate-induced changes in forest disturbance and vegetation,Nature, 343, 51-53,1990.

Park,R.A., J. K.Lee,and D. J. Canning, Potential effectsof sea-level rise on Puget Sound wetlands, GeocartoInternational, 99-110,1993.

Parsons, G. L., G. Cassis,A.R.Moldenke, J. D. Lattin, N. H.Anderson, J. C.Miller, P. Hammond,and T. D. Schowalter,Invertebrates of the H. J.Andrews experimental forest,western Cascade Range,Oregon, Part V: An annotatedlist of insects and other arthropods, US Forest ServiceGeneral Technical Report PNW-GTR-290, PacificNorthwest Research Station, Portland,Oregon,168 pp.,1991.

278


Shipman,H.,Vertical land movements in coastalWashington: Implications for relative sea level changes,Washington Department of Ecology, Olympia,Washington,1989.

Sohngen,B.,R.Mendelsohn,and R.Neilson,PredictingCO2 emissions from forests during climatic change: Acomparison of natural and human response models,Ambio, 27, 509-513,1988.

Swetnam,T.W.,and J. L.Betancourt, Fire-SouthernOscillation relations in the southwestern United States,Science, 249, 1017-1020,1990.

Swetnam,T.W.,and J. L.Betancourt,Meso-scale distur-bance and ecological response to decadal climate vari-ability in the American southwest, Journal of Climate ,11, 3218-3247,1998.

Swetnam,T.W.,and A.M. Lynch,Multicentury regional-scale patterns of western spruce budworm outbreaks,Ecological Monographs, 63, 399-422,1993.

Terleckyj, N. E.,Analytic documentation of three alter-nate socioeconomic projections,1997—2050, NPAData Services, Washington,DC,1999a.

Terleckyj, N. E.,Development of three alternate nationalprojections scenarios,1997—2050,NPA Data Services,Washington,DC,1999b.

Tett,S.B., P.A.Stott,M.R.Allen,W. J. Ingram,and J. F. B.Mitchell,Causes of twentieth century temperaturechange, Nature, 399, 569-572,1999.

Thomson,A.R.,R. F. Sheperd, J.W. E.Harris,and R. J.Silversides,Relating weather to outbreaks of westernspruce budworm in British Columbia, CanadianEntomologist, 116, 375-381,1984.

Torn,M.S.,and Fried, J. S.,Predicting the impacts ofglobal warming on wildland fire, Climate Change, 21,257-274,1992.

Trenberth,K.E.,The definition of El Niño, Bulletin ofthe American Meteorological Society, 78, 2771-2777,1997.

Tubbs, D.W.,Landslides in Seattle, Washington Divisionof Geology and Earth Resources Information Circular52,Washington Department of Natural Resources,Olympia,Washington,1974.

US Census Bureau,Historical state population series,2000,at:http://www.census.gov/population/www/esti-mates/st_stts.html

US Department of Agriculture,Economic ResearchService,US State Fact Sheets,2000(http://www.ers.usda.gov/epubs/other/usfact/),2000.

VEMAP Members,Vegetation/ecosystem modeling andanalysis project: Comparing biogeography and biogeo-chemistry models in a continental-scale study of terres-trial ecosystem responses to climate change and CO2doubling, Global Biogeochemical Cycles, 4, 407-4371995.

Walker, R.V, K.W. Myers, N. D. Davis,K.Y.Aydin,K. D.Friddland,H.R.Carlson, G.W. Boehlert,S.Urawa,Y.Ueno,and G.Anma,Diurnal variation in the thermalenvironment experienced by salmonids in the NorthPacific as indicated by data storage tags. FisheriesOceanography, in review, 2000.

Waring,R.H.,and J. F. Franklin,Evergreen coniferousforests of the Pacific Northwest, Science, 204, 1380,1979.

Weaver, J. L., P. C. Paquet, and L. F.Ruggiero,Resilienceand conservation of large carnivores in the RockyMountains, Conservation Biology, 10, 964-976,1996.

Welch, D.W.,Y. Ishida,and K.Nagasawa,Thermal limitsand ocean migrations of sockeye salmon(Oncorhynchus nerka): Long-term consequences ofglobal warming, Canadian Journal of Fisheries andAquatic Sciences, 55, 937-948,1998.

West, N. E.,and M.A.Hassan,Recovery of sage-brush-grass vegetation following wildfire, Journal of RangeManagement, 38, 131-134,1985.

Zobel, D. B.,A.McKee, G. M.Hawk,and C.T. Dyrness,Relationships of environment to composition,struc-ture,and diversity of forest communities of the centralwestern cascades of Oregon, Ecological Monographs,46, 135,1976.

279


Ruby Leung, Pacific Northwest National LaboratoryNathan Mantua,University of WashingtonEdward Miles,University of WashingtonBen Noble,Battelle Memorial InstituteHossein Parandvash, Portland Bureau of Water WorksDavid W. Peterson,US Geological SurveyAmy Snover, University of WashingtonSean Willard,University of Washington

Comments by the following reviewers are gratefullyacknowledged:Ralph Alig,Robert L.Alverts,WilliamClark,Robert Emmett, Josh Foster, Steven Ghan,Michael Haske, John Innes,Linda Joyce,Kai Lee,L.Ruby Leung,Susan M.Marcus,Steve McNulty, RonNielson,Claudia Nierenberg,Michael Scott, FrancisZweirs;Coordinated comments by U.S. Departmentof Agriculture;Coordinated comments by USDepartment of Interior;Coordinated comments byUS Department of Energy. Remaining errors are thesole responsibility of the coordinating author.


ACKNOWLEDGMENTSMany of the materials of this chapter are based oncontributions from participants on and those work-ing with the

Pacific Northwest Workshop and Assessment TeamPhilip Mote*,University of WashingtonDouglas Canning,Department of Ecology, State of

WashingtonDavid Fluharty, University of WashingtonRobert Francis,University of WashingtonJerry Franklin,University of WashingtonAlan Hamlet,University of WashingtonBlair Henry,The Northwest Council on Climate

ChangeMarc Hershman,University of WashingtonKristyn Gray Ideker, Ross and AssociatesWilliam Keeton,University of WashingtonDennis Lettenmaier, University of Washington

280


283

CHAPTER 10

POTENTIAL CONSEQUENCES OF CLIMATEVARIABILITY AND CHANGE FOR ALASKAEdward A. Parson1,2 with contributions from:Lynne Carter3, Patricia Anderson4, Bronwen Wang5, Gunter Weller4


Chapter Summary



Ecological Context




Key Issues

Thawing of Permafrost and Melting of Sea Ice

Effects on Forest and Tundra Ecosystems

Marine Ecosystems and Fisheries

Subsistence Livelihoods

Additional Issues

Freshwater

Agriculture

Tourism


Literature Cited

Acknowledgments

1John F. Kennedy School of Government,Harvard University; 2Coordinating author for the National Assessment Synthesis

Team; 3Office of the U. S.Global Change Research Program,University Corporation for Atmospheric Research;4University of Alaska, Fairbanks,co-chair,Alaska Regional Assessment Team; 5U. S.Geological Survey,Anchorage,co-chair,

Alaska Regional Assessment Team.


• Alaska’s climate has warmed about 4°F since the1950s,7°F in the interior in winter, with much ofthis warming occurring in a sudden regime shiftaround 1977.

• Alaska’s warming is part of a larger Arctic trendcorroborated by many independent measure-ments of sea ice,glaciers,permafrost, vegetation,and snow cover.

• Most of the state has grown wetter, with a 30%average precipitation increase between 1968 and1990.

• The growing season has lengthened by about 14days.

• Dramatic reductions in sea ice and permafrosthave accompanied the recent warming.


• Models project continued strong warming inAlaska, reaching 1.5-5°F by 2030,and 5-18°F by2100,strongest in the interior and north and inwinter.

• Continued precipitation increases are projected,reaching 20-25% in the north and northwest,with areas of decrease along the south coast.

• Increased evaporation from warming is projectedto more than offset increased precipitation,mak-ing soils drier in most of the state.


Spanning an area nearly a fifth the size of the entirelower 48 states,Alaska includes extreme physical,climatic,and ecological diversity in its rainforests,mountain glaciers,boreal spruce forest,tundra,peat-lands,and meadows. Lightly populated and growingabout 1.5% per year,Alaska has the nation’s highestmedian household income,with an economy domi-nated by government and natural resources. In con-trast to other regions,the most severe environmen-tal stresses in Alaska at present are already climate-related. Recent warming has been accompanied byseveral decades of thawing in discontinuous per-mafrost,which is present in most of central andsouthern Alaska,causing increased ground subsi-dence,erosion,landslides,and disruption and dam-age to forests,buildings,and infrastructure. Sea iceoff the Alaskan coast is retreating (by 14% since1978) and thinning (by 40% since the 1960s),withwidespread effects on marine ecosystems,coastalclimate,human settlements,and subsistence activi-ties.

284


Key Findings• As much as the top 30 feet of discontinuous per-

mafrost is projected to thaw over the 21st centu-ry, causing increased ground subsidence,erosion,landslides,and disruption and damage to forests,buildings,and infrastructure.

• The melting of sea ice is projected to continue,with the Canadian climate model projecting acomplete loss of summer Arctic sea ice by 2100.Loss of sea ice allows larger storm surges todevelop,increasing the erosion and coastal inun-dation,and also threatens populations of marinemammals and polar bears that depend on ice,and the subsistence livelihoods that depend onthem.

• Recent warming has been accompanied byunprecedented increases in forest disturbances,including insect attacks. A sustained infestationof spruce bark beetles,which in the past havebeen limited by cold,has caused widespread treedeaths over 2.3 million acres on the KenaiPeninsula since 1992,the largest loss to insectsever recorded in North America.

• Increases in blow-downs in forests due to intensewindstorms,and in canopy breakage from theheavy snows typical of warm winters,haveincreased vulnerability of forests to insect attack.Projected further warming is likely to increasethe risk of insect attack.

• Significant increases in fire frequency and inten-sity, both related to summer warming,have

occurred. Simultaneously, the potential damagefrom forest fires has increased due to an increasein dispersed human settlement in forests. Theprojected further warming is likely to increasenear-term risk of fire.

• In the longer term,large-scale transformation oflandscapes is likely, including expansion of bore-al forest into the tundra zone,shifts of foresttypes due to fire and moisture stress,northwardexpansion of some commercially valuablespecies,and the appearance of significant firerisk in the coastal forest for the first time sinceobservations began.

• The Gulf of Alaska and Bering Sea supportmarine ecosystems of great diversity and produc-tivity, and the nation’s largest commercial fishery.The effect of projected climate change on theseecosystems could be large.

• Present climate change already poses drasticthreats to subsistence livelihoods,practicedmainly by Native communities,as many popula-tions of marine mammals, fish,and seabirds havebeen reduced or displaced due to retreat andthinning of sea ice and other changes. Projectedclimate changes are likely to intensify theseimpacts. In the longer term,projected ecosystemshifts are likely to displace or change theresources available for subsistence, requiringcommunities to change their practices or move.

285

Chapter 10 / Alaska


286

crude oil production (8% of US consumption) pro-vide a further 35% of the state’s incomes,while fish-eries provide an additional 7%. Other significantincome shares include tourism (5%),timber (2%),and mining (2%),with the remainder miscellaneous(agriculture contributes 0.1%) (Goldsmith,1997). Inaddition,diverse forms of subsistence livelihood arepracticed throughout the state,primarily but notexclusively by native communities. These activitiesdepend on fish,marine mammals,and wildlife -including partly commercial reindeer herding - andplay a social and cultural role vastly greater thantheir contribution to monetary incomes.

ECOLOGICAL CONTEXTAlaska’s ecosystem types encompass an extraordi-nary diversity, reflecting the state’s vastness andextreme variety of climates. Ecosystem typesinclude cool Sitka spruce and Western hemlock for-est in the southeast and south-central coastalregions;boreal forest of white and black sprucewith hardwoods on well-drained uplands throughthe south-central region and interior;Alpine tundraand meadows at higher elevations on interior moun-tain ranges;maritime tundra along the west coastfrom the Alaskan Peninsula and Aleutians to theSeward Peninsula,including vast coastal wetlands inthe Yukon-Kuskokwim delta;and Arctic tundra andbarrens on the northwest coast and north of theBrooks Range. Only about 30,000 acres is in agricul-tural production,principally in the Tanana andMatanuska valleys and on the Kenai Peninsula;largerareas are used for pasture (185,000 acres) and rein-deer grazing (12 million acres,mostly on the SewardPeninsula).

The Alaskan terrestrial landscape has been alteredless by direct human intervention than anywhereelse in the United States. The most significant cur-rent environmental pressures in Alaska includeheavy stress on fish stocks and marine ecosystemsfrom large commercial fisheries,both Alaskan andinternational;local impacts from the mining andpetroleum industries,including the aftermath of theExxon Valdez spill and its cleanup,as well as smallerongoing impacts from routine operations;and strain

POTENTIAL CONSEQUENCES OF CLIMATEVARIABILITY AND CHANGE FOR ALASKA

PHYSICAL SETTING ANDUNIQUE ATTRIBUTESAlaska spans 20 degrees of latitude and 42 of longi-tude,with a land area of 570,000 square miles -nearly a fifth of the lower 48 states - and a coastlineof more than 34,000 miles,longer than those of theother 49 states combined.This enormous expanseembraces extreme physical, climatic,and ecologicaldiversity. In the south,a series of mountain rangesparallel the coast,where intense precipitation pro-duces lush cool rainforests and large mountain gla-ciers. These southern ranges culminate in the longarc of the Alaska Range,which includes McKinley(Denali),the highest peak in North America.Beyond,in the interior, lie the wide valleys of theYukon River and its tributary the Tanana. Furthernorth,the Brooks Range divides the interior fromthe cold and arid Arctic slopes. Alaska containsabout 75% of US national parklands and 90% ofnational wildlife refuge lands (USGS BRD, 1999). Itcontains roughly 40% of the nation’s surface waterresources (Lamke,1986),63% of its wetlands (Hallet al.,1994),essentially all of its permafrost,andmore glaciers and active volcanoes than all otherstates combined.

SOCIOECONOMIC CONTEXTAlaska is lightly populated,with 614,000 people in1998 distributed between the small cities ofAnchorage (260,000), Fairbanks (84,000) and Juneau(30,000),a few smaller towns,and many villages andrural settlements. Average annual populationgrowth was more than 3% per year in the 1980sdeclining to about 1.5% per year in the 1990s,where it is projected to remain, giving a projectedtotal state population that reaches 885,000 by 2025(US Census Bureau,2000). Native peoples compriseabout 16% of Alaska’s population. The state’s medi-an household income,nearly $52,800 in 1996,is thehighest in the nation.1 The economy is dominatedby government and natural resources,with Federalcivilian and military payrolls,and the State’sPermanent Fund,contributing 44% of total incomes.The North Slope oil fields,which provide 19% of US

1 Median household money income,in 1996 dollars.US Census Bureau (2000).

Chapter 10 / Alaska

287


Alaska’s recent climate has shown a strong warmingtrend. General Arctic warming began in the mid-19th century, but has accelerated in the past fewdecades (Overpeck et al.,1997). Alaska has warmed4°F (2°C) since the 1950s on average,with thelargest about 7°F (4°C) in the interior in winter(Chapman and Walsh,1993;Weller et al.,1998).Local weather records show that the growing sea-son in Alaska lengthened by 13 days since 1950(Keyser et al.,2000). Much of the recent warming,occurred suddenly around 1977,coincident withthe most recent of the large-scale Arctic atmosphereand ocean regime shifts (Weller and Anderson,1998).

on fragile ecosystems and communi-ties from rapidly growing summertourism throughout the state. Inaddition,like most of the Arctic,Alaska is already experiencing muchmore rapid climate warming than thelower 48 states,with major ecologi-cal and socioeconomic impacts andthe early signs of climate-related land-scape change increasingly evident.With the exception of direct fishingpressure on marine ecosystems,thegreatest present environmental stress-es in Alaska are climate-related.Experimental manipulations andmodel studies both suggest that pres-ent and future climate change ishighly likely to profoundly alter therange,species mix and functioning ofAlaskan ecosystems.

CLIMATE VARIABILITY AND CHANGEAlaska encompasses extreme climatic differences.The southern coastal margin,including the panhan-dle and Aleutians,has a maritime climate with coolsummers, relatively mild winters,and heavy precipi-tation,up to 200 inches (500 cm) annually in partsof the southeast, forming large glaciers on the south-ern mountains. North of the Alaska Range the cli-mate is continental,with moderate summers (Julyaverage 59°F or 15°C), very cold winters (Januaryaverage -13°F or -25°C), rapid seasonal transitions,and annual precipitation of 8-16 inches (20-40 cm).The North Slope beyond the Brooks Range has anArctic semi-arid climate,with annual precipitationless than 8 inches (20 cm), average July temperaturearound 39°F (3°C),and snow on the ground ninemonths of the year. Permafrost is present inall of the state except a narrow belt alongthe southern coast.

Alaska’s climate shows significant interannual andinterdecadal variability, associated with large-scaleshifts in ocean temperature and salinity regimes,iceconditions,and marine ecosystems in the surround-ing seas (Proshutinsky and Johnson,1997;Groismanand Easterling 1994;Serreze et al.,1995b;Thompsonand Wallace,1998; Parker et al.,1994;Royer, 1993;Francis et al.,1998;Trenberth and Hurrell,1994;NRC,1996).

Figure 1: Major Ecological Regions of Alaska. Source: NationalAtlas of the United States. See Color Figure Appendix

Figure 2: Average temperatures in Alaska have increased over the 20th century, with about 4°F warming since the 1950s. Source: Historical ClimateNetwork, National Climate Data Center. See Color Figure Appendix.

Alaska: 20th Century Annual-average Temperature

Major Ecological Regions of Alaska

of glaciers, warming and thawing of permafrost,andretreat and thinning of sea ice (Echelmeyer et al.,1996;Sapiano et al.,1998;Lachenbruch andMarshall,1986;Osterkamp,1994;Osterkamp andRomanovsky, 1996;Wadhams 1990;Cavalieri et al.,1997;Serreze et al.,2000;Krabil et al.,1999;Dowdeswell et al.,2000). Paleoclimatic evidencesuggests the Arctic is now warmer than at any timein the past 400 years (Overpeck et al.,1997). Thestart of Arctic warming in the mid-19th century indi-cates a contribution from natural factors (Overpecket al.,1997). Of the stronger high-latitude warmingof the past three decades, roughly half can beexplained by changes in storm track patterns associ-ated with natural patterns of climate variability,although it is possible that anthropogenic changesin radiative forcing may be shifting these patterns sothey tend to favor high-latitude warming (Hurrell,1995,1996). The remaining share of recent high-lati-tude warming, roughly half, is broadly consistentwith model predictions of the consequences ofanthropogenic greenhouse forcing (Serreze et al.,2000). Observations of vegetation and snowcoverfrom satellites,and of the annual fluctuation ofatmospheric CO2 concentration,further corroboratethe broad warming trend over northern mid to highlatitudes. Mean annual snowcover of the NorthernHemisphere decreased 10% from 1972 to 1992(Groisman and Easterling,1994),while the growingseason over northern mid- to high latitudesincreased by 7 to 14 days (Myneni et al.,1997).


All climate models project the largest warming tooccur in the Arctic region,principally because of

Alaska has also grown substantially wetter over the20th century. The sparse historical record since 1900shows mixed precipitation trends,with increases ofup to 30% in the south,southeast and interior, andsmaller decreases in the northwest and over theBering Sea. The trend to higher precipitation hasbeen stronger recently, a 30% average increase overthe region west of 141 degrees longitude (i.e.,all ofAlaska except the panhandle) between 1968 and1990 (Groisman and Easterling,1994).

Alaska’s recent warming is part of a strong trendobserved throughout the circumpolar Arctic, exceptfor one large region of cooling over eastern Canadaand Greenland. This broad Arctic warming has beenaccompanied and corroborated by extensive melting

288


Figure 3: Over the 20th century, precipitation in Alaska hasincreased. Source: Historical Climate Network, National ClimateData Center. See Color Figure Appendix

Alaska: 20th Century Annual Total Precipitation

Figure 4: Precipitation and temperature change projected in the 21st century by two climate models. See Color Figure Appendix.

Temperature Change - 21st CenturyPrecipitation Change - 21st Century

Canadian Model Both modelsproject precipi-tation increasesof up to 25%.With warmertemperatures,these giveslightly reducedsoil moisture inmost of Alaska.

100%80%60%40%20%0-20%-40%-60%-80%-100%

Hadley Model

Canadian ModelBy 2100, bothmodels projectlarge increases inannual averagetemperature, withthe greatestwarming in theNorth and West.

Hadley Model +15ºF

+10ºF

+5ºF

0º

-5ºF

Chapter 10 / Alaska

289

combines reviewing projections of future impactswith describing impacts that are already occurring,which are likely to provide insights into the charac-ter of the larger projected future impacts.

ice-albedo feedback (Kattenberg et al.,1996). Iceand snow are more reflective than the land orwater they cover, so after melting,the exposed sur-face absorbs more solar radiation,accelerating fur-ther warming. In both the Canadian and Hadleyscenarios (Boer et al.,1984;1999a,b,McFarlane etal.,1992;Flato et al.,1999;Mitchell et al.,1995;Mitchell and Johns,1997; Johns et al.,1997), warm-ing in Alaska increases from the southeast to thenorthwest,and is strongest in winter. In theCanadian model,Alaskan warming ranges from 2 to5°F (1.1 to 2.8°C) by 2030,and from 7 to 18°F (4to 10°C) by 2100,accompanied by complete loss ofsummer Arctic sea ice. In the Hadley model, warm-ing is 1.5 to 3.6°F (0.8 to 2°C) by 2030,5 to 12°F(3 to 6.5°C) by 2100,with smaller but still exten-sive loss of sea ice. Comparing these projectedfuture changes to the 4°F (2°C) temperaturechange already experienced in the last fewdecades,they range from half as much again to adoubling by 2030,and from a doubling to a quadru-pling by 2100.

The Hadley and Canadian scenarios also both proj-ect that annual precipitation will increase in mostof Alaska,with the largest increases reaching 20-25% in the north and northwest,with some areas ofup to 10% decrease along the south coast. Themodels differ more strongly in projecting seasonalpatterns of precipitation changes,particularly inwinter. Winters are wetter in the Hadley scenarioexcept in the extreme west and the panhandle,while the Canadian scenario has drier winterseverywhere except the Seward Peninsula andnorthwest coast. Summers have more precipitationin both scenarios except for regions along thesouth coast. In the Hadley scenario,this region ofreduced summer precipitation is confined to theextreme southeast and part of the Alaska Peninsula,while in the Canadian scenario it covers a broadswath along the entire southern coast. Increasedevaporation due to warmer summer temperaturesexceeds projected precipitation increases,however,so both scenarios project soil moisture decreasingthroughout the state, except for an interior regioncentered on Fairbanks in the Hadley scenario.

KEY ISSUESThe climate changes underway in Alaska havealready had major impacts. This synthesis focuseson four key issues - thawing and melting of thecryosphere,particularly permafrost and sea ice; for-est and tundra ecosystems;marine ecosystems andfisheries;and subsistence livelihoods. The approach

Figure 5: The largest projected warming is in winter, when bothmodels show average daily-high temperatures increasing more than15ºF over the northern half of the state. Source: B.Felzer, UCAR.See Color Figure Appendix.

Canadian Model 21st Century

Hadley Model 21st Century

Winter Maximum Temperature Change

+15ºF

+10ºF

+5ºF

0º

-5ºF

Figure 6: The Hadley model projects increased summer soil mois-ture in central Alaska and decreases in the north and south, whilethe Canadian model projects moderate decreases throughout thestate. Source: B.Felzer, UCAR. See Color Figure Appendix

Canadian Model 21st Century

Hadley Model 21st Centur y

Summer Soil Moisture Change

>100

+75

+50

+25

0

-25

-50

-75

-100


290

Thawing and Melting:Thawing of permafrost, retreat and thinning of seaice,and reduction of the river and lake ice seasonare underway and are projected to continue. Thesechanges are likely to bring widespread changes inecosystems,increased erosion,harm to subsistencelivelihoods,and damage to buildings, roads,andother infrastructure (including sanitary systems). Inthe longer term,longer ice-free seasons are likely tobring substantial benefits to marine transport andoffshore operations in the petroleum industry, andwill likely have major implications for trade andnational defense.

Effects on Forests and Tundra Ecosystems:Recent warming appears to have brought increasedproductivity in the boreal forest zone,offset to anuncertain degree by increases in summer moisturestress, fire and insect outbreaks. Future warming islikely to continue increasing both productivity andstresses,and eventually bring large-scale landscapetransformation as boreal forest advances into pres-ent tundra and mixed forest into present boreal for-est. Changes in these ecosystems will possibly havelarge effects on the global carbon cycle.

Marine Ecosystems and Fisheries:Alaskan and Bering Sea marine ecosystems showstrong signals of climate-driven variation,althoughtheir mechanisms are not known. Further climatechange is likely to bring large changes in marineecosystems including stocks important for both thecommercial and subsistence catch,but knowledgeof their specific character is very limited.

Subsistence Livelihoods:Subsistence hunting and fishing have been signifi-cantly harmed by present climate changes,through

stresses on fish,marine mammals,and wildlife driv-en by present thawing,sea ice retreat,and ecosys-tem shifts. While some specific subsistenceresources are likely to grow more abundant (e.g.,salmon near the northern limit of their range),thesestresses are likely to grow more intense, even in thenear term.

1. Thawing of Permafrost andMelting of Sea Ice

Throughout Alaska,the landscape and human activi-ties are fundamentally affected by the presence ofice,snow, and permafrost. Because annual averagetemperatures in much of the southern portion ofthe state are near 32°F (0°C) – e.g.,28°F (-2°C) inFairbanks – a small warming can transform the land-scape through thawing of permafrost,melting ofice,and reduction of snow cover. The ecological,hydrological,economic,and social effects of thesechanges to the cryosphere2 will be large,and willprofoundly affect every other domain of impact con-sidered. All components of the cryosphere in theArctic are experiencing change,including snowcover, mountain and continental glaciers,per-mafrost,sea ice,and lake and river ice. For example,glaciers in Alaska,as throughout the Arctic,haveretreated through most of the 20th century.Estimated losses in Alaskan glaciers are of the orderof 30 feet (10 meters) in thickness over the past 40years,while some have gained thickness in theirupper regions,consistent with recent increases inboth temperature and precipitation (BESIS,1997).Melting of glaciers is contributing to rising sea lev-els worldwide (Meier, 1993),while melting ofAlaskan glaciers may have pronounced regionaleffects through the contribution of their runoff toocean currents and marine ecosystems in the Gulfof Alaska and Bering Sea,as discussed below. Thediscussion here concentrates on permafrost and seaice,whose impacts on people and ecosystems werejudged to be most direct and important. Generalwarming would also reduce the ice season on lakesand rivers,impairing transport on ice roads (Cole etal.,1999).

PermafrostPermafrost underlies about 85% of Alaska,the entirestate except for a narrow belt along the southerncoast. Its character varies widely, in depth,continu-ity, and ice content. In the interior and south of thestate most permafrost is discontinuous and relativelyshallow, reaching depths of 10 to 300 feet (3 to 100meters). From the Brooks Range north and alongthe northern and northwestern coasts,it becomesFigure 7: Permafrost regions of Alaska. Source: O.J. Ferrains,

1965. See Color Figure Appendix

Permafrost Regions of Alaska

2 The cryosphere consists of the frozen components of the Earth’s sur-face:ice,snow, and permafrost.

291

Chapter 10 / Alaska

(which alter heat transfer from the surface to theground) – have also contributed. Moreover, becausemuch of the observed thawing is associated withhuman disturbance of the surface,and because sys-tematic large-scale observations of changes in per-mafrost are lacking,the degree of contribution ofsurface warming to the observed thawing is not yetknown.

Continued climate warming is highly likely to bringaccelerated thawing of warm discontinuous per-mafrost. Over the 21st century the top 30 feet (10meters) is likely to thaw throughout much of theice-rich discontinuous permafrost zone,althoughcomplete thawing is likely to take centuries even indiscontinuous permafrost (Osterkamp andRomanovsky, 1999). Canadian studies have project-ed that even present surface temperatures willcause an eventual further retreat of the southern-most permafrost fringe in the Canadian subarctic by60 to 100 miles (100 to 160 km) (Dyke et al.,1997),with further retreat of 200 to 300 miles (300 to 500km) under doubled-CO2 equilibrium (Woo et al.,1992). The actual pattern of loss of permafrost will,of course,be more complex than a simple uniformretreat. Model studies using three transient climate-model scenarios have projected a 20 - 30% increasein depth of the active layer (Anisimov et al.,1997),and a 12 to 22% reduction in total Arctic permafrostarea by 2100 (Anisimov and Nelson,1997).

Thawing is likely to benefit some activities (e.g.,construction,transport,and agriculture) after it iscompleted,but the intervening transitional periodof decades or longer is likely to bring many disrup-

thicker and continuous, reaching depths of 2,200feet (670 meters) in some locations on the NorthSlope (Ferrians,1965;Osterkamp et al.,1985;Brownet al.,1997). Permafrost has profound effects onhydrology, erosion, vegetation,and human activities.It limits movement of ground water and the rootingdepth of plants. On slopes,it allows characteristicfluid-like movement of surface soil and deposits.Seasonal thawing over continuous permafrost cre-ates a saturated surface layer in which pools ofmeltwater accumulate,conducive to marsh and tun-dra ecosystems and peat formation. Building onpermafrost requires that structures be stabilized inpermanently frozen ground below the active layer,and that they limit their heat transfer to the ground,usually by elevating them on piles. For example,toprevent thawing of permafrost from transport ofheated oil in the Trans-Alaska pipeline,400 miles ofpipeline were constructed elevated on ther-mosyphons,at an additional cost of $800 million(Cole et al.,1999).

Permafrost in Alaska has been warming for morethan a century. Continuous permafrost on theNorth slope of Alaska has warmed 4-7°F (2-4°C)over the last century (Lachenbruch and Marshall,1986). Since temperatures at the upper surface ofcontinuous permafrost are still low, typically below23°F (-5°C),no significant loss of continuous per-mafrost is projected over the 21st century, althoughthickening of the active layer may cause active layerdetachment,local subsidence,damage to structures,and hydrological changes (Osterkamp andRomanovsky, 1996). The discontinuous permafrostto the south is warmer, usually above 28°F (-2°C).Here,Osterkamp (1994) reported recently increasedwarming at multiple sites,1 to 3°F (0.5 to 1.5°C)since the late 1980s,and inferred from this andother evidence that much of the discontinuous per-mafrost south of the Yukon River and on the southside of the Seward Peninsula must already be thaw-ing. Many reports of localized thawing and associat-ed surface disruptions support this inference(Osterkamp,et al.,1998; Jorgenson et al.,2000;Osterkamp et al.,2000). In the central CanadianArctic,a general northward retreat of the southern-most margin of discontinuous permafrost by about60 miles (100 km) over the 20th century has beenreported (Kwong and Gan,1994; French andEgorov, 1998). It is highly likely that recent climatechanges have contributed to permafrost warmingand thawing. The beginning of permafrost warmingpre-dates the recent sharp increase in surface tem-perature,however, suggesting that other factorsthan warming – such as natural long-term variabilityor changes in snow depth and vegetation cover

Figure 8: For much of its length, the Trans-Alaska Pipeline is elevat-ed on refrigerated pilings to prevent local thawing and groundinstability. Source: ©David Marusek.


292

tions and few benefits. Thawing of any permafrostincreases groundwater mobility, reduces soil bearingstrength,increases susceptibility to erosion andlandslides,and can affect soil storage of CO2, there-by increasing release if the thawed soil drains anddries,or increasing storage if the soil remains flood-ed. Warming greatly reduces permafrost’s bearingstrength even if it remains below the freezing point,e.g., by 70% for a pile in permafrost that warmsfrom 25 to 30°F (-4 to -1°C) (Nixon,1990;Cole etal.,1999). Where permafrost has a high ice content,typically in about half the area of discontinuous per-mafrost,thawing can induce severe,uneven subsi-dence of the surface,called thermokarst,observedin some cases to exceed 16 feet (5 m). Human-induced thawing of ice-rich discontinuous per-mafrost has already damaged houses, roads,airports,pipelines,and military installations; required costlyroad replacements and increased maintenanceexpenditures for pipelines and other infrastructure;and increased landscape erosion,slope instabilitiesand landslides.

Present costs of thaw-related damage to structuresand infrastructure in Alaska have been estimated atabout $35 million per year,3 of which repair of per-mafrost-damaged roads is the largest component(Cole et al.,1999). Longer seasonal thaw of theactive layer could disrupt petroleum explorationand extraction and increase associated environmen-tal damage in the tundra, by shortening the seasonfor minimal-impact operations on ice roads andpads. The near-term risk of disruption to operationsof the Trans-Alaska pipeline is judged to be small,although costly increases in maintenance due toincreased ground instability are likely. The pipeline’ssupport structures are designed for specific ranges

of ground temperatures,and are subject to heavingor collapse if the permafrost thaws. Replacingthem,if required, would cost about $2 million permile. Subsidence from thawing can also destroy thesubstrate of present ecosystems,destroying them ortransforming them to other types of ecosystems, forexample changing forests to grasslands or bogs(Jorgenson et al.,2000;Osterkamp et al.,2000).Where large-scale thawing of ground ice hasoccurred,the landscape has been transformedthrough mudslides, formation of flat-bottomed val-leys,and formation of melt ponds,which canenlarge for decades to centuries (Everett andFitzharris,1998,p.94).

Sea IceAs permafrost is a prominent feature of the Alaskanlandscape,sea ice is a prominent feature of itscoasts and the adjoining marine ecosystems.Present for six months along the Bering Sea coastand ten months along most of the Chukchi andBeaufort Seas,sea ice strongly influences coastal cli-mate,ecosystems,and human activities. The area ofArctic sea ice varies up to 50% seasonally, and alsoshows strong interannual variation. Large and statis-tically significant reductions in summer sea ice,which have been proposed as an early signal ofglobal climate change (Walsh,1991),are evident inrecent decades despite this variability (Cavalieri etal.,1997;Maslanik et al.,1996,1999;Wadhams,1997). Recent reports show area declines of about3% per decade since the late 1970s,with the largestdeclines 3.6% per decade in August and September(Serreze et al.,2000). The area of multi-year ice hasdeclined by 14% since 1978 (Johannesen et al.,1999). Model calculations indicate that recent sea-ice trends are consistent with the estimated effectsof present greenhouse warming,and are highlyunlikely to be accounted for by natural climate vari-ability (Vinnikov et al.,1999). Comparison of twosatellite records suggests that the rate of area lossincreased from 2.8% per decade in the 1980s to4.5% per decade in the 1990s (Johannesen et al.,1995). Record low values of summer ice extenthave been set repeatedly since 1980 (Chapman andWalsh,1993;Serreze et al.,1995a),while September1998 ice area in the Beaufort and Chukchi seas (thewestern part of the Arctic Basin) was 25% below theprior minimum value over the 45-year record(Maslanik et al.,1999).

Arctic sea ice has also grown thinner over the pastfew decades. Local observations of sea ice thinningby 3.3 to 6.5 feet (1 to 2 meters) have been report-ed for several years (Wadhams 1990,1997;McPheeet al.,1998),but the limited spatial and temporal

Figure 9: Houses near Fairbanks, Alaska use jacks for sup-port as permafrost thawing causes uneven settling. Source:©1999, Gary Braasch.

3 This sum represents about 1.4% of the State budget.

coverage of these measurements prevented drawingconclusions about Arctic-wide trends until recently.A recent analysis of submarine ice data,however, hasprovided the first persuasive evidence of large-scalethinning over the entire Arctic basin. MeanSeptember ice draft4 observed in six trans-Arctic sub-marine cruises from 1958 to 1976 was 10 feet (3.1meters),while mean draft in three similar cruisesbetween 1993 and 1997 was 5.9 feet (1.8 meters).In addition to the 4.1 feet (1.3 meters) of averagethinning between the two sets of cruises,the recentcruises also found continued thinning at a rate of 4inches per year (10 cm/year) from 1993 to 1997(Rothrock et al.,1999). Evidence of widespread sea-ice melting is corroborated by substantial recentincrease in freshwater content of the Arctic Ocean,from depth equivalent of 0.8 meters in 1975 and 2.4meters in 1997 (McPhee et al.,1998).

Sea-ice retreat allows larger storm surges to developin the increased open-water areas,increasing ero-sion,sedimentation,and the risk of inundation incoastal areas. Moreover, coastline where permafrosthas thawed is made more vulnerable,which in com-bination with increased wave action can causesevere erosion (Brown and Solomon,2000; Forbesand Taylor, 1994;Shaw et al.,1998;Weller, 1998;Wolfeet al.,1998.). Local coastal losses to erosion of theorder of 100 feet (40 meters) per year have beenobserved in some locations in both Siberia andCanada (Semiletov et al.,1999;Solomon and Covill,1995). Aerial photo comparison has revealed totalerosive losses up to 1,500 feet (600 meters) over thepast few decades along some stretches of the Alaskancoast (Weller and Anderson,1998). Several villageson Alaska’s west coast are sufficiently threatened byincreased erosion and inundation that they must beprotected or relocated. Present plans include con-structing a $4-6 million sea wall in Shishmaref (a 10-15 year interim solution),and relocating Kivalina onhigher ground at an estimated cost of $54 million(US Army Corps of Engineers,1998).

Under further climate change,further large reduc-tions in sea ice are projected,although there is sub-stantial variation in estimates of the magnitude andtiming. Analysis with one transient climate-modelscenario projected 60% loss of Arctic summer sea icearea by the time of CO2 doubling,accompanied byan increase in the duration of the open-water seasonfrom 60 to 150 days. The same climate scenario alsosuggests an increase in the offshore distance of theice pack from 90 to 125 miles (150 - 200 km) at pres-ent,to 300 - 500 miles (500 - 800 km) (Gordon and

293

Chapter 10 / Alaska

Figure 10: Coastal thawing and sea ice retreat have allowedextreme coastal erosion in both North America and Eurasia, withsome local losses of up to 100 feet per year. Source: IgorSemiletov, Pacific Oceanological Institute, Vladivostok.

Erosion on the Arctic Coast of Siberia

Figure 11: Canadian model projections of future Arctic sea-ice retreat.

Source: B. Felzer, UCAR, 2000. See Color Figure Appendix

Projected Summer Sea Ice ChangeCanadian Model: An Ice-free Arctic Summer

Current Sea Ice Extent

2030s Sea Ice Extent

2090s Sea Ice Extent

Both models proj-ect substantial fur-ther retreat of sea icethrough the 21st century,with complete loss of summerArctic sea ice in the Canadian model bythe 2090s. Sea ice outputs were notavailable for the Hadley scenario, but areconstruction based on sea-surfacetemperature shows a 40 to 50% loss ofsummer sea ice by the 2090s.

4 Ice draft is the depth of sea ice below the water line,equal to rough-ly 90% of total ice thickness.


294

Adaptation OptionsWhere sufficient information is available,vulnerabili-ty of structures to permafrost thawing can bereduced by careful site selection to avoid per-mafrost with high ice content and favor permafrostwith high gravel content. Unfortunately, local infor-mation on permafrost characteristics is oftenunavailable or inaccurate,and many siting and devel-opment decisions fail to consider the informationthat is available,or the likely future development ofthe site and its surroundings (Smith and Johnson,2000). When site or route modifications are notundertaken or not feasible,the effects of permafrostthawing on building and infrastructure can still bereduced,although at substantial cost and difficulty,through several approaches. Local contributions tothawing can be reduced by minimizing physical dis-turbance of the surface,and through insulation andheat transfer measures to reduce local thermal dis-

O’Farrell,1997). Both the Hadley and Canadianmodels project large reduction in summer sea ice by2100. The Canadian model projects the more rapidloss,as shown in Figure 11,with complete disap-pearance of summer sea ice by 2100.

Loss of sea ice threatens large-scale change inmarine ecosystems,threats to populations of marinemammals and polar bears that depend on the ice,and to the subsistence livelihoods that depend onthem. Further retreat may also bring some benefits,principally by facilitating water transport and oilexploration and extraction. Expanded transportpossibilities from greatly reduced Arctic sea iceextent and increased open-water season,includingthe possibility of routine summer navigationthrough both the Northeast and Northwest Passages(North of the Eurasian and North American conti-nents),are likely to have major implications for bothtrade and national security.

Betting on Spring Breakup: The Nenana Ice Classic

The town of Nenana is located about 65 miles southeast of Fairbanks on the Tanana river, a major tributary ofthe Yukon. In 1917,when Nenana was a construction base for the Alaska Railroad, railroad workers ran a bet-ting pool on when the river ice would break up in the spring. Sufficiently popular to be repeated in subse-quent years,the pool became a local tradition that now has been repeated every year for 84 years. Entry tick-ets cost two dollars,and represent a bet on a single one-minute interval. The jackpot,$800 in 1917, was morethan $330,000 in 2000. The high stakes and long continuous history of this contest make it a unique localrecord of Alaska’s 20th century climate history.

The same procedure is used to define the moment of breakup each year. In early March,a large log structureis frozen into the ice about 300 feet from shore,and later joined by a cable to a watchtower on the shore. A

strong enough pull on the cable,which occurswhen the ice has shifted enough to move the struc-ture about 100 feet downriver, stops the clock.

Over the contest’s history the earliest breakup hasbeen on April 20 (in 1940 and 1998),the latest onMay 20 (in 1964). Although breakup dates varygreatly from year to year, the past few decades haveseen a strong trend toward earlier breakup.Removing some of the year-to-year variation by cal-culating an 11-year average (for each year, the aver-age of the eleven breakup dates from five yearsbefore to five years after) reveals an advance ofeight days between the 1920s and the 1990s,fromMay 7 to April 29. Nenana is a major shipping cen-ter for summer barge traffic,so the earlier breakupbrings significant local economic benefits. It is alsoa concrete local indicator of the strong warmingtrend that has occurred across Alaska over the pastfew decades.

Spring Breakup Dates in the Nenana Classic (11-year moving average)

Figure 12: The average date of spring breakup of ice on the TananaRiver at Nenana has advanced by eight days between the 1920s andthe 1990s. Source: Historical data from Nenana Ice Classic,http://www.ptialaska.net/~tripod/breakup.times.html.

turbance. Piles used to support structures can besunk deeper in the permafrost or refrigerated,tomaintain their bearing strength longer as the per-mafrost warms and active layer thickens. Withenough advance planning,local thawing can beactively induced before construction, by strippingvegetation and surface soil from the site five yearsor more in advance (Osterkamp et al.,1998). Whichof these types of measures is most promising willdepend on site characteristics,the type of project,and its intended lifetime. For roads and runways,the consequences of thawing can be reduced bybuilding with gravel rather than paved surfaces,asthey can be more readily repaired after subsidence.Coastal settlements threatened by increased stormsurge or erosion can be protected with sea walls orother fortification,or relocated further inland. Noadaptation options are likely to be available for ter-restrial ecosystems threatened by permafrost thaw-ing,or marine ecosystems threatened by sea-iceretreat.

2. Effects on Forest and TundraEcosystems

Forests cover 129 million acres of Alaska, about onethird of the state (Powell et al.,1993). Various formsof tundra cover another third,in mountainous andcoastal regions and north and west of the BrooksRange. Of the forested land, about 10% is temperatecoastal rainforest,the remainder interior bore-al forest. About 21 million acres,or 16% oftotal forest,is classified as productive,capableof average growth of 20 cubic feet per acreper year. About 4 million acres,nearly all of itin the coastal forest,is outside protected areasand has the productivity of 50 cubic feet/acre-year necessary to support commercial harvestwith road construction (Berman et al.,1999).The state’s timber harvest increased from 600to 1,100 million board feet from 1986 to1990,and has since declined to about 500 mil-lion board feet. Employment and income inthe industry followed the same pattern,peak-ing at 4,000 jobs and $200 million in 1990,declining to 2,500 jobs and $130 million by1997. The decline of the 1990s principallyreflects two economic causes:the closure oftwo pulp mills in southeast Alaska,and thedepletion of Native Corporation timber inven-tories,which were exported in large quanti-ties as round logs to convert assets to cashduring a period of high world prices (Berman et al.,1999). In addition to their commercial value,Alaskan forests provide various ecosystem services

295

Chapter 10 / Alaska

and support subsistence livelihoods and recreationactivities.

Recent warming in Alaska has increased averagegrowing degree-days by about 20% over the state,bringing apparent increases in forest productivityon sites that are not moisture limited – principallyin the southern coastal forest,but also includingsome regions of the boreal zone (Ciais et al.,1995;Myneni et al.,1997). At the northern margin of theboreal forest,the present climate already favors for-est expansion into the tundra zone,particularly onthe Seward Peninsula,with the potential for suchexpansion estimated as 35 miles per °F of climatewarming (100 km per °C) (Weller and Lange,1999).On sites that are moisture-limited,which occurthrough much of the interior, recent warming hasapparently increased moisture stress and reducedproductivity (Barber et al.,2000). Near Fairbanksthe average number of days exceeding 80°F (27°C)annually has tripled since 1950,imposing moisturestress on white spruce stands that can be observedin clear negative correlation of productivity withwarm,dry summers (Juday et al.,1998, Fig.3.13). Ithas been suggested that the past 20 years have seenthe greatest moisture stress and lowest productivityof the 20th century through much of the interiorboreal forest (Juday and Barry, 1996) .

Figure 13: Since 1992, the largest outbreak of forest insects everrecorded in North America has caused widespread tree mortalityover 2.3 million acres. Source: USDA Forest Service. See ColorFigure Appendix

The 1990s Outbreak of Spruce Bark Beetles on theKenai Peninsula

contributed to the extensive decline of YellowCedar in the coastal forests,due to freezing of theirshallow root systems during winter cold spells withno insulating snow cover (Hennon and Shaw, 1997).Over the same period,the southern coastal forestshave also seen a marked increase in the frequencyof gale-force winds,which are the primary distur-bance agent in these forests (Veblen and Alaback,1996),and outbreaks of the defoliating westernblack-headed budworm that appear to be triggeredby warm dry summers (Holsten et al.,1985;Furnissand Carolyn,1977).

Forest fire frequency and intensity have alsoincreased markedly since 1970. As Figure 14 shows,the 10-year average of boreal forest burned in NorthAmerica,after several decades of around 2.5 millionacres (1 million hectares),has increased steadilysince 1970 to more than 7 million acres (3 M ha).Boreal forest fire reached extreme values in bothEurasia and North America in 1998,with over 27million acres burned (11 M ha) in total,10 million(4.7 M ha) in North America (Kasischke et al.,1999).Analysis of historical Canadian fire data shows astrong association between area burned and anom-alous patterns of mid-tropospheric circulation thattend to bring extended warm dry periods (Skinneret al.,1999).

A major change in Alaskan settlement geographysince 1970,promoted by policies including large-scale private transfer of public lands and extensiveroad-building,has greatly increased dispersed settle-ment in forest land. At the same time,other policiesto transform native villages into permanent commu-nities created more than 60 communities with sig-nificant costly infrastructure surrounded by borealforest (Leask,1985). These trends have greatlyincreased the vulnerability of people and settle-ments to forest fires. A single major fire in June1996, for example,burned 37,000 acres of forestand peat,causing $80 million in direct losses anddestroying 450 structures including 200 homes. Asmany as 200,000 Alaskan residents may now be atrisk from such fires,with the number increasing fur-ther as outlying suburban development continues toexpand (Nash and Duffy, 1997).

Continued increases in CO2 concentrations and pro-jected further climate warming are likely to bringcontinuing increases in forest productivity (Keyseret al.,2000),although these are likely to be limitedby accompanying increases in summer moisturedeficit, fire,and insect outbreaks (Fleming andVolney, 1995;Fleming and Candau,1997;Hogg andSchwarz,1997;Hogg,1999;Oechel et al.,1997b;

Substantial changes in patterns of forest distur-bance,including insect outbreaks, blowdown,andfire,have also been observed in both the boreal andsoutheast coastal forest. Although systematic large-scale observations have not been made,localizedobservations appear to support the hypothesis thatthese changes are climate-driven. A sustained out-break of spruce bark beetles since 1992 has causedover 2.3 million acres of tree mortality on the KenaiPeninsula,the largest loss from a single outbreakdocumented in the history of North America(Werner, 1996). The association of warmer tempera-tures with both accelerated beetle developmenttimes and increased tree vulnerability through mois-ture stress makes it likely that recent warming con-tributed to the outbreak (Juday et al.,1998).Outbreaks of defoliating insects in the boreal forest,including spruce budworm,coneworm,and larchsawfly, have also increased sharply in the 1990s,affecting a cumulative total of 800,000 acres(Holsten and Burnside,1997). Susceptibility of inte-rior forests to insect attack may also have increaseddue to canopy breakage from the heavy snow loadstypical of warmer winters.

In Southeast forests, warmer winters since the1970s with more precipitation falling as rain havereduced the frequency of low and moderate-eleva-tion avalanches,allowing mountain hemlock to colo-nize alpine tundra (Veblen and Alaback,1996).Reduced low-elevation snowpack has also likely

296


Annual Area of Northern Boreal Forest Burned inNorth America

Figure 14: The Alaskan boreal forest is a small part of an enor-mous forest that extends continuously across the northern part ofNorth America. The average area of this forest burned annually hasmore than doubled since 1970. Source: Kasischke andStocks,2000. See Color Figure Appendix

Sieben et al.,1997;Volney, 1996;Kasischke andStocks,2000). Increased fire risk is likely, even inthe near term. One study projecting fire risk underdoubled-CO2 equilibrium scenarios found a largeincrease in the area facing extreme fire risk inCanada and Siberia, very similar in both size andspatial pattern under four different climate models(Stocks et al.,1998). Substantial climate-relatedchanges have also been conjectured for the coastalforest over several decades,including the appear-ance of new fungi and the appearance of significantfire risk for the first time in the observed record(Juday et al.,1998).

Over the longer term, climate change is likely tobring large landscape-level vegetation changes toboth forest and tundra regions. Experimental stud-ies in both boreal forest and tundra have shown thatwarming increases nitrogen availability (Van Cleveet al.,1990;Lukewille and Wright,1997). At one tun-dra site,a decade of experimental 6°F (3°C) warm-ing brought major reorganization of the speciesmix,principally due to increased nutrient availabili-ty through changes in nitrogen mineralization(Chapin et al.,1995). Shrubs increased in domi -nance,while mosses, forbs and lichens werereduced or eliminated. Because shrubs transpire butthe declining species do not,such a reorganizationwould increase evapotranspiration,with largeimpacts on surface water budgets at many sites like-ly, including reduced pond formation and runoff anddrying of wetlands (Rouse et al.,1997). Moreover,the declining species include some that are criticalfor lactation and winter nutrition of caribou.Although there exist different views of how sensi-tive caribou are to such climate-driven changes,it ispossible that they could greatly reduce herds,withserious consequences for native communities thatdepend on them (Gunn,1995;Callaghan et al.,1998).

Equilibrium studies using biogeochemistry and bio-geography models have projected large increases invegetation carbon under climate change in bothboreal and tundra ecosystems (McGuire et al.,1995;McGuire and Hobbie,1997). These equilibrium stud-ies exclude dynamics of ecosystem response andprovide very limited treatment of disturbance,how-ever, and may consequently either over or under-estimate the effects of climate change on ecosys-tems.

Boreal forests and tundra ecosystems also containlarge stores of carbon in their soils. Worldwide esti-mates of carbon content are about 50 gigatonnes(GtC) in tundra soils,and 200 - 500 GtC in boreal

soils (McGuire et al.,1995;Melillo et al.,1995;McGuire and Hobbie,1997;Oechel et al.,1993; Postet al.,1982;Robinson and Moore,1999). These soilscan act as either sources or sinks of greenhousegases,depending on temperature and moisture con-ditions. As temperatures rise and soils thaw and dry,they become more susceptible to oxidation andrelease of CO2. Where drainage is poor and the soilremains wet after thawing,emissions of methane

may increase (Gorham,1995;Rivkin,1998). Studiesof both boreal forest and tundra have observed sig-nificant increases in soil carbon release followingseasonal warming and thawing (Goulden et al.,1998). Growing-season observations of specific tun-dra sites have found them to operate as net CO2sinks in the cool, wet 1970s,net sources in thewarmer, drier 1980s,and net sinks in the warm, wet-ter 1990s (Oechel et al.,1993,1995;Vourlitis andOechel,1999). The seasonality and mechanism ofcarbon storage and release in tundra ecosystemshave been called into question,however, by recentevidence that carbon release in winter may predom-inate (Oechel et al.,1997a).

297

Chapter 10 / Alaska

Figure 15: Under the Hadley scenario, the MAPSS biogeographymodel projects large-scale loss of tundra and taiga ecosystems asforests expand north and west. Likely consequences include dis-ruption of wildlife migration and associated subsistence liveli-hoods, as well as the potential for large releases of soil carbon.Source: R. Neilson et al., 1998. See Color Figure Appendix

Simulated Vegetation Distribution

Current

Hadley Model 2090s

Tundra

Taiga / Tundra

Boreal Conifer Forest

Temperate Evergreen Forest

Temperate Mixed Forest

Tropical Broadleaf Forest

Savanna / Woodland

Shrub / Woodland

Grassland

Arid Lands

tion districts in high-risk areas supported by specialproperty taxes; requiring risk-adjusted assessment offire insurance rates;or encouraging rural residents atrisk to form volunteer fire and emergency-responsecooperatives at their own expense. This approachwould represent a radical departure from historicalpolicies. A related strategy might reverse presentpolicies that encourage dispersed development, byproviding infrastructure only in present or designat-ed densely settled areas.

For the projected larger-scale ecological and land-scape transformations,no adaptation strategies arelikely to be available.

3. Marine Ecosystems andFisheries

The Gulf of Alaska and Bering Sea support marineecosystems of great diversity and productivity. TheBering Sea supports at least 450 species of fish,crus-taceans,and mollusks,and 25 species of marinemammals. The population of seabirds in Alaska isthe largest and most diverse of any similar-sizedregion in the Northern Hemisphere,with 66 speciespresent at some time of the year, and 38 species -over 50 million individuals - that breed there (Piattand Anderson,1996;Meehan et al.,1999).

Roughly 25 species of fish,crustaceans,and mol-lusks are commercially exploited in the Alaskan fish-ery. In 1995,Alaska’s fisheries landed 2.1 milliontons with an ex-vessel value (the amount paid tofisherman) of $1.45 billion, representing 54% of thelandings and 37% of the value of all US fisheries. Ofthis total,pollock were the largest share of volume(1.3 million tons,$297 million) while salmon werethe largest share of value (497,000 tons,$490 mil-lion,of which sockeye contributed 175,000 tons for$321 million,and pink 218,000 tons for $80 mil-lion). A notable contributor to the value of the fish-ery is the Tanner crab or Opilio, for which the vol-ume harvested was only 37,000 tons but the ex-ves-sel value was $175 million,more than any speciesexcept sockeye salmon and pollock. The Alaskanfishery employs about 20,000 people in harvestingand processing (Knapp et al.,1999,Table 1).

The Bering Sea and Gulf of Alaska have shownmarked fluctuations in their physical and ecologicalcharacteristics over time. Observed ecological fluc-tuations have included large-scale shifts in the abun-dance and distribution of many important fish,inver-tebrates,and marine mammals. Many ecosystemcomponents show clear association with interannu-al and interdecadal climate variability, with the influ -

In addition to increased carbon storage and changesin nutrient cycling,biogeography models consistent-ly project large-scale transformation of Arctic land-scapes,in which the northern edge of the borealforest advances into the tundra (Melillo et al.,1996;Everett and Fitzharris,1998). In Alaska,northwardforest advance is likely to be constrained by theBrooks Range,but substantial westward expansionon the Seward Peninsula is possible,as occurredduring a warm climatic period 6,000 years ago(Chapin and Starfield,1997; Foley et al.,1994).Tundra,constrained by the coastline,is likely toboth change in composition and shrink in area, byas much as two thirds, worldwide (Neilson et al.,1998;Everett and Fitzharris,1998).

In southern Alaska,temperate coniferous and mixedforests are likely to advance into the boreal zone.One model study found that an 8°F (4.5°C) warmingimposed on the Seward Peninsula induced two land-scape transformations over a century, from tundra toboreal spruce forest and subsequently - principallybecause of fire - to a mixed,deciduous-dominatedforest (Rupp et al.,2000). Over one to two cen-turies,other possible landscape changes includeexpansion of the coastal forest westward on theAlaska Peninsula (Chapin and Starfield,1997);anexpansion of forests to higher elevations,includingcolonization of some formerly glaciated lands;and ashift of interior regions with the greatest precipita-tion deficit to Aspen parkland.

Adaptation OptionsProjected increases in forest productivity, includingthe possibility of northward expansion of commer-cially valuable species, would likely bring commer-cial benefits if not offset by increased moisturestress, fire,and insect outbreaks. Various adaptationmeasures could help to offset climate-inducedincreases in fire risk in commercially valuableforests or near settlements. These might includeexpanded road networks to increase fire-suppres-sion capability and facilitate salvage and sanitationlogging - assuming that subsequent increase in set-tlement in forested areas can be discouraged;peri-odic controlled burns around settled areas to createbuffers;and increased investment and staffing in firesuppression. Any strategy based on expanded firesuppression,however, will carry its own ecologicalcosts and also risks being ineffective in the longterm,because by removing risk from property own-ers it would sustain incentives to build in fire-proneareas.

An alternative approach would create incentives toreduce private risk,e.g., by creating rural fire-protec-

298


ence of interdecadal variability apparently strongerthan that of interannual variability (Royer, 1982;Parker et al.,1994;NRC,1996; Francis et al.,1998;Brodeur et al.,1996). While there is climate-drivenvariability at many different time scales,a few inter-decadal climate variations during the 20th centuryhave apparently caused rapid and extreme shifts inthe organization of these marine ecosystems,mostrecently in 1977. Previous shifts occurred in 1924and 1946 (NRC,1996,p.197),and some data sug-gest another may have occurred in the mid-1990s(Mantua et al.,1997;NOAA,1999).

The regime shift of 1977 brought warmer sea-sur-face temperatures and a sharp reduction in sea icein the Bering Sea (NRC,1996;BESIS,1997). Salmonruns soared,and have largely remained high sincethen. The catches of 1997 and 1998 were of aver-age volume but dropped sharply in value,principallybecause of large declines in the lucrative Bristol Baysockeye run, roughly offset in volume but not invalue by huge runs of Pink salmon (Kruse,1998).Groundfish species including pollock, Pacific cod,Arrowtooth flounder and Yellowfin sole dropped tolow levels in 1977 and 1978,then began a sustainedclimb to record levels in the mid 1980s,after whichthey have stabilized or declined (Witherell,1999).Greenland turbot,the groundfish in the region mostadapted to a cold climate,declined (NRC 1996, Fig4.18). As pollock and other predators increased,sev-eral species of forage fish with high nutritionalvalue,such as capelin and herring,declined sharply.Various marine mammals and seabirds that fed onthese species changed their diets to other less fattyspecies,and have in turn declined sharply (NRC1996, Fig.4.27).

Populations of many species of seabirds,includingkittiwakes,murres,cormorants,larus gulls,guille-mots,puffins,and murrelets,have declined by 50 to90% since the 1970s (NRC 1996,pp.118-120).Marine mammals show similar signs of food stress.In the Gulf of Alaska,both Stellar Sea Lions andHarbor Seals have declined by more than 80%. Theextreme decline of Stellar Sea Lions has promptedsignificant restrictions on the pollock fishery since1998,to increase the sea lions’ food supply.Northern Fur Seals declined by about 35% from1970 to 1986,then rebounded somewhat through1990. Sea otters have declined as much as 80%since 1990 over much of the west coast,but thisdecline has been attributed to predation rather thanfood shortage. Estes et al.(1998) suggest that a fewOrca whales,perhaps only a single pod,couldaccount for the observed decline if they began toprey on sea otters following a decline in their usual

prey.

While climatic effects on larger-bodied,longer-livedspecies such as marine mammals through changesin food supply are most pronounced at longer time-scales,shorter-term changes can affect them inother ways,such as changes in the extent of sea icethat provides habitat for some species and excludesothers (Fay, 1974). For example,the light ice year of1979 brought unusually large overlap in the distribu-tions of seals and walruses,and saw a high rate ofwalrus predation on seals (Lowry and Fay, 1984).Large changes in numbers and location of othercommercially important fish species have beenreported in other Arctic waters,in response to cli-mate regional warmings or coolings of the order of2°F (1°C) (Buch et al.,1994;Vilhjalmsson,1997).

These ecosystem changes reflect the joint effects ofdecadal-scale climate fluctuations and human har-vesting of fish and marine mammals,but the com-plex time and space scales of both climate variationand human pressure prevents separating the contri-butions of each (NRC,1996). Moreover, the path-ways of climatic influence on these systems are notknown. They likely reflect combined effects ofchanges in streamflow, and the nutrient content,temperature,and vertical stability of coastal waters.

For the Bering Sea and Gulf of Alaska,one likelyinfluence involves changes in the Alaskan coastalcurrent. The intense storm systems generated by

299

Chapter 10 / Alaska

Figure 16: The climatic regime shift of the late 1970s caused large-scale reorganization of the Bering Sea ecosystem. Source: simpli-fied from NRC (1996). See Color Figure Appendix

State of Bering Sea Ecosystem

have also tended to over-capitalize,leading tointense competition and rent-seeking when catchesdecline (Marasco and Arom,1991;NMFS,1996,p.34). The effect of fluctuating catch on fishery rev-enue depends on the elasticity of demand. ThePacific halibut and salmon fisheries have been esti-mated to operate in the inelastic region of theirdemand curves:increases in catch reduce prices somuch that revenue falls,as when much of therecord 1991 Pink Salmon catch was dumped at sea(Knapp et al.,1999,p.12). Demand for pollock ispresently estimated to be elastic,but would becomeinelastic with modest increases in catch (Criddle etal.,1998).

In the face of such extreme uncertainty about thedirection and magnitude of future climate effects onfisheries,the most useful adaptation options will bemeasures that increase the robustness of humanactivities and communities to shifts in the locationand abundance of dif ferent species. The present sys-tem is quite vulnerable to climate change,becausespecialization of capital and the regulatory structurelimit its robustness. Many communities specializestrongly in one or a few species (e.g.,Bristol Bay ishighly dependent on sockeye salmon,while DutchHarbor is highly dependent on pollock and crab).In extreme cases like the Bristol Bay salmon fleet,equipment is so specialized that it is only useful forone fishery in one location. Regulatory measuresthat favor Alaskan shore-based processors over off-shore processing provide jobs and secondary eco-nomic benefits in Alaska,but reduce the ability of afishery to respond efficiently to climate-driven shiftsin the distribution of stocks (Huppert,1991). Oneimportant aspect of the present regulatory systemthat does promote robustness is the use of a limited-entry program. The ability of this program torespond to stock fluctuations could be furtherimproved by allowing buyback of quotas,or bydenominating allocations in terms of shares of avariable total harvest, rather than in terms of specif-ic quantities of catch.

4. Subsistence Livelihoods

Subsistence makes an important contribution tolivelihoods in many isolated rural communities inAlaska,especially but not exclusively for native peo-ples. While subsistence is practiced to gather food,subsistence resources and the activities associatedwith their harvest also make important contribu-tions to health,culture,and identity (Callaway et al.,1999;Berkes,2000;Wenzel,1995).

the Aleutian Low drop as much as 30 feet (10meters) of snow annually on the coastal mountainsthat ring the Gulf of Alaska, forming a large glaciersystem. The runoff from these glaciers is roughly800,000 cubic feet per second,comparable to themean annual discharge from the Mississippi,andcontributes more than 40% of the freshwater inputto the northeast Pacific. This runoff forms the swiftnarrow Alaska coastal current,whose low-salinitywaters flow westward along the coast and throughthe Aleutian passes into the Bering Sea (Royer, 1981;Royer, 1982). Significant future changes in thesecoastal storm systems,which may be associatedwith either climate variability or anthropogenic cli-mate change,could consequently cause largechanges in the temperature,salinity and nutrientcontent of these waters,and hence in the organiza-tion of their ecosystems.

While it is possible that the responses of theseecosystems to future climate change will be large,their specific character is highly uncertain. Effectsof stream warming on salmon can be projected withmore confidence than any oceanic ef fects on anyspecies:salmon are likely to benefit in the northernend of their range and be harmed in the south(Berman et al.,p.15). One preliminary study con-jectured that 21st century climate change couldincrease or decrease particular Alaskan fisheries byas much as a factor of two (Knapp et al.,1999).

Adaptation OptionsAny substantial change in the abundance, age-classdistribution,or location of a commercially exploitedspecies can bring large socioeconomic effects.Fisheries have tended to develop on stocks that areabundant,as the Alaskan pollock fishery has grownfrom minimal levels over the past 30 years. They

300


Figure 17: Diverse subsistence livelihoods based on fish, marinemammals and other wildlife, are practiced throughout Alaska.Source: D. Schmitz, National Park Service archive photo

Alaska’s 117,000 rural residents are entitled to prac-tice subsistence hunting and fishing on state, feder-al,and private lands and waters,while urban resi-dents also quality for subsistence activities on stateand private lands. The subsistence harvest by ruralresidents is about 43 million pounds of food annual-ly (20 million kg),or about 375 pounds (170 kg) perrural resident. The subsistence harvest is largest inthe most remote communities, about 500 - 800pounds (225 - 350 kg) per person annually. Fishcomprise 60% of total subsistence food,but there issubstantial inter-regional variation: west coast com-munities rely predominantly on fish,interior oneson fish and land mammals,and northern communi-ties principally on marine mammals (Wolfe andBosworth,1994).

The links between subsistence harvest and commer-cial activity are complex. If subsistence food werenot available,communities would have to substitutepurchased food. With an assumed cost of $3 - 5 perpound of purchased food,a study of four rural com-munities with large wild food harvests (590 to 760pounds,or 270 to 350 kg,per person) found thatthe cost of replacing the wild food harvest wouldbe $1,800 to $3,800 per person,or 13% to 77% ofcommunity per capita income (Callaway et al.,1999,p.70). Moreover, practicing subsistence requirescash income to buy the required equipment,such asguns,boats,and snowmobiles. In one surveyedcommunity (Unalakleet in 1982),the cost of practic-ing subsistence was about $10,000,nearly half ofmean household income (Callaway et al.,1999,p.65). Consequently, particularly for fishing in coastalcommunities,the subsistence and commercial har-vests may be closely linked:profits from the com-mercial catch may help pay for required subsistenceequipment,and subsistence fish may also be takenduring commercial fishing.

Many aspects of the climate change already occur-ring,and its consequences for forests,marineecosystems,permafrost,and sea ice discussed above,are already causing multiple serious harms to subsis-tence livelihoods. Many populations of marinemammals,wildlife,and seabirds have been reducedor displaced. Reduced snow cover, a shorter riverice season,and thawing of permafrost all obstructtravel to harvest wild food. Declines in some fishstocks have harmed subsistence as well as commer-cial harvesters (Weller and Lange,1999;Mulvaney,1998).

The most extreme effects of recent changes on sub-sistence livelihoods have been from changes in seaice,which have obstructed hunting of marine mam-

mals. The ice is further from shore,thinner, andpresent for less of the year. These factors,and therougher seas encountered in the larger open-waterareas between the shore and the ice,have madehunting more difficult,more dangerous,and lessproductive. Retreat of the ice is also likely to direct-ly harm some species on which subsistence huntersrely, including bearded seals and walrus. Walrus areparticularly at risk because they need ice strongenough to hold their weight over water shallowenough that they can reach the bottom to feed(Callaway et al.,1999). Polar bears need sea iceto hunt seals,and recent ice reductions havebeen associated with declining health and birthrate. Projected further large reductions in iceduration and extent are likely to threaten themwith extinction (Stirling et al.,1999).

Some subsistence harvests,such as salmonstocks near the northern end of their range,arelikely to benefit from projected climate change.Still,most projected near-term climate changesare likely to intensify existing unfavorableimpacts through further loss of sea ice, river ice,and permafrost. In addition,shifts in the compo-sition of tundra vegetation may decrease nutri-tion available for caribou and reindeer, and inva-sion of tundra by boreal or mixed forest is likelyto curtail the range of caribou and musk-ox (Gunn,1995). As changes in the cryosphere and both ter-restrial and marine ecosystems continue,continuinglarge changes or displacements of the resourcesavailable for subsistence are likely, requiring subsis-tence communities to make major changes in theirpractices,or move.

Adaptation OptionsSubsistence cultures have historically exhibited sub-stantial adaptability to year-to-year fluctuations inabundance of different species by shifting practicesand target species,which likely implies some abilityto adapt to effects of near-term climate change(Sabo,1991). Subsistence practices are now bothextensively regulated and hotly contested,however,posing challenges to traditional means of adapta-tion. Moreover, for many subsistence-dependentcommunities,particularly northern coastal commu-nities that rely on hunting marine mammals, fewadaptation options are likely to be available.Consequently, it is possible that projected climatechange will overwhelm the available responses.Some communities may be forced to reduce theirdependence on the wild harvest,or relocate.General measures to increase the income andwealth of subsistence-dependent communities,andconsequently their ability to adapt to large-scale

301

Chapter 10 / Alaska

Changes in total and seasonal river runoff are likelyto interact with other changes in the oceans andcryosphere to yield complex patterns of ecologicaland socioeconomic impacts. Combined withincreases in sea level and storm surge,they wouldalter the hydrology of coastal wetlands and deltas,possibly impairing seabird and shorebird breeding(Meehan et al.,1999). Spring flooding due to riverice jams is likely to reduce,bringing reduced floodrisk to riverside communities but possibly dryingout lakes and riparian and delta ecosystems thatdepend on periodic flooding (Beltaos and Prowse,2000;Prowse and Conly, 1998). In contrast,smallerrivers and streams that presently freeze solid willlikely retain some flowing water beneath the ice,enhancing fish habitat but also making these riversliable to ice-jam flooding for the first time.

Agriculture

A lengthened growing season could possibly bring asubstantial increase in Alaskan agricultural land andproduction,including the potential for introductionof new crops and animals. Permafrost thawing willlikely impair agricultural potential in moisture-limit-ed regions,however, by allowing drying of surfacesoils. Thawing can also exacerbate soil erosion andloss of organic materials,or obstruct agriculture inregions of ice-rich permafrost through thermokarstformation. Some of these ef fects could be mitigatedthrough irrigation and soil conservation measures.In the long term,projected changes in tundraecosystems are likely to seriously harm reindeerherding,through increased snow or ice cover of for-age during warmer wetter winters, reduction of for-age quality in dry summers,tundra fires,and expan-sion of forest into tundra. A climatic contribution tothe large recent decline in Russia’s domestic rein-deer herds is possible (Weller and Lange,1999).

Tourism

Alaskan tourism has increased in recent decades,inparallel with climate warming,although it is notclear how much of the increase can be attributed tothe warming climate. Continued warming couldpossibly bring significant further expansion oftourism,with associated economic benefits andincreased risks to sensitive ecosystems and commu-nities (Nuttall,1998;Weller and Lange,1999).

changes in the subsistence resources on which theydepend, would likely mitigate the impacts of lostsubsistence resources on nutrition,health,andincomes,but would likely have little effect in miti-gating the associated social and cultural impacts(Nuttall,1998).

ADDITIONAL ISSUES

Due to limited time and resources in this firstAssessment,the Alaskan regional study has focusedprincipally on the four critical issues discussedabove. The choice of these four reflected the judg-ments and concerns of participants in Alaskan work-shops from 1997 to 1999,but does not imply thatthese are the only important areas of climateimpacts in Alaska. Other areas of potentially signifi-cant impact include,e.g.,freshwater, agriculture,tourism, recreation,and human health. Very prelimi-nary discussions are provided here of freshwater,agriculture,and tourism,based on contributions toworkshops from 1997 to 1999,and the scientific lit-erature.

Freshwater

Present and projected climate warming is likely toalter both seasonal and annual river flows in theYukon and other Alaskan rivers,but the aggregateeffects are quite uncertain. Over the entire Arcticbasin,a recent analysis of runoff based on stream-flow gauges found a significant increase in winterrunoff with the largest increases in Alaska andSiberia,consistent with recent winter warming. Thelarger spring and summer flows show a complexspatial mix of increases and decreases,with signifi-cant decreases in Western Canada (both Hudson Bayand the McKenzie Basin) and small increases inAlaska (Lammers et al.,2000).

Model projections of future climate ef fects on Arcticriver flows are also spatially mixed. Total Arcticbasin runoff is projected to increase (Van Blarcumet al.,1995;Shiklomanov, 1997;Hagemann andDumenil,1998;Walsh et al.,1998), by 10 - 20% annu-ally and 50 - 100% in the small winter flows by thetime of CO2 doubling (Clair et al.,1998;Shiklomanov et al.,2000). One analysis of theMackenzie basin projected a decrease in annualflow under climate change,however, consistentwith the observed spatial distribution of recent flowchanges (Kerr, 1997). Uncertainties in these projec-tions for spring,summer, and annual flows are large,however.

302


CRUCIAL UNKNOWNS ANDRESEARCH NEEDSDespite the strong evidence of present impacts ofclimate change in Alaska,there remain substantialuncertainties regarding all major domains of futureimpacts. Even for many impacts that are presentlydeveloping,there is insufficient systematic observa-tion and continuing uncertainty about importantcausal processes. Many near-term research needs follow from this lack.

Cryosphere

Systematic observations of changes presently under-way in permafrost are needed over a large scale,tospecify the rate and character of present warmingand thawing more accurately and to resolve presentuncertainties regarding the contribution of climatewarming to the observed thaw. Further understand-ing of the dynamics of permafrost are also needed,inorder to identify likely rate and character of futureloss under continued climate warming,and therebyto help identify the most appropriate responses.

Better understanding is needed of the interactionbetween thawing,thermokarst formation,surfacehydrology, and ecosystem and site characteristics,inorder to understand the effects of permafrost thaw-ing on surface plant communities and carbon stor-age in vegetation and soils. For instance, recentobservations in the Brooks Range foothills,wherethe observed depth of thawing on adjacent sites fac-ing identical climates was strongly dependent on theacidity of the tundra,illustrate the importance ofthese investigations (Walker et al.,1998).

Continued monitoring of changes in Arctic sea ice isneeded,both in area and thickness,and of coastalerosion and its relationship to both sea ice andstorm conditions. In addition,better modeling ofArctic ice dynamics is needed to improve climatemodels,particularly as regards their projections inArctic regions.

Forest and Tundra Ecosystems

More systematic large-scale observation of changesin the productivity, range,species mix,and diseaseand insect activity underway in the boreal andcoastal forest are needed,as are studies (observation-al, experimental,and model-based) of potential inter-actions between climate change, forest productivityincrease,species shift,and disturbances by fire,insects,and disease.

Better understanding is needed of how carbon stor-age in boreal and tundra ecosystems is controlled. Itis possible that warming and thawing of boreal andtundra systems will release large quantities of CO2and CH4 (Anisimov et al.,1997;Bockheim et al.,2000;Goulden et al.,1998;Lindroth et al.,1998;Chase et al.,2000),but the magnitude and even thesign of these fluxes will depend on accompanyinghydrological changes and the rate of decompositionof exposed peat under warm temperatures (Oechelet al.,1993;McKane et al.,1997 a,b;Moore et al.,1998).

Models of potential future changes in these ecosys-tems under climate change should also consider theeffect of increased surface ultraviolet (UV) radiation,which has increased 15 - 30% in Arctic regions overthe past 20 years (Taalas et al.,1997). In initialexperimental studies,enhanced UV has harmed sev-eral Arctic plant species with an apparent cumula-tive effect (Bjorn et al.,1997;Callaghan et al.,1998),and strongly stimulated the growth of one mossspecies (Gehrke et al.,1996).

Marine Ecosystems

While the history of major regime shifts in theBering Sea and Gulf of Alaska ecosystems is increas-ingly well documented,the suite of factors causingthem – and in particular, the extent of climatic influ-ence in the shifts and the mechanisms by whichthey operate – are not yet understood. Furtherstudy is needed to understand both historicalregime shifts,and the likely effects of future climatechange.

One very likely consequence of global climatechange will be significant changes in seasonal andannual runoff from the glaciers of Southeast Alaska.The potential effects of these runoff changes on theGulf of Alaska and Bering Sea ecosystems is largebut not well understood,and requires investigation.

Alaska’s climate is strongly influenced by existingpatterns of climate variability, but little is knownabout how these are likely to behave in a green-house-warmed world. Climate models are nowbeginning to reproduce ENSO, but do not reproduceobserved patterns of interdecadal variability. Ifthese patterns continue to behave as they did dur-ing the 20th century, then the changes projectedfrom climate models must be modified by theseobserved patterns of variability. But whether thesecycles will behave as they have in a greenhouse-warmed world,or will show coupled changes,is acritical unknown (Fyfe et al.,1999).

303

Chapter 10 / Alaska

One major feedback,the ice-albedo feedback,isincluded in climate model projections. Present limi-tations in modeling sea-ice dynamics introduce sub-stantial uncertainty to the representation of thisfeedback in climate models,however. A secondfeedback can change albedo through ecologicalprocesses,and is not presently represented in cli-mate models. Since forests are darker than tundra,the expansion of boreal forest into the tundra zoneas climate warms can reduce the reflectivity of theEarth’s surface. It has been suggested that thisprocess could amplify regional climate change b yup to 50% (Foley et al.,1994). Both these processesneed further study.

Other potential feedbacks would operate throughclimate change altering patterns of natural green-house gas emissions. Both boreal and tundra ecosys-tems are large carbon stores,particularly in theirsoils. As discussed above,the controls on carbonstorage or release from these systems are not yetwell understood,and it is possible that climatechange could produce large increases in carbonsequestration,or large increases in release of eitherCO2 or methane. Advancing understanding of thesecontrols,and projecting future release under climatechange,are key priorities.

A larger, though likely more remote uncertainty, con-cerns the possibility of methane release fromhydrates. Hydrates are crystal structures,in whichmethane molecules are held at high density bybeing encased in an ice lattice at high pressure orlow temperature. Methane hydrates occur world-wide in enormous quantity. Estimated worldreserves are 400 million trillion cubic feet (TCF),versus 5,000 TCF of conventional gas reserves - inocean sediments at high pressure or low tempera-ture,and at substantial depths in continuous per-mafrost. Hydrates are of interest as a fuel source,although technical challenges to their exploitationare serious. They are also of interest in the longterm for the risk of atmospheric release. While theprospect of significant releases over the next centu-ry is presently judged to be highly speculative(Kvenvolden,1999),long-term Arctic warming couldeventually release methane by warming coastalwaters to shift the depth at which hydrates becomestable,or by thawing hydrate-rich permafrost.

Human Dimensions

The possible interactions between climate-drivenchanges to natural systems and human responsesare strong,diverse,and not well understood.Research is needed in the following areas:

• Develop and refine techniques to assess vulnera-bility of communities,what determines vulnera-bility, and strategies to reduce it;

• Investigate interactions between policy, fishingpressure,and the state of marine ecosystems,andhow these are likely to adjust under climatechange;

• Investigate how changes in economic conditionsand forest management practices are likely tointeract with climate-driven changes in forestecosystems;

• Investigate response processes of subsistence-reliant communities to changes in the characterof subsistence resources available;and

• Study the potential longer-term influence of cli-mate change on aggregate prospects for Alaskanpopulation and economic growth,including,e.g.,large changes in the level and distribution ofpopulation,large-scale conversion of forestedland to agriculture or settlement,or large shiftsin the distribution of economic activity.

Arctic Feedbacks

The Arctic regions,including Alaska,are the site ofseveral key uncertainties in modeling the global cli-mate. These include the potential role of changes inthe temperature,salinity, and flow regime of theArctic Ocean,and of changes in Arctic sea ice,inchanges in the global thermohaline circulation,pos-sibly including large or rapid changes (Stocker andSchmittner, 1997;Manabe and Stouffer, 1993;Broecker et al.,1990). They also include a numberof potentially important climate-change feedbacks,processes whereby climate change can cause moreclimate change,either by increasing absorption ofsolar radiation or by increasing emissions of thegreenhouse gases that drive climate change.Gaining further understanding of these processesare key research priorities. While the influence ofthese processes on regional climate in the Arcticmay be especially large,they are also of much widerimportance for their contribution to driving climateat the global scale.

304


LITERATURE CITEDAnisimov, O.A., N. I.Shiklomanov, and F. E.Nelson,Global warming and active-layer thickness: Resultsfrom transient general circulation models, Global andPlanetary Change, 15(3-4),pp.61-77,1997.

Anisimov, O.A.,and F. E.Nelson, Permafrost zonationand climate change in the Northern Hemisphere:Results from transient general circulation models,Climatic Change, 35(2),241-258,1997 .

Barber,V.A., G. P. Juday, and B. P. Finney, Reduced growth

of Alaskan white spruce in the 20th century from tem-perature-induced drought stress, Nature, 405, 6787,668-673, 2000.

Beltaos,S.,and T. D. Prowse,Climate impacts onextreme ice jam events in Canadian rivers,Hydrological Sciences Journal, in press,2000.

Berkes, F.,Indigenous knowledge and resource manage-ment systems in the Canadian sub-Arctic,in LinkingSocial and Ecological Systems , edited by F. Berkes,C.Folke,and J. Colding, Cambridge University Press,Cambridge,United Kingdom,pp.98-128,2000.

Berman,M., G. P. Juday, R.Burnside,Climate change andAlaska’s forests: People,problems,and policies,inProceedings of a workshop,Assessing the conse-quences of climate change for Alaska and the BeringSea region, Fairbanks,29-30 October 1998,edited by G.Weller and P.A.Anderson,Center for Global Changeand Arctic System Research,University of AlaskaFairbanks,1999.

BESIS,The impacts of global climate change in theBering Sea region,an assessment conducted by theInternational Arctic Science Committee under itsBering Sea Impacts Group (BESIS),International ArcticScience Committee,Oslo,Norway, 1997.

Bjorn,L. O.,T.V. Callaghan,C.Gehrke, D. Gwynne Jones,B.Holmgren, U. Johanson,and M.Sonesson,Effects ofenhanced UV-B radiation on sub-Arctic vegetation, InEcology of Arctic Environments, edited by S. J.Woodin,and M.Marquiss,Blackwell Science,Oxford,UnitedKingdom,pp.241-253,1997.

Bockheim, J. G.,L.R.Everett,K.M.Hinkel, F. E.Nelson,and J. Brown,Soil organic carbon storage and distribu-tion in Arctic tundra,Barrow,Alaska, Soil ScienceSociety of America Journal, 63(4),934-940,1999.

Boer, G. J., N.A.McFarlane,R.Laprise, J. D. Henderson,and J.-P. Blanchet,The Canadian Climate Centre spectralatmospheric general circulation model, Atmosphere-Ocean, 22(4),397-429,1984.

Boer, G. J., G. M.Flato,M.C.Reader, and D. Ramsden,Atransient climate change simulation with historical andprojected greenhouse gas and aerosol forcing:Experimental design and comparison with the instru-

mental record for the 20th century, Climate Dynamics,16,405-426,1999a.

Boer, G. J., G. M.Flato,and D. Ramsden,A transient cli-mate change simulation with historical and projectedgreenhouse gas and aerosol forcing:projected climate

for the 21st century, Climate Dynamics, 16,427-450,1999b.

Brodeur, R. D.,B.W. Frost,S.R.Hare,R.C. Francis,and W.J. Ingraham,Interannual variations in zooplankton bio-mass in the Gulf of Alaska,and covariation withCalifornia current zooplankton biomass, CalCOFIReport 37, 80-89,1996.

Broecker,W. S., G. Bond,and M.A.Klas,A salt oscillatorin the glacial Atlantic? 1.The concept.Paleoceanography, 5, 469-477,1990.

Brown, J., O. J. Ferrians,Jr., J.A.Heginbottom,and E.S.Melnikov, Circum-Arctic map of permafrost and groundice conditions,International Permafrost Association,published as Map CP45, US Geological Survey,Washington,DC,1997.

Brown, J.,and S.Solomon,Arctic coastal dynamics:Report of an international workshop, Woods Hole,Massachusetts, A geological survey of Canada open filereport,in press,2000.Available at Web site,http://www.awi-potsdam.de/www-pot/geo/acd.html

Buch,E.,S.A.Horsted,and H.Hovgard,Fluctuations inthe occurrence of Cod in Greenland waters and theirpossible causes, ICES Marine Science Symposium , 198,158-174,1994.

Callaghan,T.V.,C. Korner, S.E.Lee,and J. H.C.Cornelison, Part 1: Scenarios for ecosystem responsesto global change,in Global Change in Europe’s ColdRegions, edited by O.W. Heal,T.V. Callaghan, J. H.C.Cornelissen,C. Korner and S.E.Lee,pp.11-63,European Commission Ecosystems Research Report 27,Luxembourg,1998.

Callaway, D., J. Eamer, E.Edwardwen,C. Jack,S.Marcy,A.Olrun,M. Patkotak, D. Rexford,and A.Whiting,Effects ofclimate change on subsistence communities in Alaska,in Proceedings of a workshop,Assessing the conse-quences of climate change for Alaska and the BeringSea region, Fairbanks,29-30 October 1998,edited by G.Weller and P.A.Anderson,Center for Global Changeand Arctic System Research,University of AlaskaFairbanks,1999.

Cavalieri, D. J., P. Gloersen,C.L. Parkinson, J. C.Comiso,and H. J. Zwally, Observed hemispheric asymmetry inglobal sea ice changes, Science, 278, 1104-1106,1997.

Chapin, F. S.,III, G. R.Shaver,A.E.Giblin,K. J.Nadelhoffer, and J.A.Laundre,Responses of Arctic tun-dra to experimental and observed changes in climate,Ecology 76, 3,694-711,1995.

Chapin, F. S.,III,and A.M.Starfield,Time lags and novelecosystems in response to transient climatic change inarctic Alaska,Climatic Change, 35, 449-461,1997.

Chapman,W. L.,and J. E.Walsh,Recent variations of seaice and air temperature in high latitudes, Bulletin ofthe American Meteorological Society, 74, 1,33-47,1993.

305

Chapter 10 / Alaska

Ferrians, O. J., Permafrost map of Alaska, US GeologicalSurvey Miscellaneous Geologic Investigations Map I-445, 1:2,500,000,1965.

Flato, G. M., G. J. Boer,W. G. Lee, N.A.McFarlane, D.Ramsden,M.C.Reader, and A. J.Weaver,The CanadianCentre for Climate Modelling and Analysis GlobalCoupled Model and its climate, Climate Dynamics,16(6),451-467,2000.

Fleming,R.A.,and J-N Candau,Influences of climaticchange on some ecological processes of an insect out-break system in Canada’s boreal forests and the impli-cations for biodiversity, Environmental Monitoringand Assessment, 49, 235-249,1997.

Fleming,R.A.,and J.A.Volney, Effects of climate changeon insect defoliator population processes in Canada’sboreal forest: Some plausible scenarios, Water, Air, andSoil Pollution, 82, 445-454,1995.

Foley, J.A., J. E. Kutzbach,M.T. Coe,S.Levis, Feedbacksbetween climate and boreal forests during theHolocene epoch, Nature, 371(6492),52-54 (September1),1994.

Forbes, D. L.,and R.B.Taylor, Ice in the shore zone andthe geomorphology of cold coasts,Progress in PhysicalGeography, 18, 59-89,1994.

Francis,R.C.,S.R.Hare,A.B.Hollowed,and W. S.Wooster, Effects of interdecadal climate variability onthe oceanic ecosystems of the Northeast Pacific,Fisheries Oceanography 7, 1-21,1998.

French,H.M.,and I.E.Egorov, 20th century variationsin the southern limit of permafrost near Thompson,northern Manitoba,Canada,in Proceedings of theSeventh International Conference on Permafrost,editedby A. G. Lewkowica and M.Allard,Centre d’ÉtudesNordiques,Université Laval,Quebec,pp.297-304,1998.

Furniss,R.L.,and V. M.Carolin,Western forest insects,USDA Forest Service Miscellaneous Publication 1339,Washington,DC,1977.

Fyfe, J.C., G. J. Boer, and G. M.Flato,The Arctic andAntarctic oscillations and their projected changesunder global warming, Geophysical Research Letters,26(11),1601-1604 (June 1),1999.

Goldsmith, O. S.,Structural analysis of the Alaska econo-my: A perspective from 1997,prepared for AlaskaScience and Technology Foundation by Institute ofSocial and Economic Research,University of Alaska,Anchorage,1997.

Gordon,H.B.,and S. P. O’Farrell,Transient climatechange in the CSIRO coupled model with dynamic seaice, Monthly Weather Review, 25(5),875-907,1997

Gorham,E.,The biogeochemistry of northern peatlandsand its possible responses to global warming,in Bioticfeedback in the global climatic system: Will the warm-ing feed the warming?,edited by G. M.Woodwell and F.T. MacKenzie,pp.169-187,Oxford University Press,Oxford,United Kingdom,1995.

Chase,T. N.,R.A.Pielke,Sr.,T. G. F. Kittel,R.R.Nemani,and S.W. Running,Simulated impacts of historical landcover changes on global climate in northern winter,Climate Dynamics, 16(2-3),93-105,2000.

Ciais, P., P. P.Tans,M.Trolier, J.W. C.White,and R.R. J.Francey,A large Northern Hemisphere ter restrial CO2

sink indicated by the C13/C12 ratio for atmospheric

CO2, Science, 269,1098-1102,1995.

Clair T.A., J. Ehrman,and K.Higuchi,Changes to therunoff of Canadian ecozones under a doubled CO2atmosphere, Canadian Journal of Fisheries andAquatic Sciences, 55(11) 2464-2477,1998.

Cole,H.,V. Colonell,and D. Esch,The economic impactand consequences of global climate change on Alaska’sinfrastructure,in Proceedings of a workshop,Assessingthe consequences of climate change for Alaska and theBering Sea region, Fairbanks,29-30 October 1998,edit-ed by G.Weller and P.A.Anderson,Center for GlobalChange and Arctic System Research,University ofAlaska Fairbanks,1999.

Criddle,K.R.,M.Hermann, J.A.Greenberg,and E.M.Feller, Climate fluctuations and revenue maximizationin the eastern Bering Sea fishery for walleye pollock,North American Journal of Fisheries Management18(1),1-10,1998.

Dowdeswell, J., J. O. Hagen,H.Bjornsson,A.Glazovsky, P.Holmlund, J. Jania,E. Josberger, R. Koerner, S.Ommanney, and B.Thomas,The mass balance of cir-cum-Arctic glaciers and recent climate change,Quaternary Research, 48, 1-14,1997.

Dyke,L. D., J. M.Aylsworth,M.M.Burges, F. M.Nixon,and F.Wright, Permafrost in the Mackenzie Basin,itsinfluence on land-altering processes,and its relation-ship to climate change,in Mackenzie Basin ImpactStudy Final Report,edited by S. J. Cohen,AtmosphericEnvironment Service,Environment Canada,Downsview, Ontario,1997.

Echelmeyer K.A,W. D. Harrison,C. F. Larsen, J. Sapiano, J.E.Mitchell, J. DeMallie,B.Rabus, G.Adalgeirsdottir, andL.Sombardier,Airborne surface profiling of glaciers: Acase-study in Alaska, Journal of Glaciology, 42(142),538-547,1996

Estes, J.A.,M.T.Tinker,T. M.Williams,and D. F. Doak,Killer whale predation on sea otters linking oceanicand nearshore ecosystems, Science, 282, 473-476,1998.

Everett, J.T.,and B.B. Fitzharris,The Arctic and theAntarctic, chapter 3,in The Regional Impacts ofClimate Change: An Assessment of Vulnerability, edit-ed by R.T.Watson,M.C.Zinyowera,and R.H.Moss,Special report of IPCC Working Group II,CambridgeUniversity Press,Cambridge,United Kingdom,1998.

Fay, F.H.,The role of ice in the ecology of marine mam-mals of the Bering Sea,in Oceanography of the BeringSea,edited by D.W. Hoos and E. J. Kelley, University ofAlaska Institute of Marine Sciences,OccasionalPublication No.2,1974.306


Goulden,M.L.,et al.,Sensitivity of boreal forest carbonbalance to soil thaw, Science, 279, 214-217,1998.

Groisman, P.Y.,and D.A.Easterling,Variability andtrends of precipitation and snowfall over the UnitedStates and Canada, Journal of Climate, 7, 184-205,1994.

Gunn,A.,Responses of Arctic ungulates to climaticchange,in Human Ecology and Climate Change, edit-ed by D. L. Peterson and D. R. Johnson,pp.90-104,Taylor and Francis,Washington DC,1995.

Hagemann,S.,and L.Dümenil,A parameterization ofthe lateral waterflow for the global scale, ClimateDynamics, 14, 17-31,1998.

Hall, J.V.,W. E. Freyer, and B. O.Wilen,Status of Alaskawetlands Alaska Region,US Fish and Wildlife Service,Anchorage,Alaska,1994.

Hennon P. E.,and C. G. Shaw,The enigma of yellow-cedar decline:What is killing these long-lived,defensivetrees?, Journal of Forestry, 95(12),4-10,December1997.

Hogg,E.H.,Simulation of inter-annual responses oftrembling aspen stands to climatic variation and insectdefoliation in western Canada, Ecological Modeling,114 2-3,175-193,1999.

Hogg,E.H.,and A. G. Schwarz,Regeneration of plantedconifers across climatic moisture gradients on theCanadian prairies: Implications for distribution and cli-mate change, Journal of Biogeography, 24 4,527-534,July 1997.

Holsten,E.H.,and R.Burnside, Forest health in Alaska:An update, Western Forester, 24 (4),8-9,1997.

Holsten,E.H., P. E.Hennone,and R.A.Werner, Insectsand diseases of Alaska forests,USDA Forest Service,Alaska Region Report Number 181, revised October1985,Anchorage,Alaska,1985.

Huppert, D.,Managing the groundfish fisheries ofAlaska:history and prospects. Reviews in AquaticSciences. 4(4),339-373.1991.

Hurrell, J.W.,Influence of variations in extratropicalwintertime teleconnections on Northern Hemispheretemperature, Geophysical Research Letters, 23, 665-668,1996.

Hurrell, J.W.,Decadal trends in the North Americanoscillations: Regional temperatures and precipitation,Science, 269, 676-679,1995.

Johannessen, O. M.,M.Miles,and E.Bjorgo,The Arctic’sshrinking sea ice, Nature, 376, 126-127,1995.

Johannessen, O. M.,E.V. Shalina,and M.W. Miles,Satellite evidence for an Arctic sea ice cover in transfor-mation, Science, 286 5446,1937-1939,1999.

Johns T. C.,R.E.Carnell, J. F. Crossley, J. M.Gregory, J. F. B.Mitchell,C.A.Senior, S. F. B.Tett,and R.A.Wood,TheSecond Hadley Centre coupled ocean-atmosphereGCM: Model description,spinup and validation,Climate Dynamics, 13, 103-134,1997

Jorgenson,M.T.,C.H.Racine, J. C.Walters,and T. E.Osterkamp,Widespread and rapid permafrost degrada-tion and associated ecological changes caused by awarming climate on the Tanana Flats,Central Alaska,Climatic Change, in press,2000.

Juday, G. P. and R.Barry, Growth response of uplandwhite spruce in Bonanza Creek LTER in Central Alaska,Proceedings of Ecological Society of America , AnnualCombined Meeting, 10-14 August 1996,ProvidenceRhode Island,222,1996.

Juday, G. P, R.A.Ott, D.W.Valentine,and V.A.Barber,Forests, climate stress,insects,and fire,in Implicationsof Global Change in Alaska and the Bering SeaRegion, Proceedings of a workshop,University ofAlaska, Fairbanks, June 1997,edited by G.Weller and P.A.Anderson,Center for Global Change and ArcticSystem Research,23-49,April 1998.

Kasischke,E.S.,K.Bergen,R. Fennimore, F. Sotelo, G.Stephens,A. Janetos,and H.H.Shugart,Satellite imagerygives clear picture of Russia’s boreal forest fires, Eos, 8013,141-147,1999.

Kasischke,E.S.,and B. J. Stocks (Eds.), Fire, ClimateChange, and Carbon Cycling in the Boreal Forest,Ecological Studies Series , Springer-Verlag,New York,2000.

Kattenberg,A., F. Giorgi,H.Grassl, G.A.Meehl, J. F. B.Mitchell,R. J. Stouffer,T.Tokioka,A. J.Weaver, and T. M.L.Wigley, Climate models: Projections of future change,in Climate Change 1995: The Science of ClimateChange, Contribution of Working Group I to theSecond Assessment Report of the Inter-governmentalPanel on Climate Change, edited by J.T. Houghton,L.G. Meira Filho,B.A.Callander, N. Harris,A.Kattenberg,and K.Maskell,pp.285-357,Cambridge UniversityPress,Cambridge,United Kingdom,1996.

Keeling,C. D., J. F. S.Chin,and T. P.Whorf, Increasedactivity of northern vegetation inferred from atmos-pheric CO2 measurements, Nature, 382(6587),146-

149,1996.

Kerr, J.A.,Future water levels and flows for Great Slaveand Great Bear Lakes,Mackenzie River, and MackenzieDelta,in Mackenzie Basin Impacts Study (MBIS), FinalReport,edited by S. J. Cohen,Environment Canada,Ottawa,1997.

Keyser,A.R., J. S.Kimball,R.R.Nemani,and S.W.Running,Simulating the effects of climate change onthe carbon balance of North American high latitudeforests, Global Change Biology, 6, 1-11,2000.

Knapp, G., P. Livingston,and A. Tyler, Human effects ofclimate-related changes in Alaska commercial fisheries,in Proceedings of a workshop,Assessing the conse-quences of climate change for Alaska and the BeringSea region, Fairbanks,29-30 October 1998,edited by G.Weller and P.A.Anderson,Center for Global Changeand Arctic System Research,University of AlaskaFairbanks,1999.

307

Chapter 10 / Alaska

Maslanik, J.A.,M.C.Serreze,and R.G. Barry, Recentdecreases in Arctic summer ice cover and linkages toatmospheric circulation anomalies, GeophysicalResearch Letters, 23(13),1677-1680,1996.

Maslanik, J.A.,M.C.Serreze,and T.Agnew, On therecord reduction in 1998 Western Arctic Sea-ice cover,Geophysical Research Letters, 26(13),1905-1908,1999.

McGuire,A. D.,and J. E.Hobbie,Global climate changeand the equilibrium responses of carbon storage inArctic and subArctic regions,in Modeling the Arcticsystem: A workshop report on the state of modeling inthe Arctic System Science Program,pp.53-54,TheArctic Research Consortium of the United States,Fairbanks Alaska,1997.

McGuire,A. D., J. M.Melillo, D.W. Kicklighter, and L.A.Joyce,Equilibrium responses of soil carbon to climatechange: Empirical and process-based estimates,Journal of Biogeography, 22, 785-796,1995.

McKane,R.B.,E.B.Rasetter, G. R.Shaver, K. J.Nadelhoffer,A.E.Giblin, J.A.Laundre,and F. S.Chapin,III,Climate effects of tundra carbon inferred fromexperimental data and a model, Ecology, 78, 1170-1187,1997a.

McKane,R.B.,E.B.Rasetter, G. R.Shaver, K. J.Nadelhoffer,A.E.Giblin, J.A.Laundre,and F. S.Chapin,III,Reconstruction and analysis of historical changes incarbon storage in Arctic tundra,Ecology 78, 1188-1198,1997b.

McPhee,M. G.,T. P. Stanton, J. H.Morison,and D. G.Martinson, Freshening of the upper ocean in the Arctic:Is perennial sea ice disappearing?, GeophysicalResearch Letters, 25(10),1729-1732,1998.

Meehan,R.,V. Byrd, G. Divoky, and J. Piatt,Implicationsof climate change for Alaska’s seabirds,in Proceedingsof a workshop,Assessing the consequences of climatechange for Alaska and the Bering Sea region, Fairbanks,29-30 October 1998,edited by G.Weller and P.A.Anderson,Center for Global Change and Arctic SystemResearch,University of Alaska Fairbanks,1999.

Meier, M. F., Ice, climate,and sea level: Do we knowwhat is happening?, in Ice in the Climate System,NATO ASI Series, Vol.112, edited by W. R. Peltier,Springer-Verlag,Berlin,pp.141-160,1993.

Melillo, J. M., D.W. Kicklighter,A. D. McGuire,W.T.Peterjohn,K.M.Newkirk,Global change and its effecton soil organic carbon stocks,in Role of nonlivingorganic matter in the Earth’s carbon cycle,edited by R.G. Zepp and C.H.Sontaff, pp.175-189, John Wiley andSons:New York,1995.

Melillo, J. M.,I.C.Prentice, G. D. Farquhar, E.-D Schulze,and O. E.Sala,Terrestrial biotic responses to environ-mental change and feedbacks to climate, chapter 9,inClimate Change 1995: The Science of ClimateChange, Contribution of Working Group I to theSecond Assessment Report of the IPCC, edited by J.T.Houghton,L. G. Meira Filho,B.A.Callander, N. Harris,A.

Krabil,W.,E. Frederick,S.Manizade,C.Martin, J.Sonntag,R.Swift,R.Thomas,W.Wright,and J.Yungel,Rapid thinning of parts of the Southern Greenland icesheet, Science, 283, 1522-524,1999.

Kruse, G. H.,Salmon run failures in 1997-1998: A linkto anomalous ocean conditions?, Alaska FisheryResearch Bulletin, 5(1),55-63,1998.

Kvenvolden,K.A., Potential effects of gas hydrate onhuman welfare, Proceedings of the National Academyof Sciences, 96(7),3420-3426,March 30,1999.

Kwong,Y.T. J.,and T.Y. Gan,Northward migration ofpermafrost along the Mackenzie Highway and climaticwarming, Climatic Change 26, 399-419,1994.

Lachenbruch A.H.,and V. Marshall,Changing climate:Geothermal evidence form permafrost in the AlaskanArctic, Science, 234, 689-696,1986

Lamke,R. D.,Alaska surface-water resource,in NationalWater Summary 1985,Hydrologic events and surface-water resources: US Geological Water-Supply Paper2300, D.W. Moody, E.B.Chase,and D.A.Aronson,(Compilers),pp.137-144,1986.

Lammers,R.B.,A.I.Shiklomanov, C. J.Vorosmarty, B.M.Fekete,and B. J. Peterson, Assessment of contemporaryArctic river runoff based on observational dischargerecords, Journal of Geophysical Research-Atmospheres, in press,2000.

Leask,L.,Changing ownership and management ofAlaska lands, Alaska Review of Social and EconomicConditions, 22:4,October 1985.

Lindroth,A.,A.Grelle,and A-S.Moren,Long-term meas-urements of boreal forest carbon balance reveal largetemperature sensitivity, Global Change Biology, 4(4),443-450,1998.

Lowry, L. F. and F. H. Fay, Seal eating by Walruses in theBering and Chukchi Seas, Polar Biology, 3, 11-18,1984.

Lükewille,A.,and R. F.Wright,Experimentally increasedsoil temperature causes release of nitrogen at a borealforest catchment in southern Norway, Global ChangeBiology, 3, 13-21,1997.

McFarlane, N.A., G. J. Boer, J. P. Blanchet,and M.Lazare,The Canadian Climate Centre second-generation gener-al circulation model and its equilibrium climate,Journal of Climate, 5, 1013-1044,1992.

Manabe,S.and R. J. Stouffer, Simulation of abrupt cli-mate change induced by freshwater input to the NorthAtlantic Ocean, Nature, 378 (6553),165-167,1995.

Mantua, N. J.,S.R.Hare,Y. Zhang, J. M.Wallace,and R.C.Francis,A Pacific interdecadal climate oscillation withimpacts on salmon production, Bulletin of theAmerican Meterorological Society, 78, 1069—1079,1997.

Marasco,R. J. and W.Arom,Explosive evolution:thechanging Alaska groundfish fishery, Reviews inAquatic Sciences, 4(4),299-315,1991.

308


Kattenberg,and K.Maskell,Cambridge University Press,Cambridge,United Kingdom,1996.

Mitchell J. F. B.,T. C. Johns, J. M.Gregory, and S.Tett,Climate response to increasing levels of greenhousegases and sulphate aerosols, Nature, 376, 501-504,1995.

Mitchell J. F. B.,and T. C. Johns,On modification of glob-al warming by sulfate aerosols, Journal of Climate,10(2),245-267,1997.

Moore T. R., N.T. Roulet, J. M.Waddington,Uncertaintyin predicting the ef fect of climatic change on the car -bon cycling of Canadian peatlands, Climatic Change,40(2),229-245,1998.

Mulvaney, K.,Arctic voices: Global warming is chang-ing the traditional Eskimo environment, New Scientist,pp.55,November 14,1998.

Myneni,R.B.,C. D. Keeling,C. J.Tucker, G.Asrar, and R.R.Nemani,Increased plant growth in the northernhigh latitudes from 1981 to 1991, Nature, 386, 698-702,1997.

Nash,C.E.,and Associates, D. Duffy, Miller’s Reach FireStrategic Economic Recovery Plan,Matanuska-SusitnaBorough Department of Planning,October 1997.

National Marine Fisheries Service (NMFS),Economicstatus of US fisheries,1996,US Government PrintingOffice,Washington,DC,1996.

NRC (National Research Council), The Bering SeaEcosystem,Committee on the Bering Sea Ecosystem,Polar Research Board,National Academy Press,Washington,DC,1996.

Neilson,R. P.,I.C.Prentice,B.Smith,T. G. F. Kittel,and D.Viner, Simulated changes in vegetation distributionunder global warming,Annex C,in The RegionalImpacts of Climate Change: An Assessment ofVulnerability, Special Report of IPCC Working GroupII,edited by R.T.Watson,M.C.Zinyowera,R.H.Moss,andD. J. Dokken,pp.439-456,Cambridge University Press,Cambridge,United Kingdom,1998.

Nixon, J. F.,Effect of climatic warming on pile creep inpermafrost, Journal of Cold Regions Engineering, 4, 1,pp.67-73,1990.

National Oceanic and Atmospheric Administration(NOAA),FOCI International Workshop on RecentConditions in the Bering Sea, NOAA-Pacific MarineEnvironmental Laboratory, Seattle,Washington,1999.

Nuttall,M.,Protecting the Arctic: Indigenous Peoplesand Cultural Survival, Harwood Academic Publishers,Amsterdam,1998.

Oechel,W. C.,S. J. Hastings, G.Vourlitis, M. Jenkins, G.Richers,and N. Gruike,Recent change of Arctic tundraecosystems from a net carbon dioxide sink to a source,Nature, 361, 520-523,1993.

Oechel,W. C., G. L.Vourlitis,S. J. Hastings,and S.A.Bochkarev, Change in Arctic CO2 flux over two

decades: Effects of climate change at Barrow,Alaska,Ecological Application, 5(3),846-855,1995.

Oechel W. C., G. L.Vourlitis,and S. J. Hastings,Cold sea-son CO2 emission from Arctic soil,Global

Biogeochemical Cycles, 11, 163-172,1997a.

Oechel,W. C.,A.C.Cook,S. J. Hastings,and G. L.Vourlitis,Effects of CO2 and climate change on Arctic

ecosystems,in Ecology of Arctic Environments, editedby S. J.Woodin and M.Marquiss,Blackwell Science,Oxford,United Kingdom,pp.255-273,1997b.

Osterkamp,T. E.,Evidence for warming and thawing ofdiscontinuous permafrost in Alaska, ElsevierOceanography Series, 75(44),85,1994.

Osterkamp,T. E., D. C.Esch,and V. E.Romanovsky,Permafrost, chapter 10,in Implications of global changein Alaska and the Bering Sea region,Proceedings of aworkshop, June 3-6 1997,edited by G.A.Weller and P.A.Anderson,Center for Global Change and Arctic SystemResearch,University of Alaska Fairbanks,1998.

Osterkamp,T. E., J. K. Peterson,and T. S.Collett,Permafrost thicknesses in the Oliktok Point,PrudhoeBay, and Mikkelsen Bay areas of Alaska, Cold RegionsScience and Engineering, 11, 99-105,1985.

Osterkamp,T. E.and V. E.Romanovsky, Evidence forwarming and thawing of discontinuous permafrost inAlaska, Permafrost and Periglacial Processes, 10, 17-37(January-March),1999.

Osterkamp,T. E.and V. E.Romanovsky, Impacts of thaw-ing permafrost as a result of climatic warming, EOS,77(46),29,1996

Osterkamp,T. E.,L.Viereck,Y. Shur, M.T. Jorgenson,C.H.Racine,A. P. Doyle,and R. D. Boone,Observations ofthermokarst and its impact on boreal forests in Alaska ,Arctic, Antarctic, and Alpine Research, in press,2000.

Overpeck, J.,et al.,Arctic environmental change of thelast four centuries, Science, 278, 1251-1256,1997

Parker, D. E., P. D.Jones,C.K.Folland,and A.Bevan,Interdecadal changes of surface temperature since the

late 19th century, Journal of Geophysical Research –Atmosphere, 99(D7),14373-14399,1994

Piatt, J. F.,and P.Anderson,Response of common murresto the Exxon Valdez oil spill and long-term changes inthe Gulf of Alaska marine ecosystem, AmericanFisheries Society Symposium, 18, 720-737,1996.

Post,W. M.,W. R.Emanuel,P J. Zinke,and A. G.Stangenberger, Soil carbon pools and world life zones,Nature, 298, 156-159,1982.

Powell, D. S., J. L. Falukner, D. R.Darr, Z.Zhu,and D.W.MacCleery, Forest Resources of the United States,1992,USDA Forest Service,Rocky Mountain Forest and RangeExperiment Station, Fort Collins Colorado,GeneralTechnical Report RM-234,1993.

309

Chapter 10 / Alaska

Serreze,M.C., J. E .Walsh, F. S.Chapin,III,T. Osterkamp,M.Dyurgerov,V. Romanovsky,W. C.Oechel, J. Morison,T. Zhang,and R. G. Barry, Observational evidence ofrecent changes in the northern high-latitude environ-ment, Climatic Change, 46, 159-207,2000.

Shaw, J.,R.B.Taylor, S.Solomon,H.A.Christian,and D. L.Forbes, Potential impacts of global sea-level rise onCanadian coasts, The Canadian Geographer, 42, 365-379,1998.

Shiklomanov, I.A.,On the effect of anthropogenicchange in the global climate on river runoff in theYenisei basin,in Runoff computations for water proj-ects,edited by A.V. Rozhdestvensky, InternationalHydrological Program,IHP-V, Technical Documents inHydrology No. 9, UNESCO, Paris,pp.113-199,1997.

Shiklomanov, I.A.,A.I.Shiklomanov, R.B.Lammers,B.J.Peterson,and C. J.Vorosmarty,The dynamics of riverwater inflow to the Arctic Ocean,in The FreshwaterBudget of the Arctic Ocean,edited by E.L.Lewis,E.P.Jones, P. Lemke,T. D. Prowse, P.Wadhamns,KluwerAcademic Publishers,Dordrecht,in press,2000.

Sieben,B. G., D. L.Spittlehouse, J.A.McLean,and R.A.Benton,White Pine Weevil hazard under GISS climatechange scenarios in the Mackenzie Basin usingradiosonde derived lapse rates,in Mackenzie BasinImpact Study Final Report,Proceedings of a conferencein Yellowknife,Northwest Territories,May 1996,editedby S. J. Cohen,Environment Canada,AtmosphericEnvironment Service,1997.

Skinner,W. R.,B. J. Stocks, D. L.Martell,B.Bonsal,and A.Shabbar,The association between circulation anomaliesin the mid-troposphere and area burned by wildlandfire in Canada, Theoretical and Applied Climatology,63, (1-2),89-105,1999.

Smith, O. P.,and L. Johnston,The warming world:Strategy for Alaskan response,Summary of a workshop,January 5-6,2000,University of Alaska,Anchorage,2000.

Solomon,S.M.,and R.Covill,Impacts of the September1993 storm on the coastline at four sites along theCanadian Beaufort Sea coast,in Proceedings of the1995 Canadian Coastal Conference,Dartmouth NovaScotia, 2, 779-795,1995.

Stirling,I., N. J. Lunn,and J. Iacozza,Long-term trends inthe population ecology of polar bears in westernHudson Bay in relation to climate change, Arctic, 52,294-306,1999.

Stocker,T. F.,and A.Schmittner, Influence of CO 2 emis-sion rates on the stability of the thermohaline circula-tion, Nature, 388, 862-865,1997.

Stocks,B. J.,M.A. Fosberg, T. J. Lynham,L.Mearns, B.M.Wotton, Q.Yang, J-Z.Jin,K.Lawrence, G. R.Hartley, J.A.Mason,and D.W. McKenney, Climate change and forestfire potential in Russian and Canadian boreal forests,Climatic Change, 38(1),1-13,1998.

Proshutinsky,A.Y.,and M.A. Johnson,Two circulationregimes of the wind-driven Arctic Ocean, Journal ofGeophysical Research, 102, No.C6,12493-12514,1997.

Prowse,T. D.,and M.Conly, Impacts of climatic variabili-ty and flow regulation on ice jam flooding of a north-ern delta, Hydrological Processes, 12, 1589-1610,1998.

Rivkin, F. M.,Release of methane from permafrost as aresult of global warming and other disturbances,PolarGeography, 22, 105-118,1998.

Robinson,S. D.,and T. R.Moore,Carbon and peat accu-mulation over the past 1200 years in a landscape withdiscontinuous permafrost,northwestern Canada,Global Biogeochemical Cycles, 13, 591-601,1999.

Rothrock, D.A.,Y.Yu and G.A.Maykut,Thinning of theArctic Sea ice cover, Geophysical Research Letters,26(23),3469-3472 (December 1),1999.

Rouse,W. R.,et al.,Effects of climate change on thefreshwaters of Arctic and sub-Arctic North America,Hydrological Processes, 11(8),873-902 (June 30),1997.

Royer,T. C.,Baroclinic transport in the Gulf of Alaska: Afreshwater driven coastal current, Journal of MarineResearch, 39, 251-266,1981.

Royer,T. C.,Coastal freshwater discharge in theNortheast Pacific, Journal of Geophysical Research,87, 2017-2021,1982.

Royer,T. C.,High-latitude oceanic variability associatedwith the 18.6-year nodal tide, Journal of GeophysicalResearch – Oceans , 98 (C3),4639-4644 1993

Rupp,T. S., F. S.Chapin,III,and A.M.Starfield,Responseof sub-Arctic vegetation to transient climatic change onthe Seward Peninsula in northwest Alaska,GlobalChange Biology, 6(5),541-555,2000.

Sabo, G.,III., Long-term Adaptations Among ArcticHunter-Gatherers, Garland Publishing,London,UnitedKingdom,1991.

Sapiano, J.J.,W. D. Harrison,and K.A.Echelmeyer,Elevation, volume,and terminus changes of nine gla-ciers in North America, Journal of Glaciology, 44(146),119-135,1998,

Semiletov, I.et al.,The dispersion of Siberian Riverflows into coastal waters: Hydrological and hydro-chemical aspects,in The Freshwater Budget of theArctic Ocean, NATO Arctic workshop,25 April - 1 May1998,Tallinn,Estonia,edited by E.L.Lewis,1999.

Serreze,M.C., J.A.Maslanik, J. R. Key, R. F. Kokaly, and D.A.Robinson,Diagnosis of the record minimum in ArcticSea-ice area during 1990 and associated snow coverextremes, Geophysical Research Letters 22 (16),2183-2186 (August 15),1995a.

Serreze,M.C.,R. G. Barry, M.C.Rehder, and J. E.Walsh,Variability in atmospheric circulation and moisture fluxover the Arctic, Philosophical Transactions of theRoyal Society of London, Series A, 352(1699),215-225(August 15),1995b.310


Taalas, P., J. Damski,E. Kyro,M.Ginzburg,and G.Talamoni,The effect of stratospheric ozone variationson UV radiation and on tropospheric ozone at high lati-tudes, Journal of Geophysical Research , 102(D1),1533-1543,1997.

Thompson, D.W. J.,and J. M.Wallace,The Arctic oscilla-tion signature in the wintertime geopotential heightand temperature fields, Geophysical Research Letters,25, 1297-1300,1998.

Trenberth,K.E.,and J.W. Hurrell,Decadal atmosphere-ocean variation in the Pacific, Climate Dynamics, 9(6),303-319,1994.

US Army Corps of Engineers,Community improvementfeasibility report,Alaska District,Anchorage,Alaska,(April),1998.

US Census Bureau,State Population Projections,athttp://www.census.gov/population/www/projections/stproj.html,2000.

USGS BRD (US Geological Service,Biological ResourcesDivision), Status and trends of the Nation’s biologicalresources, US Government Printing Office,Washington,DC,1999.

Van Blarcum,S.C., J. R.Miller, and G. L.Russell,High-lati-tude river runoff in a doubled CO2 climate, Climatic

Change, 30(1),7-26,1995.

Van Cleve K.,W. C.Oechel,and J. L.Hom,Response ofblack spruce (Picea mariana) ecosystems to soil tem-perature modifications in interior Alaska, CanadianJournal of Forest Research, 20, 1530-1535,1990.

Veblen,T.T.,and P. B.Alaback,A comparative review offorest dynamics and disturbance in the temperate rain-forests of North and South America,in High-LatitudeRainforests and Associated Ecosystems of the WestCoast of The Americas: Climate, Hydrology, Ecology,and Conservation, edited by R. G. Lawford,P. B.Alaback,and E.Fuentes,pp.173-213,Springer-Verlag,New York,1996.

Viljhamsson,H.,Climatic variations and some examplesof their effects on the marine ecology of Icelandic andGreenland waters,in particular during the present cen-tury, Journal of the Marine Research Institute,Reykjavik, 15(1), 9-29,1997.

Vinnikov, K.Y.,A.Robock,R. J. Stouffer, J. E.Walsh,C.L.Parkinson, D. J. Cavalieri, J. F. B.Mitchell, D. Garrett,andV. F. Zakharov, Global warming and NorthernHemisphere sea ice extent,Science 286(5446),1934-1937,1999. .

Volney,W. J.A.,Climate change and management ofinsects defoliators in boreal forest ecosystems,in ForestEcosystems, Forest Management, and the GlobalCarbon Cycle, edited by M. J.Apps and D.T. Price,NATO ASI Series I (Global Environmental Change), Vol.40, Springer-Verlag Academic Publishers,Heidelberg,Germany, pp.79-88,1996.

Vourlitis, G. L.,and W. C.Oechel,Eddy covariance meas-urements of CO2 and energy fluxes of an Alaskan tus-

sock tundra ecosystem, Ecology, 80(2),686-701,1999.

Walsh, J. E.,Climate change: The Arctic as a bellwether,Nature, 352(6330),19-20 (July 4),1991.

Walsh, J. E.,V. Kattsov, D. Portis, and V. Meleshko,Arcticprecipitation and evaporation: Model results andobservational estimates, Journal of Climate , 11, 72-87,1998.

Wadhams P.,Ice thickness in the Arctic Ocean: The sta-tistical reliability of experimental data, Journal ofGeophysical Research – Oceans, 102(C13),27951-27959 (December 15),1997.

Wadhams, P.,Evidence for thinning of the Arctic icecover north of Greenland, Nature, 345(6278),795-797(June 28),1990.

Walker, D.A.,et al.,A major Arctic soil pH boundary:Implications for energy and trace gas fluxes, Nature,394, 469-472,1998.

Weller, G.A.,Regional impacts of climate change in theArctic and Antarctic,Annals of Glaciology, 27, 543-552,1998.

Weller, G.A.,and P.A.Anderson (Eds.),Implications ofglobal change in Alaska and the Bering Sea region,Proceedings of a Workshop, June 3-6 1997,Center forGlobal Change and Arctic System Research,Universityof Alaska Fairbanks,1998.

Weller, G.A.,and M.Lange (Eds.),Impacts of globalchange in the Arctic Regions, report from a workshop,Tromso Norway, 25-26 April,International ArcticScience Committee,1999.

Weller, G.,A. Lynch,T. Osterkamp,and G.Wendler,Climate trends and scenarios,in Implications of GlobalChange in Alaska and the Bering Sea Region:Proceedings of a Workshop, June 3-6 1997,edited by G.A.Weller and P.A.Anderson,Center for Global Changeand Arctic System Research,University of Alaska,Fairbanks,Alaska,1998.

Wenzel, G.W.,Warming the Arctic:Environmentalismand the Canadian Inuit,in Human Ecology andClimate Change, edited by D. L. Peterson and D. R.Johnson, Taylor & Francis,Washington,DC,1995.

Werner, R.A., Forest health in boreal ecosystems ofAlaska, The Forestry Chronicle, 72(1),43-46,1996.

Witherell, D.,Status and trends of principal groundfishand shellfish stocks in the Alaska EEZ, North PacificFishery Management Council,Anchorage Alaska,February 26,1999.

Wolfe,R. J.,and R. G. Bosworth,Subsistence in Alaska:1994 update,Alaska Department of Fish and Game,Division of Subsistence, Fairbanks,Alaska,1994.

311

Chapter 10 / Alaska

ACKNOWLEDGMENTSMany of the materials for this chapter are based oncontributions from workshop participants and mem-bers of the

Alaska Workshop and Assessment Teams

Gunter Weller*,University of Alaska FairbanksPatricia Anderson*,University of Alaska FairbanksBronwen Wang*,US Geological Survey,

Anchorage,AKMatthew Berman,University of Alaska AnchorageDon Callaway, National Park ServiceHenry Cole,Hydro Solutions & Purification LLCKeith Criddle,Utah State UniversityMerritt Helfferich,Innovating Consulting Inc.Glenn Juday, University of Alaska FairbanksGunnar Knapp,University of Alaska AnchorageRosa Meehan, U. S. Fish and Wildlife ServiceThomas Osterkamp,University of Alaska Fairbanks

Comments by the following reviewers are gratefullyacknowledged: Jerry Brown, F. Stuart Chapin III,Robert W. Corell,Henry P. Huntington, Fae Korsmo,Manfred Lange,Daniel Lashof, Susan M.Marcus,A.David McGuire,Steven P. McNulty, M.GrangerMorgan,Susanne Moser, Daniel R.Muhs,Thomas D.Newbury,Thomas Osterkamp,Thomas Pagano,Eve S.Sprunt,Soroosh Sorooshian,Roger B.Street, FrancisZweirs. Remaining errors are the sole responsibilityof the coordinating author.

* Regional Assessment co-chair

Wolfe,S.A.,S.R.Dallimore,and S.M.Solomon,Coastalpermafrost investigations along a rapidly eroding shore-line,Tuktoyaktuk Northwest Territories,Canada, inPermafrost:Proceedings of Seventh InternationalConference,edited by A. G. Lewkowicz and M.Allard,Centre d’Études Nordiques,Université Laval,Quebec,Canada,1998.

Woo,M.-K.,A. G. Lewkowicz,and W. R Rouse,Responseof the Canadian permafrost environment to climatic

change, Physical Geography, 13(4),287-317,1992.

312


Chapter 11 / The Pacific and Caribbean

CHAPTER 11

POTENTIAL CONSEQUENCES OF CLIMATEVARIABILITY AND CHANGE FOR THE US-AFFILIATED ISLANDS OF THE PACIFIC AND CARIBBEANLynne M.Carter1, 2, Eileen Shea3, Mike Hamnett4, Cheryl Anderson4, Glenn Dolcemascolo3, Charles‘Chip’ Guard5, Melissa Taylor2,6,Tony Barnston7,Yuxiang He7, Matthew Larsen8, Lloyd Loope9,LaShaunda Malone2, Gerald Meehl10


Chapter Summary



Ecological Context


Key Issues

Freshwater Resources

Public Health and Safety

Island Ecosystems

Sea-level Variability

Additional Issues

Tourism

Fisheries

Agriculture

Data Collection and Availability



Literature Cited

Acknowledgments

1Coordinating author for National Assessment Synthesis Team (from April,2000); 2USGCRP National AssessmentCoordination Office; 3East-West Center;4University of Hawaii; 5University of Guam; 6Coordinating author forNational Assessment Synthesis Team (through March,2000);7NOAA,National Centers for Environmental Prediction;8USGS,Guaynabo,Puerto Rico; 9USGS, Pacific Island Ecosystems Research Center, Haleakala Field Station; 10NationalCenter for Atmospheric Research 315


Over the 20th century, average annual air temperaturesin the Caribbean islands have increased by more than1°F (0.6°C). Average annual air temperatures in thePacific Islands have increased by about 0.4°F (0.2°C).

Globally, sea level has risen by 4 to 8 inches (10-20 cm)in the past 100 years,with significant local variation.The current rate of sea-level increase in the Caribbeanand Gulf of Mexico is about 3.9 inches (10 cm) per100 years. In the Pacific,the absolute sea level is alsorising. However, long-term changes in sea level relativeto land vary considerably within the main HawaiianIslands and some of the South Pacific islands due togeologic uplift and subsidence. Trends in relative sealevel thus vary greatly from island to island,such thatthere is no discernible long-term average trend for sea-level rise for the Pacific islands as a group. There isalso extreme variability in relative sea level associatedwith transient events such as ENSO, storm surges,andextreme lunar tides.


The Hadley, Canadian,and other global climate modelssuggest it is possible that the Pacific and Caribbeanislands will be af fected by: changes in patterns of natu -ral climate variability (such as El Niño); changes in thefrequency, intensity, and tracks of tropical cyclones(hurricanes and typhoons);and changes in patterns ofocean circulation. These islands are also very likely toexperience increased ocean temperatures and changesin sea level (including storm surges and sustained rise).

For the Caribbean and the Pacific,both of the primarymodels used in this Assessment project increasing airand water temperatures. For the Caribbean,both mod-els project slightly wetter conditions in winter, whilethe Hadley model suggests slightly drier conditions insummer and the Canadian model suggests that in sum-mer parts of the Caribbean will be wetter and partsdrier. For the Pacific,both models show greater warm-ing in the central and eastern equatorial Pacific and anattendant eastward shift of precipitation.

Any changes in ENSO patterns would affect precipita-tion,sea level,and the formation and behavior ofcyclones (typhoons and hurricanes),with conse-quences for both the Caribbean and Pacific islands.


This section deals with the US-affiliated islands of theCaribbean and Pacific. Included are Puerto Rico andthe US Virgin Islands in the Caribbean and theHawaiian Islands,American Samoa,the Common-wealthof the Northern Mariana Islands,Guam,the FederatedStates of Micronesia,the Republic of the MarshallIslands,and the Republic of Palau in the Pacific. Thelatter three are independent states in free associationwith the United States. Hawaii became the 50 th stateof the US in 1959. The Virgin Islands,Guam,andAmerican Samoa are US ter ritories. The NorthernMariana Islands and Puerto Rico are commonwealths.

Islands contain diverse and productive ecosystems,andinclude many specialized and unique species. Manyislands are facing the stresses of rapid human popula-tion growth,increasing vulnerability to natural disas-ters,and degradation of natural resources. Droughtsand floods are among the climate extremes of mostconcern as they af fect the quality of water supplies inisland communities and thus can have significanthealth consequences. Due to their small size and isola-tion,many islands face chronic water shortages andproblems with waste disposal. Some are facing aspecies extinction crisis; for example,the HawaiianIslands have the highest extinction rate of any state inthe nation. Biological invasion and habitat destructionseem to be the primary causes of the extinction crisis.For most island communities,transportation and othersocial infrastructure and economic activities are locat-ed near the coast,making them highly vulnerable tostorm events and sea-level fluctuations. Over-harvest-ing of reef organisms is a serious issue in parts of thePacific.

316


Key Findings

• Islands are uniquely vulnerable to many of thepotential impacts of climate change: changes inhydrologic cycles;temperature increases;and sea-level fluctuations.

• On many islands, water resources are already lim-ited. Water supplies could be adversely affectedby changes in hydrologic cycles that result inmore frequent droughts and/or floods andchanges in sea level that result in saltwater intru-sion into freshwater lenses.

• In Puerto Rico,the US Virgin Islands,and moun-tain islands of the Pacific, climate changes thatresult in increased extreme precipitation eventsare of particular concern since flooding andlandslides often result from such extreme eventstoday. Model scenarios frequently projectincreases in rainfall intensity as global tempera-tures rise.

• Both coastal and inland island populations andinfrastructure are already at risk from climateextremes. Model projections suggest that theislands could possibly be affected by changes inpatterns of natural variability in climate (such asEl Niño) and changes in the frequency, intensity,and tracks of tropical cyclones (hurricanes andtyphoons). Both the Pacific and Caribbeanregions are familiar with severe cyclones,whichhave caused billions of dollars in damage fromthe destruction of housing, agriculture, roads andbridges,and lost tourism revenue. Any increasesin frequency or intensity would have negativeimpacts on island ecosystems and economics.

• Islands are home to rare ecosystems such astropical forests,mangrove swamps,coral reefs,and seagrass beds,with many unique and special-ized endemic species (those that occur nowhereelse). These resources are already threatened byinvasive non-native plant and animal species,aswell as from the impacts of tourism,urbanexpansion,and various industrial activities, givingthe islands the highest rates of extinction of allregions of the US. Many of these ecosystems arealso highly sensitive to changes in temperatureand are likely to change or be destroyed bychanges in climate. For example,coral reefs aresensitive to changes in water temperature,andincreases in the frequency or severity of El Niño

events could result in increased bleaching andpossible die-off of corals.

• Other ecological concerns include increasedextinction rates of mountain species that havelimited opportunities for migration,and impactson forests due to floods,droughts,or increasedincidence of pests,pathogens,or fire. Anyincrease in the frequency or intensity of hurri-canes could influence forest succession,andwould generally favor invasive species. In addi-tion,the unique cloud forests located on some ofthese islands occupy a nar row geographical andclimatological niche that would be threatened byincreases in temperature and changes in thehydrologic cycle.

• Episodic variation in sea level,and the associatederosion and inundation problems are alreadyextremely important issues for many of the US-affiliated islands. Saltwater intrusion into fresh-water lenses and coastal vegetation,and coastalerosion would only become greater problemsshould climate changes result in changes to nor-mal weather patterns, changes in ocean circula-tion patterns,and increased sea-level rise.

• Tourism remains a significant contributor to theeconomies of many of these island jurisdictions.Should climate change in these regions negative-ly impact ecosystems or freshwater supplies,tourism would suffer. As has happened in thepast,tourists are likely to select alternate destina-tions to avoid the impacts of climate changesand extreme weather events.

• For island communities to become more resilientto a changing climate they may consider imple-menting any number of adaptive strategies.Some of the strategies for the islands involve:increased use of climate forecasting capabilitiesfor planning protection and mitigation measures;consideration of changing land use policies andbuilding codes;use of improved technologies;increased monitoring,scientific research,anddata collection;and improved management poli-cies. In addition,broad public awareness cam-paigns and communication with user groups areimportant for improving public health and safetyand protecting natural resources.

317


318


PHYSICAL SETTING ANDUNIQUE ATTRIBUTES

This chapter deals with the US-affiliated islands ofthe Caribbean and Pacific. Included are:theCommonwealth of Puerto Rico (PR) and the USVirgin Islands (USVI) in the Caribbean:and theHawaiian Islands,American Samoa,theCommonwealth of the Northern Mariana Islands(CNMI),Guam,the Federated States of Micronesia(FSM),the Republic of the Marshall Islands (RMI),and the Republic of Palau (RP) in the Pacific. The lat-ter three are independent nations in free associationwith the United States;Hawaii became the 50th stateof the US in 1959;the Virgin Islands,Guam,andAmerican Samoa are US territories;and Puerto Ricoand the Northern Mariana Islands are common-wealths.

Islands contain diverse and productive ecosystems,and include many specialized and unique species.Many islands are facing the stresses of rapid humanpopulation growth,increasing vulnerability to naturaldisasters,and degradation of natural resources.Droughts and floods are among the climate extremesof most concern as they affect the quality of watersupplies in island communities and thus can havesignificant health consequences. Due to their smallsize and isolation,many islands face chronic watershortages and problems with liquid and solid wastedisposal. Some are facing a species extinction crisis;for example,the Hawaiian Islands have the highestextinction rate of any state in the nation. Biologicalinvasions and habitat loss seem to be the primarycauses of the extinction crisis in Hawaii. For mostisland communities,infrastructure including,trans-portation aspects,and economic activities are con-centrated near the coast,making them highly vulner-able to storm events and sea-level fluctuations.

The islands differ in geologic type,as well as by size,elevation,soil composition,drainage characteristics,and natural resources. There are barrier islands,con-tinental islands,coral islands,and volcanic islands.Islands can also be of mixed type,such as continen-tal islands with raised reefs or volcanic islands sur-

rounded by coral reefs. Some low-lying coral islandsor atolls rise only a few feet above sea level whereassome volcanic islands have mountains thousands offeet high. Island entities also may be single islandsor groups of islands; for example,American Samoaconsists of five volcanic islands and two coral atolls,whereas Guam is a single mountain island. Theislands discussed here also vary considerably interms of the types of human communities,their eco-nomic structure,and their lifestyles and infrastruc-ture, ranging from large densely populated cities(e.g.,Honolulu,San Juan) to small villages,dispersedpopulations,and unpopulated islands (Rapaport,1999;Lobban and Schefter, 1997;Low et al.,1998;Wiley and Vilella,1998;Loope,1998).

SOCIOECONOMIC CONTEXTThe islands of the Pacific and Caribbean are host todiverse communities,many of which are especiallyvulnerable to climate variability and change. Whilethere are many commonalities among island com-munities,an appreciation of social,political,eco-nomic,and cultural diversity can help decision-mak-ers involved in the planning of appropriate strategicresponses to climatic perturbations.

American-affiliated islands in the Pacific andCaribbean share a common history of transitionfrom indigenous to colonial and,most recently, topost-colonial society;but these histories have result-ed in an array of political structures including USstate,US territory, commonwealth,and freely associ-ated state. Many of these islands have held specialsignificance for US national security interests. Onmany islands,macro-scale political systems operatein conjunction with traditional sociopolitical struc-tures. In some of the Pacific islands, for instance,the chief-based systems still influence politicalaction,particularly at the local level. These diversepolitical organizations are instrumental in the plan-ning and implementation of climate-related policiesand programs.

The US-affiliated islands of the Pacific and Caribbeanreflect considerable economic diversity whichincludes subsistence agriculture and fishing;

POTENTIAL CONSEQUENCES OF CLIMATEVARIABILITY AND CHANGE FOR THE US-AFFILIATEDISLANDS OF THE PACIFIC AND CARIBBEAN

319


commodity to be sold or traded. While the stateshave legal authority over land for “eminent domainand condemnation,” use of that power is stronglyavoided (FSM National Communication,1999).

These complex patterns of land ownership createchallenges for many climate change response (adap-tation) options. The FSM National Communicationto the United Nations Framework Convention onClimate Change (UNFCCC) notes, for example,thatthe relocation of a large proportion of the coastalpopulation to elevated lands on high islands canonly occur with the concurrence of the owners ofthose elevated lands and requires long-term land useplanning accompanied by changes in traditional pat-terns of land tenure (FSM National Communication,1999). The FSM National Communication to the

tourism;tuna processing and transshipment;andexport-led agricultural production of crops such assugar cane,bananas,pineapple,coffee,spices,andcitrus fruits. In addition,some islands are develop-ing the infrastructure to support high-tech telecom-munications and computer industries. Some islands,such as CNMI and Puerto Rico,have established sub-stantial manufacturing sectors. Despite this,many ofthe islands continue to depend on federal subsidiesand military spending. While tourism, agriculture,fishing,and transshipment are common in most allthe islands,the relative contribution of each sectorto the overall economy varies from island to island,and most islands continue to rely on a mix of tech-no-industrial and subsistence strategies. Further,many island economies are intricately linked withwider economic structures and,as such,are espe-cially vulnerable to fluctuations in the internationalmarket,which affect the prices of both importedand exported goods. Climate change and variability,whether through extreme events or gradual change,can potentially affect all island industries.

Culturally, the islands are remarkably diverse as well.Witness the strong Boriqua culture of Puerto Rico inthe Caribbean,and the Samoan and Chamorro cul-tures of the Pacific. The Federated States ofMicronesia is home to several different language andculture groups. Further, ethnic diversity is com-pounded by historical and contemporary migrationpatterns,which include not only internal migrationof indigenous island peoples,but also some migra-tion to and from the continental US and other coun-tries. Growing manufacturing industries in theCommonwealth of the Northern Mariana Islands forinstance,have attracted migrants from thePhilippines,China,and Korea.

Cultural values and belief systems can significantlyinfluence people’s responses to climate and climatechanges as well as their responses to preventive orameliorative policies and strategies. In many Pacificislands,land has inherent cultural value to the popu-lations and therefore any potential threats areregarded as especially serious. Patterns of land own-ership/tenure and resource use in many of theisland jurisdictions addressed in this Assessmentreflect a mix of traditional practices with strong andvarying cultural roots combined with more recentpolicies that derived from periods of colonial occu-pations. In the Federated States of Micronesia, forexample,land ownership remains the most valuedright, reflecting the short supply and traditionalimportance of the land and the natural resources itsupports (FSM National Communication,1999). Inthe traditional economy of the FSM,land is not a

Figure 1: The scope of this section includes the US-affiliatedislands of the Caribbean and Pacific. In the Caribbean, thisincludes Puerto Rico and the US Virgin Islands. In the Pacific, itincludes the Hawaiian Islands, American Samoa, theCommonwealth of the Northern Mariana Islands, Guam, theFederated States of Micronesia, the Republic of the MarshallIslands, and the Republic of Palau.

Hawaiian Islands

The Caribbean has a limited resource base, yet con-tains some of the most biologically productive andcomplex ecosystems found in the world. Theseinclude coral reefs,beaches,seagrass beds,mangroveforests,and coastal estuarine ecosystems. Nativetropical rainforest (including mamey, sierra palm,and tree fern) and moist forest once dominatedmuch of Puerto Rico,but much of this land hasgiven way to cultivation,pasture,and other develop-ment. Drier areas support dry grasslands and cactias well as forests of royal palm,acacia,and othertropical trees. Mangrove swamps dominate much ofthe coastline. The US Virgin islands are hilly due totheir volcanic origin with land cover consisting pri-marily of cactus, grasses,and sparse tropical wood-land and shrubland. Some native vegetation hasbeen lost to stock grazing and the cultivation offruit and vegetable crops and sugarcane.

The native vegetation cover of the Hawaiian islandsincludes tropical coastal and lowland shrublands,grasslands, wet forests,tropical montane rain forests,and shrublands (Loope,1998). Much of the coastaland lowland vegetation,and increasingly, the interiormontane vegetation,have been lost to urban, recre-ational,and agricultural development (grazing,sugar-cane,and pineapple plantations),and to introducedspecies.

American Samoa features rain forests in the interiormountains. Much of the coastal zone is under culti-vation for coconut,taro,and other crops. Guam’sprincipal land cover consists of tropical shrublandand limestone forest in the north and sword grassand dense jungles along river valleys in the volcanicsouth. The Marianas are not considered to be tropi-cal rain forest. The flora consist of vines,shrubs,ferns,and grasses,including savanna and trees. Themore common trees include:coconut, flame tree,Formosan koa,ironwood (Casuarina),banyan,papaya,tangan,mangrove and a few other varieties.In general,the variety of botanical species is limitedon the CNMI and there are no vegetation zones(CNMI website,2000).

Over 97% of US coral reef resources are in the terri-torial waters of Hawaii and the US territories andcommonwealths. Hawaii has the widest range ofreef habitats among the islands discussed in thisreport,with a significant portion of US reefs in thenorthwest Hawaiian Islands. The US territories,com-monwealths,and freely associated states,because oftheir location in tropical waters,have extensive reefhabitats that are important culturally, economically,and biologically. For example,the abundant andwide variety of sea life within CNMI waters includes

UNFCCC also highlights the fact that such chal-lenges are even greater when considering off-island re-settlement in response to sea-level riseimpacts on low-lying atolls. The FSM NationalCommunication notes that pursuit of such anoption would “have to be achieved without dis-ruption to host communities and with sensitivityto and careful regard for the traditional values andpractices of both the displaced and host communi-ties”(FSM National Communication,1999).

Demographically, Pacific and Caribbean islandsinclude densely populated urban centers such asHonolulu,Hawaii and San Juan,Puerto Rico tosparsely populated outer islands of the FederatedStates of Micronesia. Some of the highest popula-tion densities in the Pacific are found on Majuro andEbeye in the Marshall Islands. While future popula-tion growth rates in the Caribbean appear to beattenuated,some Pacific island populations aregrowing rapidly. For example,from 1982 to 1998,the population of Majuro virtually doubled fromabout 15,000 to over 30,000. Pacific Islanders rep-resent one of the fastest growing groups of migrantsto the mainland United States. As the impact of cli -mate change is felt more strongly, this trend is likelyto continue.

ECOLOGICAL CONTEXTThe ecosystems of the islands discussed in thischapter can be characterized as tropical, rich in bio-logical uniqueness,and susceptible to disturbanceand biological invasions.1 The isolation of islandshas resulted in the evolution of unique floras andfaunas with large numbers of endemic species —those found nowhere else on Earth (Low et al.,1998;Wiley and Vilella,1998;Loope,1998). Islandsin both the Pacific and Caribbean are facing anextinction crisis as a result of steady habitat destruc-tion,and competition and predation by introducedspecies. Small population sizes and lack of room formigration exacerbates the vulnerability of islandspecies. The situation on many islands is becomingcritical as the area of undisturbed natural habitatdiminishes (Low et al.,1998;Wiley and Vilella,1998;Loope,1998). Fortunately, unique terrestrial andmarine resources are still found in the tropical USon islands which have few or no human inhabitantssuch as Baker and Howland Islands, Jarvis Island,theJohnston Atoll,and the Kingman Reef among others(OIA,1999).

320


1 This section outlines information on current island ecosystems andbiodiversity. The VEMAP modeling system providing information aboutpotential future changes to vegetation and ecosystems was not run forthe islands as it was for the conterminous US.

321


US Affiliated Islands of the Caribbean and Pacific

Name/ Geologic Type Location/Description Current Major Economic SectorsUS Affiliation Population

(Approx.)

Commonwealth One volcanic island composed • Located SE of Miami on the 4 million. Manufacturingof Puerto Rico of uplifted sedimentary rocks. northern edge of the archipelago Services

of islands on the Caribbean plate. Government• 3,425 sq mi;8,871 sq km Agriculture• Abundant rain in the North;dry Tourismclimate in the South.

US Virgin Three volcanic islands (St. • Located 1,000 miles SE of Miami 120,000 Tourism Islands Croix,St. John,and St.Thomas) on the northern edge of the Government

composed of uplifted archipelago of islands on the Territory sedimentary rocks. Caribbean plate.

• 133 sq.mi.;344 sq.km• No lakes, rivers or streams.

Hawaiian Islands Eight volcanic islands (Kauai, • Located 2,400 miles SW of 1.2 million TourismOahu,Molokai,Lanai,Maui, California and on the Tropic AgricultureKahoolawe,Niihau and of Cancer Military spending

State Hawaii). Also consists of NW • 6,450 sq.mi.;16,706 sq.km. ManufacturingHawaiian Islands and Johnston • Sea area:833,171 sq.mi. GovernmentAtoll. • Majority of US coral reefs.

American Five volcanic islands and two • Located 2,700 miles NE Australia 60,000 U.S. federal expendituresSamoa coral atolls (Ofu,Ta’u,Swains • 76 sq.mi.;197 sq.km. Canning industry

Island,Tutuila,Olosega,Rose • Rainforests in interior; Small tourism industryTerritory Island,and Aunu’u) mountainous regions Government

Commonwealth Fourteen volcanic islands • Located at the edge of the 66,000 Tourismof the Northern Philippine plate. Garment manufacturingMariana Islands • 185 sq.mi.;479 sq.km. Government

• Formed by underwater volcanoes Fish transshipmentalong Mariana Trench.

Republic of the 29 coral atolls (each made up • Located about 2,136 miles SW of 60,000 US aidMarshall Islands of many islets) and 5 low-lying Hawaii Tourism

volcanic islands • 70 sq.mi.;181 sq.km. FisheriesFreely • Each atoll is a cluster of many Subsistence AgricultureAssociated State small islands encircling a lagoon Craft items (wood carvings,

• None of these islands is more than shell and woven baskets)a few meters above sea level.

Republic of Several hundred volcanic • Located in the North Pacific 17,000 TourismPalau islands and a few coral atolls Ocean,SE of the Philippines. Craft items (wood,shell pearl)

(only 8 of the islands are Commercial fishingFreely inhabited). • 192 sq.mi.;497 sq.km. Agriculture (subsistence)Associated State • The 8 islands that are inhabited Government

include Kayangel Island,Babelthaup,Urukthapel, Koror,Peleliu,Eli Malk,Angaur, Sonsoral and Tobi.

Federated States A group of 607 small islands • Located in the Western Pacific 127,000 Tuna fishingof Micronesia consisting of volcanic islands about 2,500 miles SW of Hawaii. Tourism

and coral atolls. • 271 sq.mi.;702 sq.km GovernmentFreely • 4 states:Chuuk, Pohnpei,Yap and SubsistenceAssociated State Kosrae. Craft items (wood carvings,

shell and woven baskets)

Guam One volcanic island composed • Guam is the largest Micronesian 150,000 Tourismof uplifted sedimentary rocks. Island. Military/government

Territory • 209 sq.mi.;541 sq. km. Tuna and other cargo• Formed by the union of two transshipmentvolcanoes,northern Guam is a flat Airline hublimestone plateau while the southern part is mountainous.

322


coral gardens that harbor myriad mollusks,plant life,and tropical fish. The near-shore reef and lagoonareas are important for subsistence fishing,particu-larly net throwing. Many sea cucumbers occur inthe shallow lagoon waters (CNMI website,2000).

The Freely Associated States contain a wide range ofterrestrial and marine habitats,such as coral reefsand mangrove forests. The Exclusive EconomicZone of the Federated States of Micronesia (FSM)covers over one million square miles (2.6 million sq.km) of ocean and contains the world’s most produc-tive tuna fishing grounds. The estimated marketvalue of tuna caught within FSM waters is valued atover $200 million per year. Pohnpei State,the loca-tion of the FSM’s capital, receives an average of over180 inches of rain per year and is covered by lushrainforests,and surrounded on the southern sidewith some of the most extensive mangrove wet-lands in the Pacific Islands region. The rock islandsin the Republic of Palau are uplifted fossil coralreefs that have been rounded by erosion. Coral reefecosystems in Palau,which suffered from extensivebleaching during the 1997-1998 El Niño,are regard-ed as among the best diving areas in the world.

Unique terrestrial and marine resources of the tropi-cal US are found in insular areas in the Pacific whichhave few, if any human inhabitants. These includeBaker and Howland islands which are nationalwildlife refuges (Howland:birds and marine life;Baker:marine life). Jarvis Island,another nationalwildlife refuge,serves as habitat for seabirds andshorebirds. Johnston Atoll is home to 194 species ofinshore fishes and sea turtles,and is a site for roost-ing and nesting of 500,000 seabirds. Kingman Reefhas a rich marine fauna and the Northwest HawaiianIslands National Wildlife Refuge includes over 100islands with an estimated 616,000 pairs of Laysanalbatrosses and 21 other species of seabirds (Low etal.,1998;Harrison,1990). The CNMI islands ofAsuncion,Guguan,Maug,Managaha,Sarigan,andUracas (Farallon De Pajaros) are maintained as unin-habited places. No permanent structures may bebuilt and no persons may live on the islands exceptas necessary for the purposes for which the islandsare preserved. The islands are preserved as habitatsfor birds, fish,wildlife,and plants. The three islandsthat are collectively known as Maug have beengiven permanent status as wildlife preserves (CNMIwebsite,2000).

CLIMATE VARIABILITY AND CHANGE The islands of the Pacific and Caribbean experiencerelatively high air temperatures and low seasonalvariations in air temperature throughout the yearcompared to most of the US mainland. However,other climate variables exhibit distinct seasonal pat-terns,particularly rainfall distribution,which resultsin wet and dry seasons. For example,precipitationpatterns differ greatly over the Pacific,with a pro-nounced winter rainfall maximum in Hawaii and alate-summer/early fall maximum in Guam. In theCaribbean,the wet season is during the summer. Inthe Pacific,tropical storms and typhoons are com-mon between May and December, although intensetropical cyclones can affect Guam and the CNMIduring any month of the year. In the Caribbean,thehurricane season extends from June to November.

In addition to the seasonal variations,there arestrong year-to-year fluctuations. For example,the ElNiño/Southern Oscillation (ENSO) causes fluctua-tions in sea level, rainfall,and cyclone activity (hurri-canes or typhoons,depending on the region) inboth the Pacific and Caribbean islands. In theCaribbean,Atlantic hurricanes are suppressed duringEl Niño events,while their numbers increase duringLa Niña events. In the Pacific,during El Niñoevents,Hawaii,Micronesia,and the islands of thesouthwest tropical Pacific often receive below nor-mal rainfall. Despite this,El Niño can alsoincrease the risk of intense tropical cyclones forHawaii,American Samoa,and the eastern half ofMicronesia. Additionally, areas of above normalprecipitation,along with greater tropical cycloneactivity, typically shift eastward towards FrenchPolynesia. During La Niña events,parts ofMicronesia can receive very heavy rainfall duringwhat is normally their dry season.


Over the past century, average annual air tempera-tures in the Caribbean islands have increased bymore than 1°F (0.5 °C),while average annual airtemperatures in the Pacific Islands have increasedby about half this amount. Globally, sea level hasrisen by 4 to 8 inches (10-20 cm) in the past 100years but with significant local variation. Relativesea level, which takes into account natural orhuman-caused changes in the land elevation such astectonic uplifting and land subsidence (sinking),is

323


Africa and propagate into the tropical Atlantic.These easterly waves are responsible for hurricaneactivity in the Caribbean. Therefore, changes in THChelp regulate hurricane occurrence in the Atlantic.

Hurricane activity in the western Atlantic is reducedduring the season following the onset of an El Niñoevent in the Pacific. The reason for El Niño occur-rence leading to decreased incidence of hurricanesis because of the presence of anomalously strongwesterly winds in the upper troposphere over theCaribbean and equatorial Atlantic (Gray, 1984). Thestrengthened winds result from extra-deep cumulusconvection in the eastern Pacific Ocean that occursduring El Niño events. Hurricanes in the Atlantictend to return to normal in the second summer fol-lowing such an event. In spite of the relationshipbetween hurricanes and El Niño events,there is lit-tle net effect on the precipitation in the region.

Scenarios for Future Climate

Model projections suggest it is possible that thePacific and Caribbean islands will be affected by:changes in patterns of natural climate variability(such as El Niño); changes in the frequency, intensi-ty, and tracks of tropical cyclones (hurricanes andtyphoons);and changes in patterns of ocean circula-tion. These islands are also very likely to experience

showing an upward trend at sites monitored in theCaribbean and Gulf of Mexico. The current rate ofrelative sea-level increase in the Caribbean and Gulfof Mexico is about 3.9 inches (10 cm) per 100years. While absolute sea level is also rising in thePacific, geologic uplift is exceeding or keeping paceon many islands. As a result,across the set ofislands,there is no consistent sea-level trend for thePacific. Although showing no average net rise, rela-tive sea level fluctuates widely, due to effects suchas the ENSO cycle.

The highest frequency of cyclones in the Caribbeanoccurs in the western Bahamas and along the USeastern seaboard and adjacent Atlantic Ocean. Therehas been considerable decadal variability of hurri-canes during this century (see Reading,1990;Diazand Pulwarty, 1997;Landsea et al.,1995). Amongother factors,decadal variability is affected by ENSOas well as changes in the thermohaline circulation(THC),the large-scale oceanic circulation in theAtlantic that includes the Gulf Stream. For example,Gray et al.(1997),show that a weakening of theTHC in the latter half of this century resulted incooling of the Atlantic Ocean along the coast ofnorthwest Africa,an increase in the low-level pres-sure gradient between this region and the centralSahel,and a resulting reduction in the strength ofthe easterly waves that originate off the coast of

The El Niño/Southern Oscillation (ENSO) Phenomenon

The term“El Niño” refers to the warm phase of the ENSO cycle when ocean temperatures in the central andeastern tropical Pacific tend to increase. The term “La Niña” refers to a phase of the ENSO cycle when watersin the central and eastern Pacific tend to be cooler. The ENSO cycle,normally about one year in duration,is acontinuously evolving pattern of ocean-atmosphere behavior and represents one of the dominant patterns ofnatural variability in the climate system worldwide. Its frequency of occurrence is approximately every 2 - 7years.

El Niño is thus one phase of the ocean half of a coupled ocean-atmosphere phenomena in the tropical Pacific(see Pielke and Landsea,1999). Its atmospheric partner is known as the Southern Oscillation,a periodic oscil-lation of atmospheric pressure between the western and eastern Pacific Ocean. Under normal,non-El Niñoconditions,atmospheric pressure is lower over the western Pacific and higher over the eastern Pacific. Thispressure gradient helps drive the (surface) trade winds from east to west or from high pressure to low pres-sure and results in a tendency to maintain higher sea levels in the western Pacific Ocean than in the east.During an El Niño event,the atmospheric pressure gradient lessens, resulting in surface winds that diminishor, sometimes, reverse direction. The see-saw in atmospheric pressure,coupled with the eastward movementof warm water, results in a drop in sea level in the western Pacific Ocean and a migration of higher than nor-mal sea-level conditions to the east (along with the warm water due to reduced upwelling). In fact,this east-ward migration of sea level is one of the observational signals that an El Niño event is underway. Images of ElNiño events from satellites like TOPEX-POSEIDON show these changes in sea level as changes in the height ofthe ocean’s surface. When this occurs,islands in the far western Pacific experience lower sea level whilethose in the central and eastern Pacific experience a transient increase in sea level.

increased ocean temperatures and changes in sealevel (including storm surges and sustained rise).

For the Caribbean,most model simulations projectincreasing air and water temperatures,slightly wet-ter conditions in winter, and slightly drier conditionsin summer. Models that suggest more persistent ElNiño-like conditions across the Pacific imply areduction in Atlantic hurricane frequency in thefuture. However, not all models concur on thispoint.

324


For the Pacific,the model simulations project agradual increase of air and water temperatures.Some recent climate model studies also project thatENSO extremes are likely to increase with increas-ing greenhouse gas concentrations (Timmermannet al.,1999;Collins,2000).Additionally, recentmodel results (Knutson and Manabe,1998;Timmermann et al.,1999) have agreed with earlierstudies (Meehl and Washington,1996) in showingthat as global temperatures rise due to increasedgreenhouse gases,the mean Pacific climate tends tomore resemble an El Niño-like state,with greaterrelative surface warming in the equatorial easternPacific than the west. This implies a reduction offresh water resources in areas of the westernPacific,Micronesia,and southwest tropical Pacific(Meehl,1996). The existence of higher sea surfacetemperatures in the central and eastern Pacificcould also result in an expansion in the area of thePacific where tropical cyclones will form andmigrate,as is the case during an El Niño. Othermodels show uniform warming across the Pacificor somewhat greater warming in the western equa-torial Pacific.

Apart from the linkage with ENSO, there is signifi-cant uncertainty about how increasing global tem-peratures will affect hurricane or typhoon frequen-cy and preferred tracks in both the Caribbean andthe Pacific regions (see Landsea,et al.,1996;Henderson-Sellers,et al.,1998;Emanuel,1997,1999;Meehl,et al.,2000). Studying the effect of globalwarming at a time of doubled CO2, Bengtsson et al.(1997) conclude that there are significantly fewerhurricanes (especially in the SouthernHemisphere),but no significant change in the spa-tial distribution of storms. The model used byBengtsson,et al.(1997),however, was barely ade-quate to resolve tropical cyclones. More ocean sur-face warming could create more opportunity forstorm development, given the right atmosphericconditions. Other results,using a nested high-reso-lution model (resolution of up to 1/6 degree or 18km) indicate a 5-11% increase in surface winds anda 28% increase in near-storm precipitation(Knutson et al.,1998;Knutson and Tuleya,2000).Another recent overall assessment regardingchanges in hurricane strength suggests a 5-10%increase in tropical storm wind speed by the endof the 21st century (Henderson-Sellers,et al.,1998).

It is also a fact,however, that until capabilities aremore developed for accurately simulating ENSOand changes in the thermohaline circulation inglobal climate models,the possibility of significantchanges in the expected frequency, intensity, and

Figure 2. These model projections suggest stronger and more fre-quent El Niños and La Niñas as a result of climate change. SeaSurface temperature anomalies (SSTA) in the equatorial Pacific areused to measure the strength of El Niños and La Niñas. Thesemodel projections by the Max Planck Institute suggest a widerrange of SST deviations from normal and thus more extreme ElNiños and La Niñas in the future. The high bars in the center areoccurrences of normal SSTs. In the projections in the bottomgraph, these normal temperatures occur less frequently, whilelower (La Niña) and higher (El Niño) SSTs occur more frequently.The Max Planck model is used here because it has been able toreproduce the strength of these events better than other modelsdue to its physics and ability to resolve fine scale structure in theocean. Source: Timmermann et al., 1999

El Niños and La Niñas as a Function of Observed andProjected Sea Surface Temperatures

duration of ENSO as well as tropical storms cannotbe ruled out. It is even conceivable that changes inthe strength of monsoon circulations in the SouthPacific could have unanticipated effects on ENSO andrelated tropical storm frequency, paths,and strength.

KEY ISSUES

Two regional workshops were held to identify stake-holders’concerns regarding climate change and cli-mate variability in the context of other current stress-es for island and coastal environments and communi-ties. The first workshop was held in Honolulu,Hawaii in March 1998,and participants identifiedissues of importance to the US-affiliated islands in thePacific. The second workshop took place in July1998 and addressed concerns of the south Atlanticcoast of the US and the US-affiliated Caribbeanislands. This chapter is based on presentations anddiscussions at these workshops, extant literature,andongoing research into the consequences of climatevariability and change to islands.

While the workshop participants identified numerousissues for consideration, four of the critical climate-related issues for the islands now and in the futureinclude:• ensuring adequate freshwater supplies• protecting public safety and infrastructure from

climate extremes• protecting rare and unique ecosystems• responding to sea-level fluctuations.

It should be noted that participants at the regionalworkshops did not identify long-term sea-level rise asthe highest priority issue, focusing instead on waterresource issues and storm impacts. There is a sensethat the nature of the sea level issue for islands hasbeen somewhat misunderstood in discussions of cli-mate change. It is important to recognize thatepisodic variations in sea-level (like those associatedwith the ENSO cycle in the Pacific) and storm surgesare already extremely important issues. In addition toconsidering only the consequences of a gradual,long-term rise in sea level,island communities will contin-ue to face the consequences of episodic events (e.g.,extreme lunar tides,ENSO related changes,andstorm-related wave conditions) some of which them-selves will be af fected by climate change.Importantly, changes in sea level associated with natu -ral variability exceed long-term projections of sea-level rise in some jurisdictions.

Many of the projected consequences of long-termsea-level rise,such as salt water intrusion into fresh-

325


water lenses and coastal erosion are already prob-lems in some if not most island jurisdictions.According to USGS (1999a), for example, about 25%of the sand beaches on Oahu,Hawaii have been lostor severely degraded during the last 60 years.Climate-related changes of these conditions areseen,therefore,as magnifying existing problemsrather than as problems in isolation from otherstresses. Island communities must deal with theseproblems today and,in so doing,can develop impor-tant insights into how they might most effectivelyrespond to climate-related changes in sea level overboth the short- and long-term. Because the commu-nities will act to mitigate the increased climatechange problems periodically, they will not beadjusting to 100 years of climate change effects atone time,but rather to smaller effects more fre-quently during the centur y. This allows for stag-gered adaptation.

1. Freshwater Resources

The availability of freshwater resources is already aproblem for island communities,as a result of theirunique geography and the growth of population,tourism,and urban centers. Many islands sufferfrom frequent droughts and chronic water scarcitydue to lack of adequate precipitation. In othercases, rainfall is abundant but access to freshwater islimited by lack of adequate storage facilities anddelivery systems,or a geographic mismatch betweenthe source and the site of the need. For the PacificIslands,both a present and future key issue isdrought or water scarcity conditions. For theCaribbean,drought is also a problem,but floods andlandslides associated with hurricanes and heavyrainfall are an important addition to this region’s listof key water-related issues. According to Larsen andTorres-Sanchez (1998) rainfall-triggered landslidesare the most common type of landslide in the cen-tral mountains and foothills of Puerto Rico.

Future climate changes in the islands regions couldinclude: changes in natural variability in weatherpatterns,ocean temperature,and currents (e.g.,pat-terns of El Niño); changes in the frequency, intensity,and tracks of tropical cyclones (hurricanes andtyphoons) and their resulting precipitation;and/orchanges in sea level. Any of these changes wouldimpact the amount,timing,or availability of freshwa-ter, such that freshwater issues will be increasinglyserious concerns for the US affiliated Islands.

On the islands,demands for water are from multiplesources including agriculture,industry (includingfish processing),households,and natural ecosys-

326


tems. Tourism is extremely water-intensive and is amajor industry for the Commonwealth of theNorthern Mariana Islands,Guam,Hawaii, Palau,Puerto Rico,and the Virgin Islands. Tourism devel-opment continues in these areas and the FederatedStates of Micronesia,the Republic of the MarshallIslands,and American Samoa all see tourism as oneof their best options for economic development.

Many island communities rely upon groundwaterresources called freshwater lenses. The size of thegroundwater lens is directly related to the size ofthe island as well as incident rainfall,island rechargecharacteristics (controlled by land cover, evapora-tion,and geology),and the lateral leaking of lenswater into the ocean. Larger islands have largerlenses and are more buffered from drought condi-tions while smaller islands have no lenses or shal-low lenses that easily become depleted or contami-nated with salt water. During drought conditions,there is no recharge to the lens,and the fresh wateris depleted rapidly, especially if consumption ishigh. Low sea levels associated with El Niño lowerthe water table even more,making it more difficultto access the water and easier to damage the fragileinterface between the fresh water lens and theunderlying salt water. Low-lying atolls rely heavilyon rainwater collection and are therefore leastbuffered against drought conditions.

The geology of the island controls the amount,quality, and seasonal availability of the water sup-ply. Islands that are continental or of old volcanicorigin exhibit complex movement of groundwaterthrough layers of rock or clay. New volcanicislands generally have more dynamic movement ofwater through tunnels,holes,and layers of ash.Similarly, islands made up of coral or sand arequite permeable,so the rainwater infiltrates veryquickly. Water quality is also an issue:many vol-canic islands have highly permeable rock,whichincreases the potential for groundwater contami-nation and in some Pacific islands contaminationproblems reduce the “capacity”of the system(USGS Hawaii State Fact Sheet,1996). In addition,seawater can contaminate the freshwater lenswith salt if sea level rises,the lens is depleted,orthe delicate freshwater-saltwater interface is dam-aged by excessive extraction.

Patterns of precipitation are important in deter-mining whether islands have an adequate freshwa-ter supply. Long periods of rainfall are needed torecharge the freshwater lenses since short andlight rainfall tends not to contribute to aquiferrecharge. Land use is an important determinant ofinfiltration;how much water infiltrates into theground or flows into rivers and streams dependsupon the land cover that catches the rain. If theland is covered by forest,the forest holds the rain-water for drier periods,but if the forest has beendisplaced by urban development, for example,therain runs off faster leaving less for use during dryconditions. On some islands,destruction of forestcover has caused many formerly perennial streamsto stop flowing in the dry seasons and has con-tributed to landslides during periods of heavy rain.

Pacific Islands. Rainfall,stream flow, and ground-water are fairly abundant on the big island ofHawaii,but many areas are withdrawing water atrates close to the estimated yield of aquifers nearthe populated areas (USGS Hawaii Fact Sheet,1999a). Water use in many of the other PacificIslands is now near the limit of the known fresh-water resources,and demand exceeds the existingsupply on many islands. In some cases,the prob-lem is not the lack of adequate rainfall,but distri-bution,storage,and maintenance. Guam and theisland of Hawaii are examples of the distributionproblem due to their “geographic mismatch”between supply and demand. For example,the“Hilo side”of the big island of Hawaii consistentlyreceives abundant rainfall and is one of thewettest places in the Hawaiian Islands,while the“Kona side”is where the majority of tourism devel-

Figure 3. On many islands, the underground pool of freshwater thattakes the shape of a lens is a critical water source. The freshwaterlens floats atop salt water. If sea level increases, and/or if the lensbecomes depleted because of excess withdrawals, salt water fromthe sea can intrude, making the water unsuitable for many uses.The size of the lens is directly related to the size of the island: larg-er islands have lenses that are less vulnerable to tidal mixing andhave enough storage for withdrawals. Smaller island freshwaterlenses shrink during prolonged periods of low rainfall, and waterquality is easily impaired by mixing with salt water. Short and lightrainfall contributes little to recharge of these sources. Long periodsof rainfall are needed to provide adequate recharge. Source:Illustration by Melody Warford.

Freshwater Lens Effect in Island Hydrology

opment is occurring and has experienced significantperiods of drought. Other islands,that more easilyaccess their adequate rainfall, face storage,distribu-tion,and maintenance problems. Pohnpei normallyhas an abundance of water, including waterfalls,springs,and rivers in the central areas. However, theisland lacks storage and distribution systems,causingsevere impacts under drought conditions.

Water system development in American Samoa,Guam,and Hawaii has thus far kept pace withgrowth and almost all residents have access to reli -able supplies of quality water. On Saipan,theCommonwealth Utilities Corporation has struggledfor decades to make improvements in an inadequatewater distribution system and to keep pace withpopulation growth and tourism development. OnPohnpei,despite its abundant rainfall,the publicwater system in Kolonia,the largest town,cannotprovide an adequate supply of quality water to resi-dents under “normal”circumstances. In rural areas,people take water directly from rivers and streams,many of which dry up during severe droughts,andwhile groundwater is available,there is often no sys-tem to access it. In the outer islands of the MarshallIslands and Federated States of Micronesia, residentsrely on small rooftop catchment systems with only afew thousand gallons of storage capacity. On Majuro,the public water system is fed primarily from theinternational airport runway and that system hasbeen supplemented by wells on Laura Islet,some 17miles from the storage system at the airport.Colonia,Yap (FSM) depends on a small reservoir forwater, but the reservoir, even if full at the beginningof a severe drought,dries up.

Currently the Pacific Islands experience drought con-ditions every two to six years,usually associated withthe El Niño/Southern Oscillation (ENSO) cycle. Themost extreme droughts appear to be associated withvery strong El Niño events like those that occurredin 1982-1983 and 1997-1998. Rainfall during the1997-1998 event in some areas was well below nor-mal for the period October 1997 through September1998. In some months, rainfall was as low as 3% ofnormal. These low monthly totals meant trouble forislands relying on catchment systems.

The water resources for many of the Pacific Islandscould be most adversely affected by any increase inEl Niño or El Niño-like conditions. These increasescould take many forms such as more frequent and/orsevere El Niño events,medium-sized El Niños thatpersist for multi-year periods,or other conditionsthat diminish precipitation over these islands,such asthe “El Niño-like conditions”hypothesized to result

from increased ocean temperatures. Due to overlap-ping time-scales,these changes could result inalmost chronic drought conditions for some islandsin the Pacific. Such a sequence would not allowislands critical recovery time from one droughtbefore another sets in,or would stress their emer-gency preparedness capabilities if they were facedwith an additional natural disaster on top of any pro-tracted drought condition. A secondary conse-quence is migration:if such enhanced drought con-ditions occur over wide regions of the tropicalPacific,atoll populations may migrate in significantnumbers to high-island population centers withmore stable freshwater resources (Meehl,1996).

The Caribbean. In Puerto Rico and the US VirginIslands,high population densities and the conversionof tropical forest to other uses have affected hydrolo-gy and water resources. These changes have con-tributed to overuse of existing water supply (evenwith reduced per capita use due to conservationprograms,USGS,1999b),in-filling of public-supplyreservoirs with sediment,and contamination ofgroundwater and surface water (Zack and Larsen,1993).In the US Virgin Islands,and to a lesser extentin Puerto Rico,leaky septic tanks and inadequatesewage treatment facilities have degraded the qualityof near-surface groundwater supplies (Zack andLarsen,1993). In addition,both Puerto Rico and theUS Virgin Islands experience flash flooding and land-slides that result from the combination of steepslopes and heavy rainfall typical of this region.Deforestation and development have contributed toincreased flash flooding and landslide potential par-ticularly under steep,denuded slopes. Poor drainageon floodplains has increased the vulnerability ofthese areas to various flooding events,while land-usepatterns and inadequate construction increase theexposure of the population (Zack and Larsen,1993).

In the Caribbean,tropical storms/hurricanes are sig-nificant natural hazards,through the mechanisms ofstorm surge and high winds. In coastal locations ofPuerto Rico and the US Virgin Islands,saltwaterintrusion threatens the continued use of freshgroundwater and limits groundwater withdrawals.

Puerto Rico has abundant ground-and surface-waterresources due to relatively heavy rainfall over themountainous interior of the island and receptive,sedimentary rocks around the island’s periphery(Zack and Larsen,1993). These form an extensiveartesian aquifer system on the north coast. Water-table aquifers overlie the north coast artesian aquiferand occur along most of Puerto Rico’s coastline.Artificial reservoirs on principal water courses col-

327


328


Responding to Drought in the Pacific Islands

The 1997-1998 El Niño event offers a vivid example of how forecasts and information about potential climatephenomena and their consequences can be used to support decision making and planning that benefit socie-ty. In September 1997,the Pacific ENSO Applications Center (PEAC) predicted a near-record drought forMicronesia,beginning in the November-December timeframe and ending in the May-June timeframe.

When severe drought was projected for the Northern Marianas (where there are not regular sources of goodwater at any time) beginning in February 1998,the government implemented a tightened water-rationingschedule. Each venue was assigned “water hours”of 2 - 4 hours each day. By February, El Niño had alreadyreduced the island’s rainfall level to 75% below normal,and the shortage was to continue for many months.The Commonwealth Utilities Corporation,the government entity responsible for the island’s water system,also outlined a “severe drought scenario”in which commercial water users,such as hotels and garment facto-ries, would be cut off from the island’s water system for a few hours each day. A task force was establishedwhich pushed for legislation requiring every home and building to install a rainwater catchment system (thiswould,however, require loans or subsidies for low-income families).

For Majuro,PEAC provided a drought and typhoon forecast in September 1997 that indicated an increasedtyphoon risk and a sharp decline in rainfall beginning in December 1997. In November, the Government ofthe Republic of the Marshall Islands convened a task force to develop a drought response plan. Typhoon Pakastruck the Marshall Islands in December 1997 and brought the last significant rainfall until June 1998. Thegovernment requested assistance from the US Commander-in-Chief, Pacific Command to repair pumps at thewells on Laura Islet. The water utility on Majuro developed a conservation plan,a public education program,and imposed water hours. At the height of the drought, residents of Majuro were getting 7 hours of waterevery 15 days. The Japanese government provided two 20,000 gallon per day (gpd) and one 16,000 gpdreverse osmosis (RO) units in February 1998. Following the US Presidential disaster declaration in March,sixlarger RO units were provided with funding from the US Federal Emergency Management Agency in April. InMay 1998,the new pumps at Laura came on line and in June rainfall increased and the RO units were shutdown.

The drought in Samoa was delayed until April because of its location in the Southern Hemisphere. For theSamoas,PEAC anticipated an increase in the risk of tropical cyclones and drought as a result of the El Niño.

In April,May, and June of 1997, rainfall in PagoPago was only 64% of normal. In July 1997,itwas only 59% of normal. Their summer wascloser to normal,but PEAC forecasted that rain-fall for the period April through August 1998would be well below normal and this wasborne out. The American Samoa governmentorganized a drought response that includedsuch activities as establishing a drought taskforce and a public information campaign andhas continued to rely on the PEAC andNational Weather Service forecasts.

In May of 1998,American Samoa’s single largestprivate employer, StarKist Samoa committed toimplementing several in-house conservation

El Niño billboard used as a public information tool in the PacificIslands - US National Weather Service, Pacific Region Office.

329


measures which would eventually reduce the canneries’daily water usage by 25%. The company had also uti-lized scenario-based planning,and acknowledged that the worst case would involve importing water from off-island by tanker in the amount of 50 million gallons.

Agriculture throughout the US-affiliated Pacific Islands during 1997-1998 suf fered everywhere from thedroughts, except on Guam. There, farmers used public water for irrigation,and the delay of heavy rainstoward the end of the drought resulted in one of the most productive harvests in recent histor y. In theCommonwealth of the Northern Mariana Islands (CNMI),citrus and garden crops were most affected by thedrought,and the hospital had to buy imported fruits and vegetables rather than rely on local suppliers. A lim-ited damage assessment was done on Pohnpei and serious losses of both food and cash crops were sustained.Over half the banana trees evaluated had died or were considered seriously stressed. The loss of sakau (kava)was probably the most serious economic loss because it had become a major cash crop for Pohnpei. On Yap,taro losses were estimated at 50-65%,and betel nut prices increased more than 500%,although only 15-20% ofthe trees were lost. In Palau,imported food shipments increased from twice a month to once a week.Otherclimate-related consequences felt throughout the islands included changes in the migratory patterns of eco-nomically-significant fish stocks like tuna;stresses on coral reefs associated with increased water temperature;and increased sedimentation from erosion;and reduced local air quality in areas affected by wildfires. Guam,for example, experienced some 1400 fires and grass fire burned 20% of Pohnpei.

Relief food was supplied beginning in May 1998 to islands in the Marshalls and was also supplied to the outerislands of Pohnpei,Chuuk,and Yap. The Pohnpei Agriculture and Trade School,the US Natural ResourcesConservation Service,and Pohnpei State Agriculture Department provided assistance to the farmers as thedrought began to subside.Subsistence crops required 6-10 months for rejuvenation after the rains returned.The disaster management offices used PEAC and National Weather Service forecasts to plan food relief andreplanting programs for the droughts in the Republic of the Marshall Islands (RMI) and the Federated Statesof Micronesia (FSM).

Overall,islands in the Pacific implemented many of the following measures when faced with drought condi-tions in 1997 and 1998:

• expanded public information and education efforts;• continued emphasis on and funding for household rainwater collection systems;• accelerated drought planning efforts;• expanded drilling efforts and other activation of new water sources,such as the emergency use of moni-

toring wells to augment production;• tightened pre-existing water restrictions and delivery scheduling schemes;• more use of water-haulers to fill existing storage facilities;• shifted farm location and range management (where livestock were pastured) to take advantage of more

reliable, groundwater-based irrigation supplies;• increased attention to water conservation, for example in concentrated leak detection and repair efforts;• more extensive and effective use of ENSO forecast information;• accelerated maintenance on water pumps and lines - acquired spare parts;and• installed RO units.

Many of these improvements have long lasting positive effects – such as repair of pipes and pumps andincreased storage capacity, alleviating the need for conservation until the next drought.

Although Puerto Rico had seen lower mean annualrainfall in previous decades,the population growthchanged the nature of the needed response. Publicworks projects were initiated including a $60 mil-lion project to dredge the sediment from the Loizareservoir and $350 million for a new aqueduct sys-tem to interconnect north coast water supply sys-tems (Larsen,2000). Diminished streamflow hasalso severely affected biotic communities andecosystem processes in stream channels in PuertoRico (Covich et al.,1998).

Adaptation OptionsStrategies for providing adequate water resourcesunder changing climate conditions for island com-munities vary. Options could include:improvedrainfall catchment systems;improved storage anddistribution systems;development of under-utilizedor alternative sources;better management of supplyand infrastructure;increased water conservationprograms;construction of groundwater rechargebasins for runoff;more effective use of ENSO fore-cast information;and application of new/improvedtechnology, such as desalinization. It would also beuseful for island communities to prepare waterneeds assessments for the future and prepare,test,and improve emergency/contingency plans for peri-ods of water shortage.

For key sectors,strategies include integration of cli-mate considerations and projections of sea-level riseinto community planning, water management deci-sion making,and tourism development. Accurateassessments of current water budgets are criticalfor effective management of water resources,espe-cially on small,densely populated islands with limit-ed storage capacity (Larsen and Concepcion,1998).Strategies for potential flood conditions are consid-ered in the next section.

2. Public Health and Safety

Both coastal and inland island populations andinfrastructure are at risk from climate-relatedextreme events. As reported earlier, model projec-tions suggest it is possible that the Pacific andCaribbean islands will be af fected by: changes inpatterns of natural climate variability (such as ElNiño); changes in the frequency, intensity, andtracks of tropical cyclones (hurricanes andtyphoons);and changes in ocean currents;and thatthese islands are very likely to experience increasedocean temperatures and changes in sea level(including storm surges and sustained rise). For theCaribbean,models indicate increasing air and watertemperatures,slightly wetter conditions in winter,

lect runoff and are used for water supply, flood con-trol,and limited hydroelectric power generation.Ground water accounts for ~30% of the totalamount of water used in Puerto Rico and surfacewater accounts for ~70% (Zack and Larsen,1993).

The US Virgin Islands are much smaller and have alower maximum elevation than Puerto Rico. Theseislands receive less rainfall and retain less freshwater. The US Virgin Islands have no perennialstreams and only limited ground water resources.Retention dams have been used on ephemeralstreams to promote recharge of coastal aquifers(Zack and Larsen,1993). In the US Virgin Islands,65% of drinking water supplies are provided bydesalinated seawater, making it the most expensivepublicly supplied water in the United States (Zackand Larsen,1993). Another approximately 22% ofthe drinking water supply originates from ground-water and 13% from rooftop catchments.

Flooding is a critical issue for Puerto Rico and theUS Virgin Islands due to the topography and rainfallpatterns of the islands. In the US Virgin Islands,human infrastructure (including buildings androads) has covered many of the ephemeral channelsthat allowed for infiltration. This not only diminish-es water supply, but greatly increases flash floodhazard. In Puerto Rico, flash floods typically resultfrom rainfall that is intense in the upper basins butsparse or nonexistent on the coast. These eventscan trigger hundreds of landslides,which are com-mon in the mountainous areas of Puerto Rico wheremean annual rainfall and the frequency of intensestorms are high and hillslopes are steep. One effectof these landslides is adding sediment to river chan-nels,which is carried into downstream reservoirs.Combined with other factors such as heavy rain-storms,urban development,and agricultural prac-tices,this sediment reduces the storage capacity ofmajor reservoirs,and diminishes their efficiency atreducing flood peaks (Zack and Larsen,1993).Large quantities of sediment are also washed intothe ocean,where it is deposited on the reefs anddeteriorates reef health.

Droughts are frequent and severe in the US VirginIslands. Any minor depletions in rainfall dramaticallyaffect agriculture and require water rationing (Zackand Larsen,1993). Droughts are infrequent inPuerto Rico. However, mandatory water rationingwas implemented six times during the 1990s result-ing in significant agricultural and other economiclosses (Larsen,2000). A drought in 1994-95 af fectedmore than one million people who endured manda-tory rationing for more than a year (Larsen,2000).

330


and slightly drier conditions in summer. Some mod-els project persistent El Niño-like conditions acrossthe Pacific caused by increased ocean temperatures.This also suggests that Atlantic hurricanes maydecrease in the future because of the relationship ofreduced hurricane activity in the Atlantic followingthe season of El Niño event onset in the Pacific.Nevertheless,there seems to be little net effect ontotal precipitation in the region because rain cloudsstill form and rain still falls, even in the absence ofhurricanes.

For the Pacific,models indicate a gradual warmingof air and water temperatures with an accompany-ing eastward shift of precipitation. This would besomewhat similar to what happens during El Niñoevents,and while not exactly the same,it wouldlikely cause changes in local rainfall patterns similarto those experienced during the El Niño phase ofthe ENSO cycle. While this is still uncertain,oneresult suggests precipitation amounts associatedwith tropical storms will likely increase by morethan 25% on average (Knutson et.al,1998);otherresults suggest a 5-11% increase in surface windsand a 28% increase in near-storm precipitation(Knutson and Tuleya,et al.,2000).

Globally, sea level is projected to rise two to fivetimes faster over the 21st century than over the 20 th

century as increased warming causes glacial meltingand thermal expansion of ocean water. The US-affili-ated islands would be af fected by variations in sealevel,such as storm surge associated with tropicalstorms,and fluctuations associated with El Niñoevents.

With these types of projected climate changes,manycurrent stresses that islands face will be exacerbat-ed. Storms can directly damage structures,interferewith the provision of services to communities,

cause deaths,and increase disease transmission. InAmerican Samoa, for example,the majority of exist-ing roads are located along the shoreline and areextremely vulnerable to damage by wind drivenwaves. Each section of new or rehabilitated road-way construction must include the additional costsof sea walls, resulting in extremely expensive high-way construction. However, without properlydesigned shore protection,American Samoa couldlose a major portion of the shoreline,as well as theentire length of adjacent roadway (OIA,1999) tostorm surge. Both the Pacific and Caribbean regionsare familiar with severe hurricanes and tropicalcyclones,which have caused billions of dollars indamage from the destruction of housing, agricul-ture, roads and bridges,and lost tourism revenue(Walker et al.,1991;Scatena and Larsen,1991).

The unique topography of Puerto Rico and theVirgin Islands makes them susceptible to floods andlandslides usually resulting from the extreme precip-itation associated with these storms. On mostislands, much of the population,transportation andsocial infrastructure,and economic activity is locat-ed near the coast,leading to dense areas of vulnera-bility. Diaz and Pulwarty (1997) suggest that thesocietal impacts of hurricane activity are potentiallymore worrisome in the future because of severalsocioeconomic changes that have occurred duringthis century. Population increases put more peoplein the path of potential hurricane devastation. Sincetransportation infrastructure has not kept pace withthis population increase,it takes longer to evacuatepeople (e.g.,Hurricane Floyd and East Coast evacua-tions), even though hurricane-forecasting ability hasimproved. Finally, the value of insured property atrisk is increasing,so the ability of insurance compa-nies to keep pace with potential damage is dimin-ishing.

331


Potential Consequences of Climate Extremes onInfrastructure and Human Health

Disruption of lifeline systems, including water supply, energy supply, waste management,and sanitation dueto fluctuations in temperature and precipitation,as well as storm events. Medical services can also be disrupt-ed during storm events;

Effects on people and public health, due to decreased water quality and sanitation,impact on agriculture andconsequently nutrition,and ecological changes that may increase the risk of disease transmission,and directimpacts from severe storms;and

Damage or destruction of infrastructure, including housing,lifelines,transportation,economic,and industri-al infrastructure due to storms, related storm surge,and sea-level rise.

show that increased flow resulting from storms car-ries 1,000 times the normal concentration of fecalsanitary bacteria (Alvarez,1998).

The Caribbean’s worst hurricane season since 1933came in 1995. Hurricanes Luis and Marilyn hit theVirgin Islands in September 1995 (see table for spe-cific damages from Hurricane Marilyn). Damagewas greatest in St.Thomas. Every telephone poleand 80% of the homes and businesses weredestroyed. It took eight weeks to get desalinizationplants running. Every cistern was contaminated bysalt water and there was no electricity and no handpumps to get water out of cisterns. Critical facilitiesthat were affected included hospitals,power anddesalinization facilities,sewer systems, fire houses,police stations,public shelters,ports,communica-tion systems,and docks. Notably, hospital roofswere destroyed;sewage distribution systems weredisrupted,leading to surface water and marine baycontamination;and power and telephone distribu-tion systems required months of repairs and cur-tailed service. The US Virgin Islands Bureau ofEconomic Research estimated the economic loss at$3 billion. (Potter et al.,1995;National HurricaneCenter, 2000).

The Pacific Islands have also suffered from severestorms. According to Pielke (2000),between theyears 1957 and 1995,Hawaii suffered over $2.4 bil -lion in hurricane damages. For the Pacific islands,1992 and 1997 were particularly active years. In1992 FEMA responded to “major disasters”in theMarshall Islands and Micronesia after Typhoon Axel

Leatherman (2000) also suggests that impacts fromextreme events on the human community are afunction of complex interactions between meteoro-logical phenomena and the environment,includinginfrastructure (e.g.,transportation and other parts ofthe built environment),land-use patterns,and social,political,and economic systems. He also asserts thatany future increases in destructive meteorologicalevents,as are suggested by some of the climatechange scenarios,will negatively impact the islandseconomically as well as socially.

Severe Storms. Hurricanes and typhoons are amongthe most socially devastating natural disasters,affect-ing more people on a yearly basis than earthquakes.Cumulative losses are also greater for hurricanesthan earthquakes (Leatherman,1998). Hurricanesand typhoons cause billions of dollars in losses dueto destruction of transportation and other infrastruc-ture,life,and property. Increases in population den-sity, changes in age structure and population health,urban sprawl,insufficient transportation infrastruc-ture,and human occupation of coastal and flood-prone areas have all increased the vulnerability ofpopulations and the infrastructure they depend onto hurricanes and typhoons. Deforestation hasincreased the amount of runoff and erosion thatresult from precipitation events contributing to thenegative impacts of severe storms. One of thosenegative impacts is the increased risk of some dis-eases (e.g., leptospirosis and other water-borne dis-eases) following floods caused by storms (evenstorms that are far less intense than hurricanes). Forexample,data collected in certain areas of Florida

332


Estimated Damages for Hurricane Marilyn in the US Virgin Islands

Category of Damage Estimated Costs Sewage Treatment Facilities $1,000,000 Roads and Bridges 1,000,000 Damage to Manufacturing 1,000,000 Agriculture 1,000,000 Water 3,000,000 Protective Measures 10,000,000 Debris Removal 18,000,000 Telephones 30,000,000 Electrical 70,000,000 Lost Employment 80,000,000 Public Buildings 210,000,000 Damage to Hotels 253,000,000 Lost Tourist Revenue 293,000,000 Private Housing 1,300,000,000 Total 2,271,000,000

Source: The Virgin Islands Natural Hazard Mitigation Plan,B. Potter et al.,1995.

storm surge),on flood plains,or on hillsides proneto landslides. In the meantime,human modifica-tions are increasing the frequency of landslides. A

333


Hurricane Georges of 1998: Effects on Puerto Rico

On September 15,1998,a tropical weather system off the coast of West Africa was upgraded to a tropicaldepression. Within 24 hours,the tropical depression intensified to be the tropical storm that would becomethe fourth hurricane of the 1998 season,Hurricane Georges. At its strongest,Hurricane Georges had sus-tained winds of 150 mph,becoming a category 4 hurricane,the second most intense level of hurricane onthe five-level scale used to measure intensity.

On September 21,Hurricane Georges (reduced to a category 3) began to sweep across Puerto Rico. It was tobe the first hurricane since the 1932 San Ciprian hurricane to cross the entire island. The eye of the hurri-cane,measuring 25-30 miles wide,passed within an estimated 15 miles of San Juan,Puerto Rico’s capital,leav-ing a trail of devastation in its wake.

While rainfall across the island varied greatly, some areas received up to 26 inches of rain within a twenty-four hour period. The rainfall resulted in flooding,landslides,and catastrophic losses in infrastructure.Twelve people on the island of Puerto Rico were killed;three as a direct result of the storm,and nine othersas an indirect result. Infrastructure impacts were enormous. The island lost 75% of water and sewage serv-ice,96% of electrical power, and suffered damage to 50% of utility poles and cables. An estimated 33,100homes were destroyed. Road damage was estimated at $22 million and damage to public schools was esti -mated at $20-25 million. Agricultural damages were also significant with 75% of the coffee crop,95% of theplantain and banana crops,and 65% of all poultry destroyed (USGS,1999;NOAA,1999).

Overall costs to the US mainland and island territories reached $5.9 billion,with an estimated $2 billion ormore in Puerto Rico alone. On October 15,1998,the Red Cross announced that Hurricane Georges had infact been “the most expensive disaster relief effort in the organization’s 117-year history.” They estimate thatsolely within the US mainland and island territories, about 187,300 families were affected (American RedCross,1998,1999).

Figure 4. On September 21, 1998, Hurricane Georges swept acrossPuerto Rico. The eye of the hurricane was 25-30 miles wide andpassed within 15 miles of the capital, San Juan, leaving a trail ofdevastation in its wake. The path of the hurricane and rainfalltotals are shown here. Some areas received up to 26 inches of rainwithin 24 hours. Flooding, landslides, and catastrophic losses ininfrastructure resulted. Hurricane Georges Map –USGS:http://water.usgs.gov/pubs/FS/FS-040-99/images/PR_fig01.gif SeeColor Plate Appendix

Path of Hurricane Georges in Relation to Puerto Ricowith Precipitation Totals

struck in March;then in Guam in the aftermath ofTyphoon Omar ($500 million dollars in damages).September 1992 saw Hurricane Iniki hit the Hawaiianisland of Kauai. It resulted in 7 deaths,approximately$1.8 billion in damages,and $259.7 million in FEMAdisaster relief costs. In December 1996,Typhoon Fernstruck Yap and in April of 1997,Typhoon Isa struckMicronesia. Later that year, in December,Supertyphoon Paka ($650 million in damages) pum-meled Guam and the Northern Marianas,as well asthe Marshall Islands causing widespread destruction(Hamnett et al.,2000;FEMA website,2000).

Landslides. Each year in Puerto Rico,landslides causeextensive damage to property and occasionally resultin loss of life (Larsen and Torres-Sanchez,1998).Although landslides can be triggered by seismic activi-ty and construction on hillslopes,the leading cause oflandslides in Puerto Rico is intense and/or prolongedrainfall (Larsen and Simon,1993). Population densityin Puerto Rico is high, about 1,036 people per squaremile (400 people per square kilometer),and isincreasing. This increase is accompanied by the useof less desirable construction sites. As a result,humanpopulations are becoming more vulnerable to land-slide hazards (Larsen and Torres Sanchez,1998).People tend to live in the coastal zone (vulnerable to

millions of years from the continual challenge ofsome of the selective forces that shape continentalorganisms,oceanic island fauna and flora areuniquely vulnerable to the human introduction ofpreviously absent predators,diseases,and competi-tors (Loope and Mueller-Dombois,1989). Naturalecosystems also appear to be most vulnerable tomany effects of climate changes (e.g.,temperatureand carbon dioxide increases,precipitation changes,sea-level rise,and changes in the frequency andintensity of extreme events) since climatic condi-tions strongly determine where particular plantsand animals can live, grow, and reproduce. Somespecies and ecosystems are so strongly influencedby the climate to which they are adapted,that theyare very vulnerable to even modest climate changes.Some island ecosystems are already constrained byclimate as well as geography and are likely to faceextreme stress from projected climate changeswhile some species may disappear entirely. Forexample,increased ocean temperatures and possiblechanges in ocean circulation patterns will affectcoastal ecological systems,such as mangrove forestsand coral reefs and important marine resourcessuch as fisheries. Often there is very little that canbe done to assist ecosystems in adapting to the pro-jected speed and amount of change.

Pacific Islands. The Hawaiian Islands are larger, havemore topographical and climatic diversity, and havehigher biological uniqueness than most other PacificIslands. Hawaii also has the highest proportion ofextinct and endangered species of anywhere in theUS (USGS,1999a). Hawaii has lost over 70% of itsoriginal (pre-Polynesian) 140 land bird and 1200+land snail species. Of 1,302 known taxa (89%endemic) in the Hawaiian vascular flora,106 (8%)are extinct and an additional 373 (28%) are consid-ered at risk of becoming extinct in the near future.The “at risk”taxa include over 200 that are federallylisted as endangered, roughly one-third of the totallisted plants nationwide (Loope,1998). Loss ofabout two-thirds of the area of Hawaii’s natural habi-tats has been a primary reason for the losses to date;however, at present,biological invasions pose thegreatest threats to further dismemberment of thebiota.

Although invasive non-indigenous species are prob-lematic throughout the Pacific,Hawaii,being in themainstream of commerce,has suffered significantly(Holt,1998). Hawaii is inherently vulnerable to inva-sion due to great isolation from source areas of natu-rally colonizing species and high endemism(Mueller-Dombois and Fosberg,1998). This is not tosay that other Pacific Islands have not suffered dev-

study conducted in Puerto Rico in 1998 found that,although mean annual rainfall is high,intense stormsare frequent,and hillslopes are steep, forested hill-slopes are relatively stable as long as they are notmodified by humans. However, the greater the mod-ification of a hillslope from its original forestedstate,the greater the frequency of landslides (Larsenand Torres-Sanchez,1998).

Adaptation Options.Strategies for improving public health and safetyunder changing climate conditions could includenumerous options. For example:1) upgrading andprotecting infrastructure (including transportationinfrastructure);2) initiating comprehensive disastermanagement (and avoidance) programs (includingmapping and risk analysis);3) changing and enforc-ing land use policies to avoid hazards/hazardousareas;4) adopting and enforcing increasingly strin-gent building codes;5) improving sanitation andhealth care infrastructure;6) improving emergencyplans;and 7) increasing public and official informa-tion and outreach programs related to the range ofhistorical disaster experiences and to future risksassociated with climate change.

For potential flood conditions,adaptation measuresinclude changes in land use policies that discourageconstruction in flood plains or areas at risk for land-slides,and that allow for natural patterns of runoff.Improved monitoring,observation,alert, warning,and evacuation systems would also be desirable.

3. Island Ecosystems

Islands are extremely valuable as living laboratoriesfor understanding species adaptation and evolution.Much as the Galapagos archipelago provided excep-tional insight for Charles Darwin in the 1830s,themore ancient Hawaiian island chain has sincebecome recognized as a premier site in the worldfor scientific studies of evolution in isolation. Beforehuman influence,the Hawaiian Islands had no ants,reptiles,amphibians,or mammals (except for onebat species). The roughly 1,500 species of animalsand plants that successfully colonized this climati-cally and topographically complex archipelago overthousands of miles of open ocean — on the winds,by floating,or attached to storm-driven birds(Carlquist,1980) — gave rise to roughly 15,000species, over 90% of them endemic (found nowhereelse).

Since the time of Darwin,it has been recognizedthat organisms of oceanic islands are notoriouslyvulnerable to extinction. Having been isolated for334


astation by humans and their invasive species intro-ductions. For example,Steadman (1995) estimatesthat more than 2,000 bird species may have beeneliminated by early human settlers in Polynesia. Theexample of the brown-tree snake (Boigairregularis) on Guam (Fritts and Rodda,1995)should serve as a cautionary note for all oceanicislands that non-indigenous species may explode inisland habitats with unexpected consequences.According to Fritts (1999),the brown tree snakeprobably arrived on Guam via ship cargo from theNew Guinea area and the first inland sighting was inthe early 1950s. Some forested areas now have pop-ulations up to 13,000 snakes per square mile. Thebrown tree snake is responsible for essentially wip-ing out the native forest birds of Guam. Twelvespecies of endemic birds have disappeared from theisland,and several others survive in such low num-bers that they are close to extinction. The snake isresponsible for economic impacts as well. Forexample,snakes crawling on electrical lines fre-quently cause power outages. More than 1,200power outages have been attributed to snakes since1978. The power interruptions have reportedlyresulted in numerous problems ranging from foodspoilage to computer failures and all include signifi-cant expenses for the population.

A comparable botanical example of the unintendedconsequences of non-indigenous species invasionsand impact is seen in the invasion of the SouthAmerican tree Miconia calvescens in Tahiti (Loope,2000). Miconia now dominates 65% of Tahiti’s for-est canopy and its presence threatens some 40-50endemic plant species to the point they are now onthe verge of extinction (Meyer and Florence,1997).Miconia is a serious issue in Hawaii as well.Probably because of its attractive purple and greenfoliage,it was introduced to Hawaii as an ornamen-tal — a single tree in the Oahu Wahiawa BotanicalGarden in early 1960. By 1964,it had reachedHawaii and by the late 1960s or early 1970s it hadreached Maui. Based on a 1995 discovery of a fruit-ing tree over 10 meters tall,it is estimated that itreached Kauai by the early 1980s. Its presence inHawaii poses a threat to endemic plant species inhabitats receiving 1800-2000 mm (75-80 inches) ormore of annual precipitation (Loope,2000). Thevulnerability of oceanic islands to invasions cannotbe underestimated.

The USGS (2000) Hawaiian Ecosystems at Risk(HEAR) project asserts that “the silent invasion ofHawaii by insects,disease organisms,snakes, weeds,and other pests is the single greatest threat toHawaii’s economy and natural environment and to

the health and lifestyle of Hawaii’s people.”According to HEAR,millions of dollars in crop loss-es,the extinction of native species,the destructionof native forests,and the spread of disease canalready be attributed to alien pest (plant and animal)invasions,with many more harmful pests nowthreatening to invade Hawaii and cause further dam-age. For example,an invasion of ant species such asthe “red imported fire ant”(Solenopis invicta) andthe little fire ant (Wasmannia auropunctata) maypose a threat to Pacific islands’biodiversity, tourism,agriculture,and quality of life that rivals the threatof global climate change (R.Thaman,University ofthe South Pacific,personal communication,1999).

In spite of profound human modification of theHawaiian landscape and biota (Cuddihy and Stone,1990),large tracts of near-pristine ecosystemsremain at high-elevation. Even with the high inci-dence of extinction and endangerment,Hawaii hasmore non-endangered endemic species of vascularplants,birds,and insects than any other state exceptCalifornia (Loope,1998). In Samoa,as well as onother high islands of the Pacific (e.g.,the Societyand Marquesas,Tonga,the Cook Islands,and theHawaiian islands) relatively intact biodiversity is cen-tered in high-elevation cloud forests,which may beexceptionally vulnerable to climate change (see boxon Montane Cloud Forests).

Caribbean Tropical Forests. The islands in thePuerto Rican Bank have been largely denuded ofnative vegetation,have extremely dense human pop-ulations,and face a formidable array of environmen-tal problems,including extinction of native plantsand animals (Wiley and Vilella,1998). The introduc-tion of non-indigenous species,particularly verte-brate animals,is yet another difficulty facing thenative plant and animal communities. Still,withincreasingly aggressive conservation efforts,PuertoRico and the US Virgin Islands may retain the best-preserved examples of certain natural ecosystems ofthe West Indies. Being showcases of natural ecosys-tems is part of the image of these tropical islands,and is deemed vital to their tourist economies(Wiley and Vilella,1998).

Past policy considerations for tropical forest conser-vation have generally dismissed the threat of climatechange as quite insignificant in comparison to land-use change and other human impacts. However, abetter understanding of tropical ecology is nowleading many scientists to the conclusion that tropi-cal forests may be very sensitive to climate change.It is being increasingly recognized that factors otherthan warming,including changes in hydrology, rain-

335


These pines depend on lenses of freshwater that arebeing gradually eliminated by salt water as sea levelrises. The result has been that mangroves havereplaced the pines. This transformation will likelycontinue as sea level rises (Alvarez,1998). On theother hand,Snedekar (1993) has argued that man-groves in the Caribbean are more likely to be affect-ed by changes in precipitation than by higher tem-peratures and rising sea levels because they requirelarge amounts of fresh water to reach full growthpotential. A decrease in rainfall in the Caribbeancould thus reduce mangrove productive potentialand increase their exposure to full-strength seawa-ter. (For more information see the Coastal chapter.)

Coral Reefs. It is projected that coral reefs areamong the most sensitive ecosystems to long-termclimate change (IPCC,1995). Widespread coralbleaching has already been observed in the Pacificand Caribbean in association with ENSO events andwould be expected to continue and accelerate withincreased ocean temperatures. Bleaching occurswhen the coral animal expels all or part of its sym-biotic algae,when the pigments in the algae declinedrastically, or when there is some combination ofthe two. Bleaching is a stress reaction that can beinduced by many individual or combinations of con-ditions:high or low water temperature,high fluxesof ultraviolet radiation,prolonged aerial exposure,freshwater dilution,high sedimentation,or variouspollutants (Glynn and de Weerdt,1991). While lowtemperatures have generally been considered a lim-iting factor for the development of coral reefs incooler climates, reefs are also susceptible toincreased temperatures because they are alreadynear their maximum threshold temperatures in sum-mer. While various species and populations mayrespond differently, in general,coral are likely tobleach but survive if high temperature anomaliespersist for less than a month. Such a short-termexposure usually allows corals to recover. However,sustained high temperatures can result in chronicstress that can cause physiological damage that maybe irreversible (Wilkinson et al.,1999;Glynn,1996).Even a sub-lethal stress may make corals highly sus-ceptible to infection by a variety of opportunisticpathogens.

In the late 1980s and 1990s,after localized bleach-ing in 1982-1983, bleaching became a regular andpervasive problem in the Caribbean and began toappear in the Pacific and Indian Oceans as well.Bleaching occurred in American Samoa during thewarm event of spring 1993. During the El Niño of1997-1998, bleaching began in the eastern Pacificbut then expanded to an unprecedented depth and

fall patterns,and the frequency and intensity ofstorms and fires,may have far-reaching conse-quences (Markham,1998). In comparison toHawaiian forests,Caribbean forests are well-adaptedto disturbance,but increased frequency of hurri-canes, floods,droughts,and fires could lead tounprecedented stresses and drastic changes in foreststructure and composition. For example,a comput-er model run for the Luquillo reserve of Puerto Ricoshowed that increasing intensity of hurricanes couldreduce the density of trees in the forest and theirtotal biomass,and favor the development of fast-growing,short-lived and weedy species,includinginvasive species (O’Brien,et al.,1992).

Mangrove Forests. Coastal wetlands and mangrovesof the Caribbean and the Pacific are vital areas forsustaining populations of birds,juvenile fishes,andinvertebrates. They also provide important stormand flood buffers between the sea and coastal com-munities,are coastal stabilizers,and nutrient sinks.Mangrove forests have the capacity to adapt to sea-level rise,although this is being limited by humanactivities that interfere with forest regeneration.One study suggests that mangrove communities willdo better in macrotidal,sediment-rich environmentswhere strong tidal currents redistribute sedimentrather than in microtidal,sediment-starved environ-ments,like most small islands (Parkinson et al.,1994). In the pine rocklands (e.g.,Sugarloaf Key) inthe lower Florida Keys is an area where pine treesgrow directly on rock on very low-lying islands.

336


Coral reefs are among the most sensitive ecosystems to long-termclimate changes. Source: ©P. Grabhorn

Mangrove trees grow at the land-ocean interface, helping to protectcoastal landsfrom erosion. - Source: ©P. Grabhorn

337


region across to and including the Indian Ocean,and in the Pacific Ocean from Australia to Polynesia.Bleaching was widespread in the Republic of Palauduring the 1997-1998 El Niño.

Corals in the Caribbean are currently impacted byenormous numbers of international tourists,highvolume of ship traffic,and fishing. Pacific Oceancoral reefs are some of the world’s healthiest over-all; about 70% are rated in good-to-excellent condi-tion. But 30% are rated fair-to-poor and many aredying,while human impacts are growing. Theregion’s extremely diverse corals,mangroves,andsea-grasses are pressured by deforestation, agricul-ture,construction,pollution,and fishing.Climate-related changes could affect coral through

more than simply increased water temperatures:insome Pacific jurisdictions, changes are expected inaerial exposure due to sea-level alterations associat-ed with ENSO. In the Caribbean,erosion/sedimenta-tion may be caused either directly through floods orindirectly through increased erosion following firesassociated with drought conditions. In addition,hurricanes and typhoons can damage coral reefs.For example,Hurricane Hugo in 1989 caused exten-sive damage to reefs outside of Puerto Rico by frag-menting, overturning,or destroying virtually allelkhorn coral colonies that were over one meter insize (USGS,1999). Elkhorn coral are a principal reefbuilder. Unhealthy coral reefs are the most suscepti-ble to destruction by hurricanes and typhoons.Effects of increasing atmospheric carbon dioxide

Montane Cloud Forests: Unique and Vulnerable Ecosystems

The tropical montane cloud forests of islands are rare,diverse,and vulnerable natural habitats,occurring insuch Pacific islands as the Society and Marquesasislands,Cook Islands,Tonga,Samoa,and theHawaiian Islands as well as elsewhere in the world(e.g.,Madagascar, Sri Lanka,Greater and LesserAntilles). The upper and lower limits of cloud for-est vegetation are determined by the altitude of apersistent cloud zone. The upper limit of this zoneis generally set by the elevation of the trade windinversion and the lower limit is determined by tem-perature and humidity. Cloud forests are likely tobe extremely sensitive to climate change. Theseunique forest environments would be affected byany potential changes in the altitudinal level of thetrade wind inversion,as well as changes in tempera-ture,precipitation zones,increased drought,or hur-ricanes. On high islands such as Maui and Hawaii,thetwo largest of the Hawaiian islands,the cloud forest veg-etation now changes abruptly to grassland or shrubland above the mean elevation of the inversion layer.

Loope and Giambelluca (1998) have identified Pacific island cloud forests as particularly vulnerable to human-induced climate change,since relatively small climate shifts are likely to trigger major local changes in rainfall,cloud cover, and humidity. Such climatic disruptions will undoubtedly favor the penetration of invasive non-indigenous species into previously intact ecosystems. Pounds et al.(1999) have already demonstrated a con-stellation of population crashes of native birds, reptiles,and amphibians in biological communities in the mon-tane cloud forest of Monteverde,Costa Rica,apparently resulting from oceanic and atmospheric warming anddrying in the extreme eastern tropical Pacific.

The cloud forests of East Maui,Hawaii (Haleakala National Park and neighboring reserves), for example,arehome to numerous endemic species,with approximately 90% of the native flowering plants and invertebratesendemic to Hawaii. Nine endemic bird species in the Hawaiian honeycreeper family occupy these cloudforests. One reason these birds survive in the higher forest is because the mosquitoes that carry avian malariaare limited at higher elevations by cooler temperatures. Therefore,any warming that allows the mosquitoesto move farther up the mountain may threaten the birds, five species of which are already listed as endan-gered by the US Fish and Wildlife Service.(From Loope,personal communication.)

The Hawaiian anianiau (Hemignathus parvus) an endangeredspecies, is now restricted to high elevation montane rain forest onKauai. Like other native Honeycreepers, it is highly susceptible toavian pox and malaria, and as a result of mosquito transmitted birddiseases the Anianiau has disappeared from its former low eleva-tion habitats. Source: (http://water.usgs.gov/pubs/FS/FS-012-99/ )

4. Sea-Level Variability

Sea-level rise and the associated erosion and inunda-tion problems are currently extremely importantissues for many of the US affiliated islands. Thereare two factors that affect the impacts of sea-levelrise on islands:the rate and extent of global sea-levelrise and the occurrence of episodic events,such asextreme lunar tides,ENSO related changes,andstorm-related wave conditions. Rising sea levelsover the past century have already resulted in saltwater intrusion into freshwater lenses,inundation ofcoastal vegetation (such as taro,pulaka,and yams),and coastal erosion,since a typical beach erosionrate can be 150 times the amount of sea-level rise(Alvarez,1998). The Gulf and Caribbean region

concentrations could also possibly have adverseeffects on reef building corals (see Coastal chap-ter for details). Without the reefs,low-lyingislands would be susceptible to rapid erosion anddestined to eventual disappearance.

Adaptation Options There are a limited number of specific optionsthat can be employed to assist island ecosystemsto cope with a changing climate. Some of thoseoptions include the following:attempting to slowbiological invasions;strengthening and enforcingpolicies that protect critical habitats;improvingunderstanding of the local effects of climate vari-ability and change;and increasing awareness oftourists and the public concerning the value ofspecies and biodiversity.

338


Pacific Island Observed Sea-Level Trends (based on Merrifield,personal communication)

Location/Name Observed Rise in Relative Sea Years of Data Level (averaged forthe 20th century) Collection

HawaiiHilo 13.6 +/- 2 in (34 +/- 5 cm) 1927- 1999Honolulu 6 +/- 0.8 in (15 +/- 2 cm) 1905- 1999Nawiliwili 6 +/- 1.6 in (15 +/- 4 cm) 1954- 2000Kahului 8.4 +/- 2 in (21 +/- 5 cm) 1950- 1999Mokuoloe 4 +/- 2 in (10 +/- 5 cm) 1957- 1999

Guam 0 +/- 2.4 in (0 +/- 6 cm) 1948- 1999

American SamoaPago Pago 6.4 +/- 2.4 in (16 +/- 6 cm) 1948- 1999

Commonwealth of Northern MarianasSaipan -0.4 +/- 8.8 in (-1 +/- 22 cm) 1978- 1999

Republic of the Marshall IslandsWake 7.2 +/- 2 in (18 +/- 5 cm) 1950- 1999Kwajalein 3.6 +/- 1.6 in (9 +/- 4 cm) 1946- 1999

Republic of PalauMalakal - 1.6 +/- 7.2 in (-4 +/- 18 cm) 1969- 1999

Federated States of MicronesiaKapingamarangi -6.4 +/- 9.2 in (-16 +/- 23 cm) 1978- 1999Pohnpei 6.4 +/- 7.2 in (16 +/- 18 cm) 1974- 1999Yap -5.6 +/- 7.2 in (-14 +/- 18 cm) 1969- 1999

experienced on average a relative sea-level rise of3.9 inches (10 cm) during the 20th century. In thePacific region,as around the globe, absolute sealevel is also rising.There is,however, a great deal ofinter-island variability in relative sea level (seetable). This occurs because a number of PacificIslands are rising due to geologic uplift. As a resultof this uplift, relative sea-level changes for thoseislands can appear to be negative. The Pacificregional averages can,therefore,tend toward zerodepending on which islands are included in theaverage;hence,the global sea-level trend is not evi-dent in the long-term average trend in relative sea-level in the Pacific (Merrifield,personal communica-tion).

Long-term global rates of sea-level rise are projectedto be 2 to 5 times faster in the 21st century than dur-ing the 20th century. Future sea-level rise,both glob-al and episodic (since increasing global sea levelwill also raise the level from which episodic eventsoccur),will increasingly contribute to negative con-sequences for populations and ecosystems (Titus etal.,1991).

Risk of flooding,inundation,and coastal erosion onparticular islands depends on both physical proper-ties of the island (elevation, rock and soil-type,loca-tion) and on biological properties (presence orabsence of biotic protection such as coral reefs andmangroves). The most at-risk island types are low-lying coral atolls because of both their elevation andcomposition. The vulnerability of specific sitesdepends on the value of what is at risk,includingsuch aspects as natural resources,transportation andsocial infrastructure,businesses,and human popula-tions. Among other factors,sensitivity increasesbecause of dependence on coastal activities fortheir economies,dependence on coastal aquifers forfreshwater supply, the presence of culturally valuedstructures along the coast,or because of previousland-use changes. Ultimately, vulnerability to sea-level rise is a function of both the natural resilience

of the shoreline and biological resources,as well asthe value of the resource at risk,and the ability ofhuman populations to respond and implement adap-tation measures and strategies. Vulnerability alsodepends on the timeframe over which recovery ispossible and the resources that are available.

Islands will be particularly vulnerable to sea-levelchanges and consequent impacts if they are low-lying and have limited resources for protecting theircoastline. In the Pacific,potentially vulnerableisland groups include the atolls of the Republic ofthe Marshall Islands and the Federated States ofMicronesia. In the Caribbean, much of the metro-politan area of San Juan in Puerto Rico is alreadyclose to sea level. Islands that are high volcanic orlimestone,mountainous,naturally rising,or lackresources and populations in coastal areas will beless vulnerable.

Adaptation OptionsCoping with sea-level change and the resulting con-sequences (e.g., flooding,inundation of freshwaterand agricultural systems,erosion,destruction oftransportation and other built infrastructure) willrequire a variety of strategies. Some of the optionsthat are likely to be needed could include:protect-ing coastal infrastructure (both social and trans-portation), water systems, agriculture,and communi-ties;integrated coastal zone management;and cropshifts and the use of salt-resistant crops. Retreatfrom risk-prone,low-lying areas is also likely to benecessary in some cases,but will be complicateddue to land ownership,and could have significantconsequences for social and cultural identity.

For more details on projected sea-level rise and itsimpacts,see the Coastal chapter in this volume andthe additional volume: The Potential Impacts ofClimate Change on Coastal and Marine Resources:Report of the Coastal and Marine Resource SectorTeam of the US National Assessment.

ADDITIONAL ISSUESClimate variability and change present island gov-ernments,businesses,and communities with chal-lenges and opportunities in a number of other keyareas including:

Tourism

Tourism remains a significant contributor to theeconomies of the Caribbean islands,the State ofHawaii,Guam,and the Commonwealth of the

339


A typical beach erosion rate can be 150 times the amount of sea-level rise. Source: ©P. Grabhorn

in the Pacific during an El Niño is associated withan eastward displacement of Skipjack tuna stocks.As global patterns of water temperature change,species shifts can be expected to follow thesechanges.

Climate variability and change present a number ofchallenges for coastal and marine fisheries thatremain significant components of the culture andeconomies of many of the island jurisdictionsaddressed in this chapter. The movement of thecommercially-important Skipjack tuna noted aboveresulted in the stocks being closer to the MarshallIslands and away from the waters of Micronesia.Industry representatives in FSM noted reducedcatches in late 1997 and early 1998 (coinciding withthe 1997-1998 El Niño),while government officialsin the Marshall Islands noted an increase in fishingactivity within their waters in late 1997, resulting ingreater access fees from vessels within theirExclusive Economic Zone.

Changes in the ENSO cycle or other climate-relatedchanges in ocean circulation and productivity couldbring significant changes to the location of tunastocks,thus providing opportunities for those juris-dictions that find themselves close to large stocks,and problems for other jurisdictions that find them-selves far away from commercially-importantspecies. For many of the Pacific island jurisdictionsaddressed in this National Assessment,the develop-ment of a viable tuna industry is considered animportant component of their economic growth.StarKist Samoa and Samoa Packing, for example,aretwo of the largest tuna canneries operating withinthe US and they are the largest private sectoremployers in American Samoa. Fishingaccess/license fees from Distant Water FishingNations catching tuna in the FSM’s ExclusiveEconomic Zone are a major source of revenue forthe country (constituting a 17% share of the nation-al GDP in 1994). Further development of thatindustry is currently the primary focus for economicdevelopment in FSM.

In addition to their importance to the tourismindustry, many inshore species,including a numberof reef fish, remain important contributors to thesubsistence diet in small island communities. Anyclimate-related changes in the habitats that supportthese fisheries would have consequences for thosecommunities.

Northern Mariana Islands,and is considered to offereconomic growth potential for many of the jurisdic-tions addressed in this chapter. Unique ecosystems(both terrestrial and marine) and attractive coastalareas are among the natural assets that draw touriststo the islands of the Pacific and Caribbean. Thesenatural assets are already under stress as a result ofpollution and the growing demands of an increas-ingly coastal population. Many of the projected cli -mate changes for the islands would exacerbatenumerous of the stresses that the islands are alreadyunder. For example increasing rates of sea-level risewould contribute to:sea water inundating low-lyingareas and threatening key infrastructure (includingairports and roads);increasing the risks of damagefrom tropical storms (particularly damages associat-ed with storm surge);threatening coastal water sup-plies through salt water intrusion;and reducing theextent and quality of sandy beaches through bothinundation and increased erosion associated withstorm surge. The South Atlantic Coast-Caribbeanworkshop report suggests that at the present rate ofsea-level rise,there is already a loss of 9 feet (2.7meters) of coastline due to erosion every decade,insome areas. The projected increased rate of sea-levelrise would produce over 33 feet (10 meters) of ero-sion per decade since a typical beach erosion ratecan be 150 times sea-level rise.

Other climate-related changes of concern to tourisminclude: changes in tropical storm patterns; changesin temperature and rainfall patterns with attendantconsequences for terrestrial ecosystems and ani-mals;and changes in ocean temperature,circulation,and productivity that could affect important marineresources like coral reefs and the fish they support.In addition,the climate itself is a magnet fortourists. Locations could become undesirable totourists because of detrimental temperature andcomfort-level changes. It is also evident that just thethreat of some impacts has caused tourists to planvacations or conventions elsewhere.

Fisheries

Climatic conditions influence where specificspecies can live, grow, and reproduce. The appropri-ate temperature ranges of some fisheries areextremely narrow and as a result they are sensitiveto even small temperature variability within theirranges. Recent scientific studies are demonstratingan important link between patterns of natural cli-mate variability, such as the ENSO cycle,and themigratory patterns of important pelagic species liketuna. Hamnett and Anderson (1999), for example,suggest that the eastward expansion of warm water340


341


Agriculture

Agriculture, for both commercial and subsistencepurposes, remains an important part of theeconomies of many of the island jurisdictionsaddressed in this National Assessment. Much of themost productive agricultural land on these islands islocated in low-lying coastal areas that are at-riskfrom climate-related changes in sea level. In addi-tion to problems associated with inundation,salt-water intrusion associated with sea-level rise wouldalso present challenges to agricultural productionunless appropriate salt-tolerant species could be uti-lized. Climate-related changes in rainfall and tropicalstorm patterns also present problems for agriculturein island communities as evidenced by the impactsof the 1997-1998 El Niño event in the Pacific. Forexample, agriculture in all Pacific Island jurisdictionsaffiliated with the US except Guam suffered as aresult of the droughts associated with the 1997-1998 El Niño event. In Yap,taro losses were estimat-ed at 50-65% and over half of the banana trees evalu-ated on Pohnpei had died or were seriouslystressed. Agriculture in Guam,on the other hand,suffered substantially from the El Niño-inducedTyphon Paka that occurred in December 1997.

Data Collection and Availability

One other issue of concern to the islands relates tothe complexities of data collection, availability, andreliability, both for climate data and impact studies.These data are necessary to enhance the ability ofscientists and decision makers throughout the islandregions to understand and respond to the challengesand opportunities presented by changes in climateand other critical environmental conditions.

Climate prediction data availability is at times limit-ed for islands because most climate models in usetoday are unable to effectively capture island-scaleor sub-island scale processes. Also, geographic isola-tion and dispersion makes data collection andresearch in island settings,particularly the Pacific,costly and difficult. The harsh environments affect-ing most islands pose related difficulties in mainte-nance of monitoring stations. At the same timeresources for small-scale,locally-based researchfocused on individual islands or island ecosystemsare difficult to secure,especially since many of theseecosystems are unique and,therefore, researchresults tend to have more local than global signifi-cance. Another difficulty is related specifically toimpacts research where many impacts investigationsare undertaken by non-local scientists who toooften fail to capture/integrate valuable information

associated with traditional knowledge of specificisland cultures. In the case of many of the islands,the communication and information managementinfrastructure is limited and already stressed.Establishing and maintaining an interactive dialoguebetween scientists and beneficiaries of their workoutside the scientific community will require bothtechnological upgrading and institution of a newparadigm of collaboration and integration of infor-mation related to climate changes and their impacts.These challenges point to a significant need forlocal capacity building (e.g.,technical training indata gathering,monitoring technology, informationdistribution,impacts research,and communicationtechnologies).

ADAPTATION STRATEGIES

Overall,strategies helpful for islands to cope withclimate variability and change involve a wide rangeof options. While identifying many of the possibleoptions is an important part of this Assessment,it isonly a first step. It was not possible at this time toevaluate any of these potential options for practicali-ty, effectiveness,or cost. Nevertheless,some poten-tial adaptation options for the islands include:increased use of climate forecasting capabilities;consideration of changes in land use policies andbuilding codes;use of new and emerging technolo-gies to both mitigate the causes and reduce theimpacts of climate change;increased monitoring,sci-entific research,and data collection abilities;andimproved management policies. In addition,broadpublic awareness campaigns and communicationwith user groups are important mechanisms forimproving public health and safety, and protectingnatural resources.

Land-use policies and building codes can mitigateflood conditions (from sea-level rise as well as fromextreme precipitation events) and can encourageconstruction in areas where life and property willnot come into danger from floods,landslides,andsevere storms including storm surge. Implementingsuch land-use policies could be an important copingoption. Planting of trees and reforestation encour-ages natural infiltration,which increases under-ground water resources and prevents runoff thatcan lead to flood conditions. Building and trans-portation infrastructure can be renovated orreplaced in the natural renewal cycle with higherstandards. It is important to protect the species andecosystems that currently exist and find strategies tolimit sprawl and population growth that are impact-ing presently healthy ecosystems.

342


Possible Climate Change Adaptation Options as identified by the Federated States of Micronesia

Water • Conduct a comprehensive inventory of existing water resourcesResources • Assess the status of storage and distribution systems and secure resources for necessary

improvements.• Encourage improvements to residential and commercial catchment systems and identi-

fy/support new technology.• Identify opportunities to adjust water conservation and management policies to incorpo-

rate information about climate variability and change.• Document the experience gained during the 1997-1998 El Niño and build on the con-

cept of drought management task force(s) to assist governments,communities,and busi-nesses in responding to climate-related events.

• Identify opportunities to improve watershed management.

Coastal • Identify buildings,infrastructure,and ecosystems at risk and exploreResources opportunities to protect critical facilities.

• Develop and implement integrated coastal management objectives that enhanceresilience of coastal systems to climate change and sea-level rise.

• Consider the need for beach nourishment and shoreline protection programs in high-riskareas.

• Integrate considerations of climate change and sea-level rise in planning for future con-struction and infrastructure.

Agriculture • Document the experience gained during the 1997-1998 El Niño and build on the con-cept of drought management task forces to assist governments,communities and busi-nesses in responding to climate-related events.

• Develop policies that protect both subsistence and commercial crops during extremeevents.

• Explore opportunities to diversify crops and select drought and/or salt-tolerant specieswhere appropriate.

• Document low-lying agricultural areas at-risk from the effects of sea-level rise and consid-er protection measures where appropriate and necessary.

Fisheries • Enhance data collection and analyses required to improve understanding of the impactsof El Niño and La Niña events on tuna and other critical fisheries.

• Identify and protect critical habitats for key inshore and near-shore species,particularlythose important for subsistence fisheries.

• Support monitoring and monitoring programs designed to improve understanding of theregional and local consequences of climate variability and change for tuna and otherimportant fisheries.

From the National Communication to the United Nations Framework Convention on Climate Change,October 1999.

Improved technologies and water conservation arecritical for protection of water resources. Theseinclude more and better rainfall catchment systems,improved storage and distribution systems,artificialrecharge of aquifers,and desalinization. In addition,it is important for communities to take advantage ofnew capabilities in climate forecasting which,among other advantages,can allow for preparationfor water scarcity conditions.

Improved management is critical for both ecosys-tem and human system protection. This includesmore efficient management of existing water sup-plies,active management of natural areas,attentionto policies that attempt to slow biological invasions,increased disaster response planning,and appropri-ate land uses.

At the regional workshops,participants advocatedfor effective,unified planning and response systems,and for incorporating climate change in growthmanagement regulations. In particular, they stressedthe need to design solutions that will be acceptedby the local communities,and the need for coordi-nated institutional structures to manage the coastalresources. An example of a regional response planfor the Federated States of Micronesia is shown inthe table opposite.

CRUCIAL UNKNOWNS ANDRESEARCH NEEDSTo more effectively evaluate the potential conse-quences of climate variability and change impactson Islands,a number of research needs have beenidentified by the Pacific and Caribbean workshopparticipants. The following is a brief listing of someof the identified research needs:

• Regional-scale information on changes in wateravailability, frequency, and intensity of extremeevents such as hurricanes or typhoons,and interan-nual variability of climatic factors such as ENSO;specifically, downscaling by nesting regional climatemodels within global models.

• How to build local capacity to use such regionally-focused nested models and other data collectiontechniques and to interpret the information intouseful products for island decision makers andisland populations;

• How identified key parameters affect communi-ties,the built infrastructure including transportationaspects,businesses,and economic sectors,as well ascritical habitats and natural resources;

• The value of forecasting and understanding cli-mate variability and change and how that under-standing supports stakeholders in decision makingprocesses;

• Environmental and socioeconomic changes onislands and how the two are related to one another;the interactions of multiple stresses;and identifica-tion of the specific parts of society that are mostvulnerable to climate change and variability;

• An assessment of the status quo related to waterresource management,health care systems,andemergency systems;

• Implications of potential climate-change-responsepolicies for islands and island economies;

• How to improve risk analysis,monitoring systems,and evacuation systems on islands;and

• The impacts of changes in sea level,both short-term variations and long-term trends,on freshwatersupply and other potential threats to coastal com-munities and ecosystems;

• How to incorporate appropriate traditional knowl-edge and response strategies into response options.

343


Emanuel,K.A.,Thermodynamic control of hurricaneintensity, Nature, 401, 665-669,1999.

FSM (Federated States of Micronesia),National commu-nication to the United Nations Framework Conventionon Climate Change, Palikir, Pohnpei,FM. CountryCoordinator for national climate change program —John Mooteb,PICCAP Coordinator - Department ofEconomic Affairs,PO Box PS-Palikir, Pohnpei, FederatedStates of Micronesia,(email:[email protected]),1999.

FEMA (Federal Emergency Management Administration)Web site:http://www.fema.gov/library/cat4o5.CitedAug 1,2000.

Fritts,T. H.,The brown tree snake of Guam: Fact Sheet,USGS,1999,(http://159.189.24.10/btreesnk.htm)

Fritts,T. H.,and G. H.Rodda,Invasions of the browntree snake,in Our living resources: A Report on thedistribution, abundance, and health of US plants,animals, and ecosystems, edited by E.T. LaRoe, G. S.Farris,C.E.Puckett, P. D. Doran,and M. J. Mac,pp.454-456,US Department of the Interior, National BiologicalService,Washington,DC,1995

Glynn, P.,Coral reef bleaching: Facts, hypotheses,andimplications, Global Change Biology 2, 495-509,1996.

Glynn, P.W.,and W. H.de Weerdt,Elimination of tworeef-building hydrocorals following the 1982-1983 ElNiño warming event, Science 253(5015),69-71,1991.

Gray,W. M.,Atlantic seasonal hurricane frequency, Part1: El Niño and 30 mb Quasi-biennial oscillation influ-ences, Monthly Weather Review, 112, 1649-1668,1984.

Gray,W. M., J. D. Sheaffer, and C.W. Landsea,Climatetrends associated with multidecadal variability ofAtlantic hurricane activity, in Hurricanes: Climate andSocioeconomic Impacts, edited by H. F. Diaz and R.S.Pulwarty, Springer-Verlag,New York,292 pp.,1997.

Hamnett,M. P.,and C.L.Anderson,Impact of ENSOevents on tuna fisheries in the US-affiliated PacificIslands,Report of research supported by the Office ofGlobal Programs,National Oceanic and AtmosphericAdministration,Washington, D.C.,October 1999.

Hamnett,M. P.,C.L.Anderson,C.Guard,and T.A.Schroeder, Cumulative ENSO impacts:Tropical cyclonesfollowed by drought,in The Pacific ENSO ApplicationsCenter: Lessons learned for regional climate forecast-ing (revised report), Pacific ENSO Applications Center,Honolulu,HI,2000.

LITERATURE CITEDAlvarez,R.(Ed.),Proceedings of the Climate Changeand Extreme Events Workshop for the US SouthAtlantic Coastal and Caribbean Region,a regional work-shop,hosted by the International Hurricane Center,Florida International University, Miami,November1998.

American Red Cross website:(http://205.214.45.11/news/inthenews/98/10-16-98.html) and (http://205.214.45.11/news/inthe-news/98/10-16c-98.html) Accessed June 1,1999.

Bengtsson,L.,M.Botzet,and M.Esch, Numerical simu-lation of intense tropical storms,in Hurricanes:Climate and Socioeconomic Impacts, edited by H. F.Diaz and R.S.Pulwarty, Springer-Verlag,New York,292pp.,1997

Carlquist,S.,Hawaii: A natural history, published byPacific Tropical Botanical Garden,Lawai,Kauai,Hawaii,1980.

CNMI (Commonwealth of the Northern MarianaIslands) official website:(http://www.saipan.com/cnmi -info/info_pg4.htm#sec100),Accessed May 19,2000

Collins,M.,The El Niño Southern oscillation in the sec-ond Hadley Centre coupled model and its response togreenhouse warming, Journal of Climate, in press,2000.

Covich,A. P.,T.A.Crowl,S.L. Johnson,and F. N. Scatena,Drought effects on pool morphology and neotropicalstream benthos in Proceedings, Third InternationalSymposium on Tropical Hydrology and WaterResources, San Juan,Puerto Rico, July 13-17,1998,edit-ed by R.I.Segarra-Garcia,American Water ResourcesAssociation,Middleburg,VA.1998.

Cuddihy, L.W.,and C. P. Stone,Alteration of nativeHawaiian vegetation: Effects of humans,their activities,and introductions,University of Hawaii National ParkStudies Unit,Honolulu,138 pp.,1990.

Diaz,H. F.,and R.S.Pulwarty, Decadal climate variability,Atlantic hurricanes,and societal impacts: An overview,in Hurricanes: Climate and Socioeconomic Impacts,edited by H. F. Diaz and R.S.Pulwarty, Springer-Verlag,New York,292 pp.,1997.

Emanuel,K.A.,Climate variations and hurricane activi-ty: Some theoretical issues,in Hurricanes: Climateand Socioeconomic Impacts, edited by H. F. Diaz andR.S.Pulwarty, Springer-Verlag,New York,292 pp.,1997.344


Harrison,C.S., Seabirds of Hawaii: Natural Historyand Conservation, Cornell University Press,Ithaca,NY,249 pp.,1990.

Henderson-Sellers,A.,et al.,Tropical cyclones and glob-al climate change: A post- IPCC assessment, Bulletin ofthe American Meteorological Society, 79, 19-38,1998.

Holt,R.A.,An alliance of biodiversity, health, agricul-ture,and business interests for improved alien speciesmanagement in Hawaii,in Invasive Species andBiodiversity Management, edited by O.T. Sandlund, P.J. Schei,and A. Viken,pp.65-78,Kluwer AcademicPublishers,The Netherlands,1998.

(IITF) International Institute of Tropical Forestry web-site (http://www.fs.fed.us/global/iitf/cnf/Georges.html)Accessed June 1,1999.

IPCC,Intergovernmental Panel on Climate ChangeClimate Change 1995:Impacts,Adaptations andMitigation of Climate Change:Scientific-TechnicalAnalyses, edited by R.T.Watson,M.C.Zinyowera,and R.H.Moss,pp 289-324,Cambridge University Press,NY,1995,

Knutson,T. R.,and S.Manabe,Model assessment ofdecadal variability and trends in the tropical PacificOcean, Journal of Climate, 11, 2273-2296,1998.

Knutson,T. R.,and R.E.Tuleya,Increased hurricaneintensities with CO2-induced warming as simulatedusing the GFDL hurricane prediction system, ClimateDynamics, in press,2000.

Knutson,T. R.,R.E.Tuleya,and Y. Kurihara,Simulatedincrease of hurricane intensities in a CO2-warmed cli-mate, Science, 279, 1018-1020,1998.

Landsea,C.W., G. D. Bell,W. M.Gray, and S.B.Goldberg,The extremely active 1995 Atlantic Hurricane season:Environmental conditions and verification of seasonalforecasts, Monthly Weather Review, 126, 1174-1193,1995.

Landsea,C.W., N. Nicholls,W. M.Gray, and L.A. Avila,Downward trends in the frequency of intense Atlantichurricanes during the past five decades, GeophysicalResearch Letters, 23, 1697-1700,1996.

Landsea,C.W.,R.A.Pielke,Jr.,A.M.Mestas-Nunez,and J.A.Knaff, Atlantic basin hurricanes: Indices of climaticchanges, Climatic Change, 42, 89-129,1999.

Larsen,M.C.,Drought, rainfall,streamflow, and waterresources during the 1990s in Puerto Rico, PhysicalGeography, in press,2000.

Larsen,M.C.,and I.M.Concepcion,Water budgets offorested and agriculturally developed watersheds inPuerto Rico,in Proceedings, Third InternationalSymposium on Tropical Hydrology and WaterResources, San Juan,Puerto Rico, July 13-17,1998,edit-ed by R.I.Segarra-Garcia,American Water ResourcesAssociation,Middleburg,VA,199-204,1998.

Larsen,M.C.,and A.Simon,Rainfall-threshold condi-tions for landslides in a humid-tropical system,PuertoRico, Geografiska Annaler 75A(1-2),13-23,1993.

Larsen,M.C.,and A. J.Torres Sánchez,The frequencyand distribution of recent landslides in three montanetropical regions of Puerto Rico, Geomorphology, 24,309-331,1998.

Leatherman,S.,Opening plenary session speaker #3 pg.16 in Proceedings of the Climate Change and ExtremeEvents Workshop for the US South Atlantic Coastal andCaribbean Region,A regional workshop,hosted by theInternational Hurricane Center, Florida InternationalUniversity, Miami,Florida,November 1998.

Leatherman,S.International Hurricane Center, FloridaInternational University, home page –(http://www.ihc.fiu.edu/ihc/research/researchagenda.html),2000.

Lobban,C.S.,and M.Schefter, Tropical Pacific IslandEnvironments, University of Guam Press,Mangilao,Guam,US,399 pp.,1997.

Loope,L.L.,(http://www.hear.org/MiconiaInHawaii/MiconiaSummaryByLLL.htm ),2000.

Loope,L.L.,Hawaii and Pacific islands in Status andTrends of the Nation’s Biological Resources, edited byM. J. Mac, P.A.Opler, C.E.Puckett Haecker, and P. D.Doran,pp.747-774,US Geological Survey, Reston,Virginia,1998.

Loope,L.L.,and T.W. Giambelluca,Vulnerability ofisland tropical montane cloud forests to climatechange,with special reference to East Maui,Hawaii,Climatic Change, 39, 503-517,1998.

Loope,L.L.,and D. Mueller-Dombois,Characteristics ofinvaded islands,in Ecology of Biological Invasions: AGlobal Synthesis, edited by H.A.Mooney, et al., pp.257-280, John Wiley & Sons,Chichester, UnitedKingdom,1989.

345


Pielke,R.A.,Jr.,(http://www.esig.ucar.edu/HP_roger/sourcebook/hurri-cane.html),2000.

Pielke,R.A.,Jr.,and C.W. Landsea,La Nina,El Niño,andAtlantic hurricane damages in the United States,Bulletin of the American Meteorological Society,80(10),2027-2033,1999.

Potter, B.,E.L.Towle,and D. Brower,The Virgin IslandsNatural Hazard Mitigation Plan, by Island ResourcesFoundation for the Virgin Islands Territorial EmergencyManagement Agency (VITEMA).St.Thomas,December,1995.

Pounds, J.A.,M. P. L. Fogden,and J. H.Campbell,Biological response to climate change on a tropicalmountain, Nature, 398, 611-615,1999.

Rapaport,M.(Ed.), The Pacific Islands: Environmentand Society, The Bess Press,Honolulu,Hawaii,442 pp.,1999.

Reading,A. J.,Caribbean tropical storm activity over thepast four centuries, International Journal ofClimatology, 10, 365-376,1990.

Scatena, F. N.,and M.C.Larsen,Physical aspects ofHurricane Hugo in Puerto Rico, Biotropica, 23, 317-323,1991.

Snedaker, S.C.,Impact on mangroves,in ClimaticChange in the Inter-Americas Sea, edited by G.A.Maul,United Nations Environment Programme andIntergovernmental Oceanographic Commission EdwardArnold,London,United Kingdom, pp.282-305,1993.

Steadman, D.A.,Prehistoric extinctions of Pacificislands’birds: Biodiversity meets zooarcheology,Science, 267, 1123-1131,1995.


Titus, J. G.,R.A. Park,S. P. Leatherman, J. R.Weggel,M.S.Greene, P.W. Mausel,S.Brown,C.Gaunt,M.Trehan,andG.Yohe,Greenhouse effect and sea-level rise: The costof holding back the sea, Coastal Management, 19,171-204,1991. On web at (http://www.epa.gov/global-warming/publications/impacts/sealevel/cost_of_hold-ing.html)

Low, L.-L.,A.M.Shimada,S.L.Swartz,and M. P.Sissenwine,Marine resources,in Status and Trends ofthe Nation’s Biological Resources, edited by M. J. Mac,P.A.Opler, C.E.Puckett Haecker, and P. D. Doran,pp.775-866,US Geological Survey, Reston,Virginia,1998.

Markham,A.(Ed.), Potential impacts of climate changeon tropical forest ecosystems, Climatic Change, 39, 1-603,1998.

Meehl, G.A.,Vulnerability of fresh water resources toclimate change in the tropical Pacific region, Journalof Water, Air, and Soil Pollution, 92, 203-213,1996.

Meehl, G.A.,and W. M.Washington,El Niño-like climatechange in a model with increased atmospheric CO2concentrations, Nature, 382, 56-60,1996.

Meehl, G.A., F. Zwiers, J. Evans,T. Knutson,L.Mearns,and P.Whetton,Trends in extreme weather and climateevents: Issues related to modeling extremes in projec-tions of future climate change, Bulletin of theAmerican Meteorological Society, 81(3), 427-436,2000.

Meyer, J.-Y.,and J. Florence,Tahiti’s native flora endan -gered by the invasion of Miconia calvescens DC.(Melastomataceae), Journal of Biogeography, 23, 775-781,1997.

Mueller-Dombois, D.,and F. R. Fosberg, Vegetation ofthe Tropical Pacific Islands, Springer-Verlag,New York,733 pp.,1998.

National Hurricane Center,(http://www.nhc.noaa.gov/1995marilyn.html),accessed September, 2000.

(NOAA) National Oceanic and AtmosphericAdministration (http://www.ncdc.noaa.gov/o1/reports/georges/georges.html) Accessed May 26,1999.

O’Brien,S.T.,B. P. Hayden,and H.H.Shugart,Global cli-mate change,hurricanes,and a tropical forest, ClimaticChange, 22, 175-190,1992.

OIA (Office of Insular Affairs),US Department ofInterior,A Report on the State of the Islands,URL(http://www.doi.gov/oia/report99.html),1999.

Parkinson,R.W.,R.D. de Laune,and J.R.White,Holocenesea-level rise and the fate of mangrove forests withinthe Wider Caribbean region, Journal of CoastalResearch, 10, 1077-1086,1994.

346


US Department of State, Coral Bleaching,CoralMortality, and Global Climate Change:cited on 5/16/00(http://www.state.gov/www/global/global_issues/coral_reefs/990305_coralreef_rpt.html)

US Geological Survey (USGS),2000,HawaiianEcosystems at Risk (HEAR) project:(http://www.hear.org/)

US Geological Survey (USGS),1999a,USGS Hawaii FactSheet:(http://water.usgs.gov/pubs/FS/FS-012-99/)

US Geological Survey (USGS),1996,USGS Hawaii StateFact Sheet: (http://water.usgs.gov/pubs/FS/FS-011-96/)

US Geological Survey (USGS),1999b,USGS Fact Sheet040-99:Puerto Rico,(http://water.usgs.gov/pubs/FS/FS-040-99/).

Walker, L.R., D. J. Lodge, N.V. L.Brokaw, and R.B.Waide,An introduction to hurricanes in the Caribbean,Biotropica, 23, 313-316,1991.

Wiley, J.W.,and F. J.Vilella,Caribbean islands,in Statusand Trends of the Nation’s Biological Resources, edit-ed by M. J. Mac, P.A.Opler, C.E.Puckett Haecker, and P.D. Doran, pp.315-349,US Geological Survey, Reston,Virginia,1998.

Wilkinson,C., O. Linden,H.Cesar, G. Hodgson, J.Rubens,and A.E.Stong,Ecological and socioeconomicimpacts of 1998 coral mortality in the Indian Ocean:An ENSO impact and a warning of future change?,Ambio, 28, 188-196,1999.

Zack,Allen,and M.C.Larsen,Puerto Rico and the USVirgin Islands: Research & exploration, NationalGeographic Society, Water Issue, 9, 126-134,1993.

BACKGROUND REFERENCESAckerman, J. D.,L.R. Walker, F. N. Scatena,and J.Wunderle,Ecological effects of hurricanes, Bulletinof the Ecological Society of America, 72,178-180,1991.

Basnet,K.,Controls of environmental factors on pat-terns of montane rain forest in Puerto Rico, TropicalEcology, 34, 51-63,1993.

Basnet,K., G.E.Likens, F. N. Scatena,and A.E.Lugo,Hurricane Hugo:Damage to a tropical rain forest inPuerto Rico, Journal of Tropical Ecology, 8,47-55,1992.

Brokaw, N.V. L.,and J. S.Grear, Forest structurebefore and after Hurricane Hugo at three elevationsin the Luquillo Mountains,Puerto Rico, Biotropica,23, 386-392,1991.

Brokaw, N.V. L.,and L.R.Walker, Summary of theeffects of Caribbean hurricanes on vegetation,Biotropica, 23, 442-447,1991.

Felzer, B.,and P.Heard,Precipitation differencesamongst GCMs used for the US National Assessment,Journal of the American Water ResourcesAssociation, 35(6), 1327-1340,1999.

Frangi, J. L.,and A.E.Lugo,Hurricane damage to afloodplain forest in the Luquillo Mountains ofPuerto Rico, Biotropica, 23, 324-335,1991.

Guariguata,M.R.,Landslide disturbance and forestregeneration in the upper Luquillo Mountains ofPuerto Rico, Journal of Ecology, 78, 814-832,1990.

Herbert, P. J., J. D. Jarrell,and M.Mayfield,The dead-liest,costliest,and most intense hurricanes of thiscentury (and other frequently requested hurricanefacts). NOAA Technical Memorandum NWS TPC-1,National Hurricane Center, Miami,Florida,1996.(http://www.nhc.noaa.gov/pastcost.html)

Karl,T. R.,R.W. Knight,and N. Plummer,Trends inhigh-frequency climate variability in the twentiethcentury, Nature, 377, 217-220,1995.

Larsen,M.C.,and A.Simon,Landslides triggered byHurricane Hugo in eastern Puerto Rico,September1989, Caribbean Journal of Science, 28(3-4),113-125,1992.

Lobban,C.S.,and M.Schefter, Tropical PacificIsland Environments, University of Guam Press,Mangilao,Guam,US,399 pp.,1997.

Lodge,D. J.,and W. H.McDowell,Summary of ecosys-tem-level effects of Caribbean hurricanes,Biotropica, 23, 373-378,1991.

Loope,L.L., F.G. Howarth, F. Kraus,and T.K.Pratt,Newly emergent and future threats of alien speciesto Pacific landbirds and ecosystems.in Studies inAvian Biology, No. 22. Ecology, conservation, andmanagement of Hawaiian birds:a vanishing avi -fauna, edited by J. M.Scott et al.,published byCooper Ornithological Society,Allen Press,Lawrence,Kansas,in press,2000.

347


Vitousek, P. M.,Biological invasions and ecosystemproperties: Can species make a difference?,in Theecology of biological invasions of North Americaand Hawaii, edited by H.A.Mooney and J. Drake,pp.163-176,Springer-Verlag,New York,1986.

Walker, L.R., J.Voltzow, J. D.Ackerman, D. S.Fernández,and N. Fetcher, Immediate impact ofHurricane Hugo on a Puerto Rican rain forest,Ecology, 73,691-694,1992.

Wiley, J.W.,and F. J.Vilella,Caribbean islands,inStatus and Trends of the Nation’s BiologicalResources, edited by M. J. Mac, P.A.Opler, C.E.Puckett Haecker, and P. D. Doran,pp.315-349,USGeological Survey, Reston,Virginia,1998.

Willig,M.,and L.R.Walker, Disturbance in terrestrialecosystems: Salient themes,synthesis,and futuredirections,in edited by L.R.Walker, pp.747-767,Elsevier Science,Amsterdam,1999.

Zack,A.,and M.C.Larsen,Island hydrology: PuertoRico and the US Virgin Islands, NationalGeographic Research & Exploration.Water Issue,126-134,1994

Lugo,A.E.,Reconstructing hurricane passages overforests: A tool for understanding multiple-scaleresponses to disturbance, Trends in Ecology andEvolution, 10, 98-99,1995a.

Lugo,A.E.,Tropical forests: Their future and ourfuture,in Tropical forests: Management and ecolo -gy.Ecological Studies, edited by A.E.Lugo and C.Lowe,Springer-Verlag,New York,1995b.

Lugo,A.E.,Mangrove forests: A tough system toinvade but an easy one to rehabilitate, MarinePollution Bulletin, 37(8-12),427-430,1998.

Lugo,A.E.,Comparison of island and continentalecosystems with a focus on disturbance, ActaCientífica, 9(2-3),129-130,1999a.

Lugo,A.E.,Ecological aspects of catastrophes inCaribbean islands, Acta Científica, 2(1):24-31,1999b.

Pringle,C.M.,and F. N. Scatena,Aquatic ecosystemdeterioration in Latin America and the Caribbean,inManaged Ecosystems: The MesoamericanExperience, edited by L. U. Hatch and M.E.Swisher,Oxford University Press,New York,104-113,1999a.

Pringle,C.M.,and F. N. Scatena, Freshwater resourcedevelopment: Case studies from Puerto Rico andCosta Rica, Managed Ecosystems: TheMesoamerican Experience, edited by L. U.Hatchand M.E.Swisher, Oxford University Press,NewYork,Oxford,NY, chapter 13,1999b.

Rapaport,M.(Ed.), The Pacific Islands:Environment and Society, The Bess Press,Honolulu,Hawaii,442 pp.,1999.

Scatena, F. N.,The management of Luquillo cloud for-est ecosystems: Irreversible decisions in a non-sub-stitutable ecosystem,in Tropical Montane CloudForests.Proceedings of an InternationalSymposium, edited by L.S.Hamilton, F. N. Scatena,and J. Juvik,pp.191-198,East-West CenterPublications,Honolulu,Hawaii,1993.

Silver,W. L.,The potential effects of elevated CO2and climate change on tropical forest biogeochemi-cal cycling, Climatic Change, 39, 337-361,1998.

Steudler, P.A., J. M.Melillo,R. D. Bowden,M.S.Castro,and A.E.Lugo,The effects of natural and human dis-turbances on soil nitrogen dynamics and trace gasfluxes in a Puerto Rican wet forest, 356-363,1991.

348


ACKNOWLEDGMENTSMuch of the material for this chapter is based on

contributions from participants on and thoseworking with the:

Pacific Islands Workshop and Assessment TeamsEileen Shea*,East-West CenterMichael Hamnett*,University of HawaiiCheryl Anderson,University of HawaiiAnthony Barnston,NOAA,National Centers for

Environmental Prediction,Climate Prediction CenterJoseph Blanco,Office of the Governor (State of

Hawaii)Kelvin Char, Office of the Governor (State of

Hawaii) and NOAA National Marine FisheriesService, Pacific Islands Area Office

Delores Clark,NOAA National Weather Service,Pacific Region Office

Scott Clawson,Hawaii Hurricane Relief FundTony Costa, Pacific Ocean ProducersMargaret Cummisky, Office of the Honorable Daniel

K.Inouye,United States SenateTom Giambelluca,University of HawaiiCharles “Chip”Guard,University of GuamRichard Hagemeyer, NOAA National Weather

Service, Pacific Region OfficeAlan Hilton,NOAA Pacific ENSO Applications CenterHonorable Daniel K.Inouye,United States SenateDavid Kennard,FEMA Region IX, Pacific Area OfficeRoger Lukas,University of HawaiiFred Mackenzie,University of HawaiiClyde Mark,Outrigger Hotels and Resorts-HawaiiGerald Meehl,National Center for Atmospheric

ResearchJerry Norris, Pacific Basin Development CouncilDavid Penn,University of HawaiiJeff Polovina,NOAA National Marine Fisheries

ServiceRoy Price,Hawaii State Civil Defense (retired)Barry Raleigh,University of HawaiiKitty Simonds,Western Pacific Regional Fishery

Management CouncilPeter Vitousek,Stanford UniversityDiane Zachary, Maui Pacific Center

South Atlantic Coast and Caribbean Workshop TeamRicardo Alvarez,International Hurricane CenterKrishnan Dandapani,Florida International UniversityShahid Hamid,Florida International UniversityStephen Leatherman,International Hurricane CenterRichard Olson,International Hurricane CenterWalter Peacock,International Hurricane Center and

Laboratory for Social and Behavioral ResearchPaul Trimble,South Florida Water Management

District

Additional ContributorsRobert Cherry, USGCRP National Assessment

Coordination OfficeBenjamin Felzer, University Corporation for

Atmospheric ResearchMark Lander, University of GuamKatherin Slimak,USGCRP National Assessment

Coordination Office


349


351

CHAPTER 12

POTENTIAL CONSEQUENCES OF CLIMATEVARIABILITY AND CHANGE FOR NATIVEPEOPLES AND HOMELANDSSchuyler Houser1,Verna Teller2, Michael MacCracken3,4,Robert Gough5, and PatrickSpears6


Chapter Summary

Introduction

Historical Context

Geographical and Socioeconomic Context

Climate and Ecological Context

Scenarios for Future Climate

Key Issues

Tourism and Community Development

Human Health and Extreme Events

Rights to Water and Other Natural Resources

Subsistence Economies and Cultural Resources

Cultural Sites,Wildlife and Natural Resources

Coping and Adaptation Strategies

Enhance Education and Access to Information and Technology

Promote Local Land-use and Natural Resource Planning

Participate in Regional and National Discussions and Decision-making


Literature Cited

Acknowledgments

1 Lac Courte Oreilles Ojibwa Community College,Hayward,Wisconsin, 2 Isleta Pueblo, 3 Lawrence LivermoreNational Laboratory, on assignment to the National Assessment Coordination Office of the U. S.Global ChangeResearch Program, 4 Coordinating author for the National Assessment Synthesis Team , 5 Intertribal Council onUtility Policy, 6 Lakota

among Native peoples. These interests arisebecause the consequences will affect both theirreservation lands and the much larger land areasencompassed in the concept of Native homelands.While each tribe will face its own challenges,thischapter focuses on a few general issues facing largenumbers of Native peoples,particularly AmericanIndians. More region-specific issues are covered inthe various regional sections of the report,notablyin those dealing with the Pacific Northwest andAlaska.

Climate of Past Century

• During the 20th century, much of the western US,where most reservation lands are located,haswarmed several degrees Fahrenheit,contributingto apparent changes in the length of the seasons.

• After relatively dry periods in the central andwestern US during the first half of the 20th centu-ry, the second half has been wetter, with morerunoff in key river basins.


• During the 21st century, the Hadley and Canadianmodel scenarios both project that temperaturesin the western US are likely to rise significantly(typically 5 to 10˚F) with even larger increasespossible in the Great Plains and Alaska.

• Greater precipitation is projected for the south-western US,although projections are uncertainas to how far inland the increase is likely toextend.

• Although rainfall is likely to increase in manyregions,the higher temperatures will increaseevaporation,likely causing a soil moisture deficitin some regions that will tend to dry out forestsand grasslands.

CHAPTER SUMMARYContext

Native peoples,including American Indians and theindigenous peoples of Alaska,Hawaii,and the Pacificand Caribbean Islands,currently comprise almost1% of the US population. Formal tribal enrollmentstotal approximately two million individuals,morethan half of whom live on or adjacent to hundredsof reservations throughout the country, while therest live in cities,suburbs,and small rural communi-ties outside the boundaries of reservations. The fed-eral government recognizes the unique status ofmore than 565 tribal and Alaska Native governmentsas “domestic dependent nations.” The relationshipsbetween tribes and the federal government aredetermined by treaties, executive orders,tribal legis-lation,acts of Congress,and decisions of the federalcourts. These actions cover a range of issues thatwill be important in adapting to climate change,from responsibilities and governance to use andmaintenance of land and water resources.

Tribal land holdings in the 48 contiguous states cur-rently total about 56 million acres,or about 3% ofthe land. This area is approximately the size of thestate of Minnesota.Additionally,Alaska Native corpo-rations hold approximately 44 million acres of land.Despite the relatively extensive total land holdings,most individual reservations are small,supportingcommunities with populations of less than 2,000.Larger reservation populations,while unusual,doreach as high as 200,000 on the Navajo Reservation.

The federal government has recognized that tribesand tribal governments also have legal rights in ter-ritories that lie beyond the boundaries of theirrespective reservations. Treaties in the PacificNorthwest and the north-central states ofMinnesota,Wisconsin,and Michigan recognizerights of tribes to fish,hunt,and gather off-reserva-tion. Further, federal legislation has recognizedtribal interests in historical and cultural interestareas beyond reservation boundaries. These inter-est areas cover a significant fraction of the 48 con-tiguous states, generally matching the “NativeHomelands”that Native peoples inhabited prior toor since European settlement.

With the beginning of clearly observable climatechange,and because of the relationships of plants,water, and migrating wildlife with ecosystems out-side reservation boundaries,the potential conse-quences of climate change create significant interest

352


Key Findings

Sustaining economic vitality will require thoughtfulplanning because many reservation economies andtribal government program budgets depend heavilyon agriculture, forest products,and tradition- andrecreation-based tourism,which are all likely to beaffected as the climate shifts and warm extremesbecome more frequent. For example,hotter anddrier summer conditions are likely to affect recre-ational use of forest campgrounds and lakes.Economic diversification has the potential to reducethe existing vulnerability.

Preparing for the health and welfare implications ofunusual climate episodes is likely to become moreimportant as climatic patterns shift and landscapeschange. This is likely to be particularly importantbecause Native housing is typically more sensitive tothe prevailing climatic conditions than national aver-age for housing. As a result,increasing use of air-conditioning is not as ready a means of addressingan increasing frequency of very hot and dry condi-tions. In addition,increased dust and wildfiresmoke may well exacerbate respiratory conditions.

Ensuring stable water supplies for tribal lands andsurrounding users will require consideration of cli-mate variability and change in management of waterresources and in negotiations concerning Indianwater rights. Financial resources are also likely tobe needed to address infrastructure issues. It is pos-sible that increased precipitation in the Southwestcould help to diminish water concerns in that areawhile other areas are likely to face decreased wateravailability.

Although only a few tribal economies in Alaska andother regions are primarily based on subsistence,

many tribal communities depend on their environ-ment for many types of resources. A changing envi-ronment puts such resources at risk,which willaffect both sustenance and cultural dependence onenvironmental resources.

Sacred and historically significant sites (and theexperiences associated with them) and cultural tra-ditions are likely to be significantly af fected.Because some sites are located in vulnerable loca-tions, changes in climate and ecosystems are likelyto alter the site environment. Changes in the timingof animal migrations and changes in the seasonalappearance and abundance of both plants and ani-mals are also likely.

Coping Options

Increasing education about basic science issues andhealth and wellness as well as increasing scientificand technical expertise could help to build econom-ic resilience. Increased monitoring of ongoingchanges and improved projections of future changesare needed to improve the quality of planning andpreparing for climate change. Enhancing access toinformation and technology has the potential to alle-viate some of the stresses from climate variabilityand change.

Promoting and enabling local land use and naturalresource planning are likely to help create the con-ditions for prudent preparations for and responsesto climate change. Increasing participation ofNative peoples in regional and national decision-making is needed in recognition of the connected-ness of tribal reservations,surrounding lands,andlarger regional and even national landscapes.

353

Chapter 12 / Native Peoples and Homelands

354


INTRODUCTIONWith projections of significant changes in the cli-mate and in the character of the land and itsresources over the next century,American Indiansand the indigenous peoples of Alaska,Hawaii,andthe Pacific and Caribbean islands that are part of theUS will face special challenges. Although countingmethods vary, Native peoples currently comprisealmost 1% of the US population (see end note 1).Tribal enrollments total approximately two millionindividuals,of which about 1.2 million live on oradjacent to hundreds of reservations throughout thecountry. The other 40% of Native peoples live incities,suburbs,and small rural communities outsidethe boundaries of reser vations (BIA,2000a). Thefederal government recognizes the unique status ofmore than 565 tribal and Alaska Native governmentsas “domestic dependent nations.” The relationshipsbetween tribes and the federal government aredetermined by treaties, executive orders,tribal legis-lation,acts of Congress,and decisions of the federalcourts. These agreements cover a range of issuesthat will be important in adapting to climatechange,from responsibilities and governance,to useand maintenance of land and water resources.It important to note that many of the phenomena ofa changing climate addressed in this chapter areprecisely those which affect rural areas throughoutthe United States. At least half of the Native peoplesof the US live in reservations in rural areas. Why,then,if climatological phenomena are so similar,should there be a separate chapter addressing theissues of Native peoples? The answer is at leasttwofold.

First,the institutional,legal,economic,and politicalstructures in Indian country differ significantly fromthose of other rural communities. These structuresdetermine and constrain,to a great degree,theresources that Native peoples can deploy in address-ing issues that arise from changes in climate.

Secondly, different cultures provide their memberswith different kinds of tools for addressing life’s situ-ations:language,preferences of technology, respons-es to innovation,tastes in food,attitudes towardsstrangers, relationships to the physical environment.Native peoples,with traditional cultures as differentfrom each other as they might be from those ofSwedish-Americans,bring a wide variety of culturaltools and experience to issues of climate changeand adaptation. This chapter examines a number ofthose tools and their implications for adaptation tosignificant environmental changes.

The chapter is based on the Native Peoples/NativeHomelands portion of the Assessment (NPNH,2000),which is examining the potential for climate-related impacts that are likely to affect Native peo-ples and the interactions of these impacts with con-temporary Native cultures,communities,and futuregenerations. Although the diversity of land areasand tribal perspectives and situations makes general-ization difficult,Native peoples recognize thatbecoming more resilient to variations in the climateand preparing for and adapting to climate changewill require special attention. The preceding region-al sections of this report, for example the chapterson Alaska,the West,and the Pacific Northwest,high-light a number of region-specific issues. This sectionreports on aspects that reach beyond what citizensof particular regions will experience.

POTENTIAL CONSEQUENCES OF CLIMATEVARIABILITY AND CHANGE FOR NATIVE PEOPLESAND HOMELANDS

The Earth and Nature are inseparable from Indians themselves. Land sustains thelifestyle of countless tribes, even when little acreage is in production or has littleproductivity. Land sustains far more than subsistence, and indeed many Indiansrecognized decades ago the folly in attempting to sustain their daily needs onacreage that is marginal, both in resources and in per capita size. But land hasemotional meaning, a psychological significance for the Indian that is far moreintense than our nostalgic longing for the family farm and a rural way of life.

James Rattling Leaf,Sicangu Lakota


in climate at the continental and regional levels. Inaddition,as Native peoples were displaced andnational development occurred (Brown,1991),Native peoples experienced continental-scalechanges in their surroundings that are not unlikethe types of changes that all Americans may face incoming decades. The changes were substantial inmagnitude,surprising in their occurrence,unman-ageable by available technologies and existing formsof government,and generally irreversible. In thoserespects,the changes may provide insights of thekinds of transformations – cultural,economic,andsocial – that global changes in climate may bring,both for Native peoples and for America as a whole.

GEOGRAPHICAL ANDSOCIOECONOMIC CONTEXTThe lands held by Native peoples are extensive. Inaddition to the 40 million acres of land held byAlaska Natives,tribal lands in the rest of the US cur-rently total about 56 million acres (Department ofthe Interior, 1996). The lands outside Alaska amountto about 3% of the land area of the 48 contiguousstates,or approximately the size of the state ofMinnesota (see Figure 1). The largest portion ofIndian lands are held on reservations,so namedbecause they consist of lands that were reserved for

HISTORICAL CONTEXTOver the last 500 years,essential environmental bal-ances that sustained Native peoples in NorthAmerica for many millennia began to shift rapidly.Forests were cut for homesteads and farming. Alienplants replaced native grasslands. Dry lands flood-ed, rivers changed their courses,and ponds andswamps drained away as watercourses weredammed and channeled. Important providers ofnourishment and protection – buffalo,salmon,andshad – were harvested to near extinction. New andstrange creatures – horse,cow, pig,sheep,andpheasant – shoved aside nature’s important spiritualguardians and came to dominate local economies.Exotic new diseases eradicated whole villages.Tribal governments and relationships fell. Spiritualleaders lost their followers. Communities – evenentire tribal nations – were forced to relocate.

Five hundred years ago,the population of Nativepeoples in North America is thought to have rangedbetween about 10 and 18 million. Between 1500and 1890,the population of Native peoples on thecontinent was declining at an average rate ofbetween 500,000 and 850,000 individuals each 20-year generation,and by 1890 had dropped to only228,000 (Snipp,1991). Some opinion leaders pre-dicted that Native people would soon disappear.However, those who predict-ed the “vanishing of the RedMan”substantially underesti-mated the endurance andadaptability of Native peoples,and the strength of Nativeperspectives and values. Overthe last 100 years,the popula-tion of Native peoples hasgrown almost ten-fold asNative communities havebeen rebuilt;artists,craftwork-ers,and writers have created arenaissance of beauty andmeaning;and economic devel-opment has accelerated(Cornell,1998).

In that the global tempera-tures have varied relatively lit-tle over the past millennium,the environmental changesthat drastically altered thelives and circumstances ofNative peoples as a wholesince 1492 seem to have beenmore influenced by changes

Figure 1: Map of Indian lands in the conterminous United States (BIA, 2000b). Thelargest areas are located in the central to western US. See Color Plate Appendix 355

work on Indian lands because of the leasing of triballands to non-Indian farmers and ranchers – or, forexample,in the case of Agua Calienta,near PalmSprings,California, for commercial development.The leasing of reservation lands is a long- standingpractice and a vital source of income to thelandowners (Lawson,1982). Complicating mattersfurther, a major portion of the lands that were allot-ted to Indian heads of household are now managedeither by the Bureau of Indian Affairs (BIA) or bythe appropriate tribal government. This land is alsofrequently leased to non-Indian farmers or rancherswith the proceeds from the leases then being divid-ed among the descendents of the original allottee.Maps of land ownership and tribal jurisdiction onmany individual reservations thus resemble checker-boards, greatly complicating planning efforts. At thesame time,judicial decisions have sharply limitedthe jurisdiction of tribal governments and tribalcourts over non-Indians.

As a result,many tribes face severe legal difficultiesin creating or enforcing comprehensive plans forland use or natural resource management,a situa-tion that will complicate planning for climatechange. For example,if a tribal government createsan environmental code,enforcement over an entirewatershed or forest may be impossible without thevoluntary consent of non-Indian owners of propertywithin and outside of reservation boundaries. If atribe leases cropland, grazing rights,or timber tonon-Indians,environmental regulations can conceiv-ably be written into the terms of the leases,although long-term traditions are likely to be diffi-cult to change and the practical job of enforcingnew regulations is likely to stretch the resources ofsmall and understaf fed tribal governments (Getcheset al.,1998; Pevar, 1992).

Tribal governments also have some legal rights inlands beyond the boundaries of reservations – rightsthat may establish precedents for collaboration onissues involving climate and environmental changes.For example,the federal government has recognizedhistorical and cultural interests of tribes and tribalgovernments concerning broader regions,oftencalled “Native Homelands,”which include landsoccupied by Native peoples at present or in thepast. Within the pre-determined boundaries of his-torical and cultural interest areas (generally home-land areas inhabited by a particular Native peopleprior to contact with Europeans),tribes are entitled,for example,to establish claims to human remains,ifevidence of kinship or ancestry can be established.1

These historical and cultural interest areas cover a

the sole use and occupancy of Indian peoples fromthe vast expanses of land which were ceded to theUnited States government (Brown,1991). Propertyownership by Native peoples of the Pacific andCaribbean islands varies greatly because of a varietyof circumstances,including traditional rights andleadership and historical legal rights. As indicated inthe Islands chapter, however, some island lands areoverseen by clans with responsibility for steward-ship on behalf of their members,whereas on otherislands there are no longer reserved land rights.

By far the majority of reservations are small,bothgeographically and demographically, with popula-tions less than 2,000 (Tiller, 1996). These lands,although they are owned by tribes or individualIndian people,are held in trust for the owners bythe federal government,in much the same way thatbanks or other trustees hold property for heirs untilthey come of age and can assume personal manage-ment of the property. One result of this system oftrusteeship is that tribes and individual Indian peo-ple have had very limited control over the use,envi-ronmental management,or profits of their ownlands. For much of the 20th century, in fact,many ofthe decisions over these matters rested with the fed-eral government,not with the tribes themselves.Only in the last several decades have tribal govern-ments taken over more control of and responsibilityfor their lands.

From the most basic perspectives of the Americanlegal system, reservations may be viewed as jurisdic-tional islands,largely exempt from the laws of thestates that surround them. Tribal governments holdthe authority to levy taxes, regulate commerce,passand enforce civil and criminal codes and,in princi-ple, regulate the use of tribal lands and water. Whilefederal laws prevail,state authorities generally haveno rights of enforcement within these jurisdictionalislands.

However, from the perspective of tribal environmen-tal and land management policies and practices,theparadigm of reservations as islands is inadequate.First,the paradigm is inadequate environmentallybecause these “islands”are surrounded not byoceans,but by land,and so these lands are intimate-ly tied to the forests, grasslands, watersheds,andother ecosystems surrounding them;thus,thechanges on Native and surrounding lands are closelycoupled. Second,because many reservations haveconsiderable populations of non-Indians residingwithin the exterior borders of reservations (see endnote 2),the paradigm is inadequate administratively.Third,throughout the country, non-Indians also356


1For further information,see http://www.doi.gov/oait/docs/

eo13007.htm .

357


significant fraction of the 48 contiguous states,widening greatly the areas of interests of Native peo-ples.

CLIMATE AND ECOLOGICALCONTEXTReservations are present in all of the major ecosys-tems across the US,including the unique environ-ments represented by Alaska and the islands of thePacific and Caribbean regions. Native peoples havebeen experiencing the vagaries of climate on thiscontinent for many thousands of years. Theresource-rich environment created by the wood-lands of the northeastern,southeastern,and GreatLakes regions,especially the presence of deer, rab-bit,beaver, fish,berries,and many other resources,allowed tribes to occupy particular regions for longperiods. The Great Plains provided a source ofgrains and buffalo,along with fish and otherresources,but the wide range of climate extremescaused these tribes to be relatively mobile in orderto survive. The western US provided a wide array ofenvironments,from coastlines to mountains andriver valleys to deserts,and is now home to thegreatest number of Indian reservations. The Nativepeoples of Alaska have developed a lifestyle thatdepends,in large part,on very cold winters. Thoseliving on islands depend on the reliability of therains,and are adversely affected by either too muchor too little precipitation.

Their adaptability, and the histories of the experi-ences and the lessons that have been learned aboutcoping with climate fluctuations,have sustainedNative cultures through many generations. Nativeoral histories are now being linked with past cli-mate data derived from tree rings and other sourcesin ways that are enriching understanding of past cli-matic conditions. Oral histories often correlate withevents identified in the geological record,such asperiods with high or low rainfall,periods of warmor cold winters,and periods of flooding or drought(e.g.,Deloria,1997). What makes these historiesespecially valuable is that they often record not onlythe consequences of these climate fluctuations forpeople and for the environment around them,butalso the responses that helped the communities toadjust and survive. Thus,the retelling of theseevents by tribal elders has created a populace that isrelatively well informed about how to adapt toexternal stresses.

There are,however, two key changes that will limitthe application of some of the lessons to the issueof climate change,and are likely to create greatervulnerability than in the past. First,the changes inclimate are likely to be larger and more long lastingthan the fluctuations experienced in the past.Second,earlier coping strategies of Native peoples,on which many of their histories and traditions arebased, relied on shifting and moving,sometimesfrom one food source to another, sometimes fromone place to another, or sometimes to find alterna-tive sources of food and water, or to intersect withthe annual migrations of wildlife. In the Southwest,archeological evidence and Native oral historiesindicate that the great regional drought of the 13th

century caused the ancestral Pueblo people to aban-don their permanent homes in the mesas and val-leys of marginal areas (see box,“Lessons from theRelocation of the Ancestral Pueblo People in the13th Century”). When the ability to cope in oneplace was exceeded,Native peoples moved,laterreturning if and when climate permitted.

Over recent decades,Native peoples have beenobserving that changes in the environment havebeen occurring,some due to regional or global-scalechanges in the climate and some due to changingpractices of land management and use. Thesechanges are noteworthy, both because Native peo-ples are changing their practices and because of thenature of the observations themselves. In north-western Alaska, for example,elders lament that win-ter temperatures have become so warm (now typi-cally only -20˚F instead of -70˚F) that the traditionalecosystem on which they have depended for gener-ations is deteriorating and is no longer able to pro-vide the needed resources. In the Southwest, recol-lections by elders (corroborated by Army recordsfrom the early 1800s) are of valleys full of tall saca-ton grasslands,whereas the region now is scarredby deep arroyos and supports only sparse vegeta-tion,likely as a result of overgrazing and subsequentdrought. All across North America,tribal historiesindicate that change is occurring.

Native peoples today feel vulnerable to significantenvironmental changes because they are no longerable to cope easily with changes by relocating. Fewcontemporary tribes can afford the purchase oflarge tracts of new land,and federal laws hinder thetransfer or expansion of tribal jurisdiction. Tribestherefore see their traditional cultures directlyendangered by the magnitude of the projected cli-mate change. Had the ancient Anasazi been com-pelled to remain in place,the culture and way of lifeof an indigenous people that can be traced back

358


thousands of years would likely have been lost for-ever. This history provides a context for thinkingabout the potential consequences of future changesin the climate.

SCENARIOS OF FUTURECLIMATE

Most of the large Indian reservations in the US arelocated in the central and western United States.Coincidentally, most climate models project thatchanges in temperature and precipitation will begreater over the western than over the eastern US.The Canadian and Hadley model scenarios projectwarming of as much as 5 to 10˚F (3-6˚C) over the21st century over much of the western US,withmore warming during winters than during summersin many areas (see Chapter 1 for more details).

The warming is already being experienced by someNative peoples. Regions in Alaska are already expe-riencing significant warming,with some melting ofpermafrost and sea ice. Across northern NorthAmerica, warming is already lengthening the periodof plant growth. Such changes are expected tointensify. These trends will affect agriculture,wildlife migrations, flowering seasons,and otherimportant ecological events or occurrences that areclosely tied to seasonal change.

Some models project that,particularly in theSouthwest, warmer winters will also bring increas-ing wintertime precipitation,a rising snowline,andearlier springtime runoff, thereby affecting the tim-ing and volume of river flows. Increased precipita-tion would be expected to increase river runoff,which,if it occurs in seasons when the water canbe used or stored,may help to augment water sup-plies. Increased precipitation,however, can causeerosion on sparsely vegetated surfaces,distribute

Lessons from the Relocation of the Ancestral Pueblo Peoplein the 13th Century

About 2000 years ago,ancestors of today’s Pueblo people inhabited expansive areas of the Southwest,includ-ing much of what is now known as New Mexico,Arizona,Utah,Nevada,and Colorado. They developedadvanced architecture,moved water via extensive and complex irrigation systems to grow crops,made bas-kets and pottery, and created sophisticated social and religious structures. Today, those ancestral homelandsare dotted with the ruins of large villages that show deserted centers of trade and commerce, abandoned roadsystems and multi-story houses, forsaken art,and elaborate ceremonial chambers.

What happened to these communities is believed to have been a result,quite possibly in large part,ofchanges in the climate of this region. In the late 1200s, evidence from the growth rings of trees documentssignificant changes in the weather and climate. In particular, proxy evidence indicates that droughts becamemore frequent in southwestern North America. These conditions in turn would have led to disruption of theagricultural lifestyle of the communities dependent on the runoff from the southern Rocky Mountains.Although these regional climate changes were not particularly evident in the global climate record (see theClimate chapter’s description of the 1000-year climate record),the regional changes were apparently so largethat they contributed to the abandonment of these ancient cities. Cross-sections of soil combined withpollen records tell a similar story of changing conditions,with increasing erosion, ruined farmlands,and possi-bly a prehistoric “dust bowl.” As communities became unable to adapt their abilities to grow food to the dete-riorating environment,these changes apparently triggered a mass migration of what had been a typically sta-tionary, agriculture-based people.

Pueblo people living today pass on migration stories that recall the early journeys that brought them to thevillages they still occupy, mostly along the Rio Grande in New Mexico. Some clans migrated as far to the westas today’s Hopi tribe in Arizona. Others settled in the Pueblos of Zuni,Acoma,and Laguna. Since thesemoves, much has changed that limits the ability of Native peoples to move again to escape the consequencesof climate fluctuations and change. For the Pueblo people,Spaniards claimed their ancestral homelands inthe early 1500s. Subsequently, tribal land holdings were generally confined to the very limited boundariesthat persist to this day. Thus,if another disruptive climate fluctuation or change were to occur, the Pueblopeople could no longer easily move as a group to places with better conditions in order to preserve theirsocieties and cultures.

359


Subsistence Economies and Cultural Resources.Although only a few tribal economies in Alaska andother regions are primarily based on subsistence,many tribal communities depend on their environ-ment for many types of resources. A changing envi-ronment puts such resources at risk,which willaffect both sustenance and cultural dependence onenvironmental resources.

Cultural Sites, Wildlife, and Natural Resources. Theenvironment,both climate and the landscape,pro-vides an important sense of place for Native peo-ples,both for historical and cultural reasons. As theclimate changes and vegetation patterns and thepresence of wildlife and migrating species shift,thecultural context of Native peoples,who view them-selves as tightly coupled with and integral to thenatural environment,will be disrupted,with littlerecourse as the shifts occur.

1. Tourism and CommunityDevelopment

The most urgent priorities for tribal governmentsand communities over the past thirty years havebeen economic development and job creation.Many tribes have based their development initiativesaround land-based enterprises,including drylandand irrigation-based agriculture in the central andwestern US; forestry and forest products in the cen-tral, western,and sub-Arctic regions;and recreation-and tradition-based tourism in areas ranging fromHawaii and Alaska to the central, western,and south-western US. All of these activities are dependent onfavorable weather and climatic conditions. Althoughfew of these enterprises have generated enoughincome to develop a strong economic base forentire tribes,they are all vital to economic develop-ment for tribal communities. Adverse conditions,from severe winter storms to unusually wet or dryconditions,can have very severe economic effects,especially because tribal communities are alreadyeconomically stressed. The 1990 census indicatedthat 31.6% of all Indian people were below thepoverty line,compared to 13.1% of the total popula-tion (US Bureau of the Census,1990). From 1969though 1989,of the 23 reservations which have datafor the period,per capita income declined on 18reservations during the second decade (Trosper,1996). The sustained growth of the American econ-omy over the past decade has, for the most part,bypassed Native American households and reserva-tions.

Although none of the primary barriers to develop-ment on reservations is currently a result of long-

contaminants more widely, and create greater poten-tial for flash flooding. Moreover, warmer conditionswill also lead to increased evaporation,especially insummer, that will dry summer soils and vegetationin ways that may more than offset the increase inrunoff in some regions. For example,these changeswill likely lead to lower river and lake levels in thenorthern Great Plains and Great Lakes.

KEY ISSUESThe national workshop on Native Peoples/NativeHomelands identified a wide variety of issues facingNative peoples. Assessment studies of issues of spe-cific interest to Native peoples in particular regionsare just beginning;information on the first suchstudy, reporting on issues in the southwestern US,can be found at http://native-peoples.unm.edu/. Inthis report,issues that are primarily regional areaddressed in more detail in the chapters dealingwith particular regions (e.g.,Alaska) whereas issuesthat have broader national implications for Nativepeoples are addressed in this chapter. These issuesinclude:

Tourism and Community Development. In a num-ber of regions,Native economies are stronglydependent on tourism, agriculture,and other envi-ronmentally sensitive activities,so that shifts in tem-perature and precipitation and the ecosystems thatare based on the prevailing climate are likely torequire adjustments away from traditional activities.For example,hotter and drier summer conditionsare likely to limit recreational use of forest camp-grounds and lakes at lower elevations while poten-tially allowing more use at higher elevations.

Human Health and Extreme Events. As a result ofinsufficient economic development,Native housingis typically more sensitive to the prevailing climaticconditions than the national average. As a result,increasing use of air-conditioning is not as ready ameans of addressing an increasing frequency of veryhot and dry conditions. In addition,increased dustand wildfire smoke may exacerbate respiratory con-ditions.

Rights to Water and Other Natural Resources.Native water rights are established in a variety oftreaties, agreements,and court decisions. Providedfinancing is available,significant amounts of poten-tially irrigable land exist on reservations,and thepotential exercising of these rights along with pre-cipitation changes is likely to complicate waterresource allocations,negotiations,planning,andmanagement.

term climate change, recent variations and changesin weather patterns are requiring tribes to adapt andadjust their actions and plans. Tribes have alreadyidentified many local needs in response to anincreasing frequency of disruptions from severestorms,including improving or re-routing roads(many of which are unpaved), flood control andbank stabilization,providing new or more reliablewater and drainage services for industrial sites,strengthening communications links and power sup-plies,and altering schedules and calendars atschools and medical clinics to adapt to changingweather conditions. Tribal communities now expe-riencing sharp changes in precipitation patterns arealready modifying reservation infrastructures to dealwith such situations2. Projections of increasedoccurrence of extreme rains are likely to intensifythe need to improve infrastructure on reservations.Significant warming and the rise in the heat indexare likely to necessitate alterations in communitybuildings and water supply systems. For the future,particularly for tribes in the Southwest,longer-termchanges in water resources on which many tribesdepend are likely to have significant consequencesfor resource-based sectors such as agriculture andindustry that depend on stable water supplies.

Many tribes also are basing an increasing share oftheir economic development on recreation andtourism (Tiller, 1996). Tourism and recreation-basedactivities take advantage of the attractions of rivers,lakes,mountains, forests,and the other elements ofthe natural aesthetic beauty of reservations without,in most cases,causing long-term change. Culturaland historical sites and ceremonies of Native peo-ples can also be used to attract tourists. These activ-ities provide income while also encouraging the re-establishment of customs and traditions that hadbeen suppressed for many decades by federal poli-cies.

The economic viability of many aspects of recre-ation and tourism,however, is based on naturalattractions that depend on the prevailing climate –rivers and lakes provide water-based recreationopportunities, forests provide campsites and trails,and the wildlife,including migrating fish and birds,and flowering of plants,attract many visitors. As theclimate changes,these environments will change:reduced winter runoff from reduced snow cover is

360


likely to reduce the flow in many streams;driersummer conditions are likely to increase the firerisk and require closure of campgrounds;and thecombined effects of climate and ecosystem changeare likely to disrupt wildlife and plant communities.

Cultural traditions that draw visitors (and theirmoney) are often tied to the cycles of the seasonalrhythms in plant and animal life,with some tradi-tions honoring annual weather-related events thatare likely to be significantly af fected by climaticchange. The willingness to visit reservations forsuch events is dependent on the existence (andeven the perception) of a safe and healthy environ-ment. Such conditions can be disrupted by unusualclimatic occurrences. In 1993, for example,a han-tavirus outbreak associated with unusually heavyrains induced by El Niño conditions created a per-ception of an unhealthy environment (Schmaljohnand Hjelle,1997). The rains were conducive to highproduction of piñon nuts and other food sources,leading to an explosion in the hanta-bearing mousepopulation. This high mouse population thenencroached on human populations, resulting in anumber of virus-caused deaths (Engelthaler et al.,1999). This one event led to a significant reductionin tourist visits to the Southwest,especially toPueblo country, indicating how vulnerable sensitivetourist-based economies can be in the event of theoutbreak of rare,but frightening diseases, evenwhen these outbreaks are not occurring primarilyon Indian reservations. It is possible that a changetoward more intense El Niño/La Niña variationscould increase the likelihood of such conditions inthat such variations would likely upset the predator-prey relationships that develop during less variableconditions.

While some economic diversification of reservationeconomies is underway, and casino gambling isbecoming a basis for attracting tourists in a numberof regions,tribal economies tend to be more closelytied to their environments than is typical for theeconomies in regions where the reser vations arelocated. Because of this,tribal economies tend to bemore vulnerable to adverse changes and,on theother hand,more likely to benefit from climaticchanges that provide opportunities by enhancingwater availability.

Adaptation OptionsWith many tribal lands already under climatic stress,and with the economies of Indian Nations stronglydependent on climatically sensitive activities such asagriculture and tourism,the lands and economies ofmany Native peoples are vulnerable to climatechange. As a result of the less-diversified Native

2Repeatedly during the assessment process,participants raised issues

about the development of rural transportation systems.Reliable trans-portation is a major concern in many tribal communities,where manyindividuals cannot afford newer, more reliable,and fuel-efficient vehi-cles.Pr oviding effective public transportation in rural communities –most particularly communities with limited incomes – improves accessto employment,education,and medical services,and has the additionalbenefit of improving the efficiency of the use of fossil fuels,thus reduc-ing the production of greenhouse gases.

economies,a larger share of financial resources willlikely need to be devoted to ongoing adaptationthan is the case for society as a whole,thereby pos-sibly making tribal economies less competitive. Toaddress this issue,enhancement and diversificationof reservation resources and the strategic integra-tion of tribal economies with local non-Indianeconomies could help to make the tribal economiesmore resilient and sustainable.

At the same time, climate change and resulting poli -cy actions are likely to create some economicopportunities. For example,an increased demandfor renewable energy from wind and solar energycould create new opportunities because undevel-oped tribal lands could be an important resource forsuch energy in areas that are already developed. Fortribes in the Great Plains,this region could utilize itstremendous wind resource and development couldhelp to reduce greenhouse gas emissions as well asalleviate management problems created by demandsfor Missouri River hydropower, thereby helping tomaintain water levels for power generation,naviga-tion,and recreation. In addition,there may beopportunities for carbon sequestration.

2. Human Health and ExtremeEvents

The rural living conditions of many Native peoplesincrease the likelihood of impacts due to variationsin the weather. Due to the poor economic condi-tions,housing on many reservations is old and offerslimited protection from the environment. Althoughmany traditional structures are designed to takeadvantage of the natural warmth or coolness of thelandscape (e.g., by being located below ground,hav-ing thick walls,and being exposed to or shelteredfrom the sun),acclimation,both physiological andthrough use of appropriate clothing,is criticalbecause homes in many areas lack effective heatingand cooling systems. A recent study of energy con-sumption on Indian lands (EIA,2000) found thatreservation households are ten times more likely tobe without electricity (14.2%) than the nationalaverage (1.4%). While warming in colder regionswill alleviate some home heating needs,some accli-mation has already occurred so this will not signifi-cantly reduce stresses. In presently hot regions,however, there is likely to be a significant increasein natural heating that will require new acclimationand responses as new extremes are reached. Whilean increase in the presence of air-conditioned facili-ties would help,it would also require changes inbehavior toward a more indoor lifestyle along withimproved housing and access to electricity.

Climate change is also likely to exacerbate the deliv-ery of and need for health services. The delivery ofhealth care to rural communities throughout theUnited States has already been affected significantlyby widespread changes in demographic and eco-nomic patterns. Rates of depopulation in the leastdensely populated portions of the country are accel-erating. As communities lose population and eco-nomic conditions stagnate,the numbers of doctors,pharmacists,and other health care professionalsattracted to rural communities have been declining.Full-service health care is increasingly concentratedin regional centers,and consultation with a special-ist increasingly involves, for rural residents,an exten-sive trip.

Reservation populations,on the other hand,are con-tinuing to increase,as birth rates remain high,longevity increases,and tribal members move backto their home communities from urban areas. Theinstitutional structure of health care delivery toNative peoples,however, differs sharply from themarket-driven system that provides medical care forrural non-Indians. The federal government,as partof its trust responsibility for Indian people,providesthe health care systems for reservation residents.The Indian Health Service (IHS),part of the UnitedStates Department of Health and Human Services,isthe primary provider of medical services to Nativepeoples. IHS operates clinics,pharmacies,and hos-pitals in many tribal communities;in others,tribeshave contracted with the federal government tooperate health care facilities themselves.

Access to these health care facilities,however, is notalways easy for reservation residents,particularlyunder extreme weather conditions. A single hospi-tal, for example,serves the entire Rosebud Siouxreservation in South Dakota,where many roads areunpaved. The external boundaries of the reserva-tion are approximately 120 miles east-to-west,and60 miles north-to-south. The hospital is located inthe southwestern part of the reservation and accesscan be disrupted or cut-off by extreme precipitationconditions (whether rain or snow). Because of thedistances that many Native people must travel tohealth care facilities and the conditions of the roads,their access to health care is more subject to sharpvariations in the weather than those living andworking in cities,and extreme weather events arelikely to cause significant interruptions in access.

Changes in climate would also create new chal-lenges for community health. Drier summer condi-tions and the projected increase in forest fire inci-dence would likely lead to increased lofting of dust

361


362


and dust-borne organisms and an increase in forestfire incidence. The poorer air quality resulting fromincreases in smoke and dust would likely increaserespiratory illnesses such as asthma. Hypertensionand adult-onset diabetes are pandemic in tribal com-munities. As life spans increase,larger portions ofthe populations are becoming increasingly vulnera-ble to extreme temperatures,and increasinglydependent on uninterrupted access to such thera-peutic interventions as kidney dialysis. Further,direct and side effects of many standard medicationsare affected by climatic conditions.

Adaptation Options Health care delivery systems that serve Native com-munities will need to educate health professionals,paraprofessionals,and patients on the potentials fordehydration, overexposure to sunlight,and the needto restrict activity during hot weather. This need foreducation and appropriate care is intensifiedbecause many Indian people live in substandardhousing that exposes them to heat and cold. Muchof the most effective health education in Nativecommunities is carried out by Community HealthRepresentatives (CHRs),paraprofessionals employedand trained by the Indian Health Service or individ-ual tribal health programs. Specific training, focus-ing on climate-related health and wellness issues,and targeted towards CHRs could enhance use of asystem that has already demonstrated its abilities toreduce infant mortality rates and improve care forthe elderly among many Native peoples.

In addition,consideration should be given to thedesign and construction of appropriate housing forNative peoples. Most housing construction on manyreservations is financed through the federalDepartment of Housing and Urban Development.Tribal housing authorities,however, have taken overmuch of the responsibility for design,construction,and management of both rental and mutual self-help(purchasable by the occupant) housing units. Giventhe likelihood of significant changes in average tem-peratures,precipitation levels,and severe weatherevents,training and regular updates for decision-makers and technical specialists within tribal hous-ing authorities are needed to assist in improving thedesign,construction,or remodeling of homes toincrease resilience to extreme weather conditions.

3. Rights to Water and OtherNatural Resources

For Native peoples, water is recognized as a culturalas well as a physical necessity. Water is vital for lifeand livelihood,especially for those relying on the

resources provided by natural ecosystems. Water isnecessary for community use and the production offood as well as for fish, riparian plants,and wildlife.Water is particularly valued where it is most scarce,such as in the southwestern US. Prior to Europeansettlement, water was not owned,but was viewed asa gift to be shared by all. Settlement – both bytribes on reservations and by other Americans –brought increasing demands for water and conceptsof ownership that were not traditional to Nativepeoples. As a result, rights to water, including accessto sufficient quantity and quality, are now estab-lished and guaranteed by treaties,statutes,and deci-sional law. Changes in the amount,timing,and vari-ability of flows will affect the exercise of theserights.

Despite the many agreements and treaties,quantita-tive determination of existing Native rights to waterremains a contentious legal issue over much of thewestern US. Access to water was a key issue inmany of the treaties negotiated between tribes andthe US government,especially when relocation oftribes to reservations restricted or eliminated theiraccess to traditional homelands. These provisionsbecame the subject of litigation as expectationswere not met. The concept of federal reservedwater rights for Indians originated in the landmarkcase Winters v.United States [207 US 564 (1908)]involving withdrawal of water from the Milk Riveralong the Fort Belknap Reservation. The WintersDoctrine provided that in the treaties the federalgovernment entered into with various tribes,thegovernment had implicitly reserved a quantity ofwater necessary to supply the needs of the reserva-tion3. Under this provision,the federal govern-ment’s commitment of water rights would be para-mount because it had created the Indian reserva-tions and therefore had a fiduciary duty to protectwater implicitly reserved by and for the tribes. Thedoctrine was based on a set of principles underpin-ning reservation establishment that retained forIndian tribes all rights not explicitly waived.

Indian reserved water rights have become a subjectof considerable importance to tribes,states,the fed-eral government,and private water users due to:(i)the scarcity of water, particularly in the Southwestand Great Plains;(ii) the reality of drought condi-tions;and (iii) uncertainties arising because of fully(or even over) appropriated watersheds and interna-tional water commitments (some of which do notaccount for tribal water rights at all). The courts

3The Winters Doctrine later also formed the basis for federal reserved

water rights asserted by the government to obtain water for nationalforest, wetland,wildlife refuge,military, and other reservations.

have often had to resolve conflicting interests in theallocation of water rights,usually using a standardknown as “practicable irrigable acreage” for deter-mining the allotment to Indian reservations. Thisstandard quantifies Winters rights by providing thatallotments include sufficient water to provide foragriculture,livestock,domestic, recreational,cultur-al,and other uses. In addition, for some tribes,spe-cific legal language also reserves water to maintainin-stream flows necessary to sustain fish or riparianareas. Figure 2 compares the acreage of Indianlands that are currently being irrigated in elevenwestern states with the areas that could potentiallybe irrigated under the Winters doctrine. Quiteclearly, substantially increasing the areas of irrigatedlands would significantly increase the amount ofwater being withdrawn from current resources,increasing the competition for water among the var-ious users.

In some cases,tribes have not had the financialwherewithal to develop their water rights. In othercases,despite having high priority (i.e.,senior)rights to sufficient water, tribes have often had tocompete for access to water with non-tribal waterusers,including federal,state,and local govern-ments. This has required negotiations,which haveoften proven to be time-consuming,costly, and com-plex. The potential for changes in the amounts andtiming of water flows caused by climate change arelikely to add to the complexity of the allocations,negotiations, agreements,and management of waterresources.

As an indication of the complexity of the issuesinvolved,consider the Southwest,where climatemodels project that there is likely to be additionalprecipitation (most probably as rain) in winter, butwith earlier snowmelt and generally warmer anddrier conditions during the summer. If this occurs,overall winter runoff is likely to increase while sum-mer runoff is likely to decrease significantly. Whilean increase in annual precipitation could ease over-all management of water resources (at least to theextent that increases in the amount of vegetation donot counterbalance the likely increase in runoff),ifthe extra runoff in winter is not retained in reser-voirs,then the increased needs for water in summerwould likely increase water demands. Further, waterstorage that benefits some could be detrimental toothers,an example of which is described in the boxon “Shifting Ecosystem Boundaries”later in thischapter. With Native peoples exercising their histor-ical water rights more fully, and with federal policiesrequiring additional amounts of water to protect theenvironment,the amount of water available for irri-gation,communities,and other uses is likely to

363


decline,particularly during years in which climatefluctuations lead to reduced overall water flows.Water quality is a major issue that is coupled to theissues of water rights and water quantity;as waterquantity changes, water quality is likely to change aswell. Water quality affects everything from the envi-ronment for fish to the purity of drinking water to thequality of aquifers. In addition,because of the culturalconnection of Native peoples, water pollution andpoor water quality can have unusual ramifications. Inthe Northeast, for example,streams have been pollut-ed by persistent organic pollutants and heavy metals,so fishing and fish consumption are not permitted.These prohibitions affect not only the diets of Nativepeoples who observe such restrictions (many do notfor cultural reasons),but the restrictions have alsoreduced the opportunity for intergenerational connec-tion while fishing. In the Southwest, water purity hasbecome an issue based on water’s use in religious cer-emonies (see box“Ensuring Water Quality forReligious Ceremonies”). The prospect of diminishedsummertime water flows,along with more intermit-tent flows of some streams as a result of an increase infrequency and intensity of extreme rains,introducesthe potential for water quality to become an issue insome regions.

Adaptation OptionsWhere water is scarce,it merits careful stewardship.Improving the efficiency of water capture and use

Figure 2: Comparison of existing acreage being irrigated on Indianlands in eleven western states with the maximum acreage thatcould potentially be integrated based on the Winters doctrine that isapplied for determining Indian water rights (Riebsame et al., 1997)See Color Plate Appendix

Currently Irrigated vs. Potentially Irrigable Indian Land

resources for Native peoples for thousands of years.Forests, grasslands,streams,coastal zones,and morehave provided,and for many groups still provide,substantial amounts of food, fiber, fish,medicines,and culturally important materials. Native traditionsare also very closely tied to natural events andresources,including migrating birds and fowl,landanimals, fish,and medicinal plants,creating animportant cultural link to the land. Indeed,subsis-tence economies were the predominant form ofcommunity organization in North America prior tothe colonization of the continent by Europeans. Itis clear from oral traditions of Native peoples them-selves throughout the continent,and from theaccounts of the first Europeans who contactedthem,that subsistence economies were able to sus-tain communities in lives of comfort, relative stabili-ty, and abundance,with sophisticated artistic andintellectual traditions.

With the spread of the market economy across theUS,especially during the 20th century, subsistenceeconomies began to disappear as the resources nec-essary to support them were absorbed by commer-cial markets. By the end of the 20th century, subsis-tence economies among Native peoples were signif-

through more efficient irrigation practices andchoosing and growing crops that need less waterare steps that can be taken now, provided resourcesare available. As an additional possibility, increaseduse of cisterns and other water harvesting tech-niques of the type used several centuries ago bytribes living in the Southwest could also provide avery localized means of conserving water.Increasing in-ground storage of water through artifi-cial recharge is another possibility. Other historicalpractices of Pueblo cultures may also be applicable;mapping of ancient land use practices show exten-sive use of water management techniques,includinggridding of gardens to slow runoff, pebble mulchfields to reduce evaporation,stone-lined drainagechannels to reduce leakage,and terraces on hillsidesto retain water and prevent erosion. Quite clearly,however, the climate changes,ensuring the reliabili-ty and availability of water resources for tribal landsand surrounding users will require special consider-ation.

4. Subsistence Economies andCultural Resources

Native lands have provided a wide variety of

364


Ensuring Water Quality and Quantity forLegally Protected Religious Ceremonies and Cultural Practices

In 1986,the US Environmental Protection Agency (EPA) was authorized to treat tribes as states for purposesof grant and contract assistance, regulatory program development,and permitting and enforcement. This hasallowed some tribes to propose water quality standards. As for other states,these standards must then beapproved by EPA. One example where this has occurred is the Pueblo of Isleta,which is located immediatelysouth of Albuquerque,New Mexico. This was the first Native government in the US to obtain approval andTreatment-as-State (TAS) status from the EPA,pursuant to Section 518 of the Clean Water Act.

When the Pueblo of Isleta acted to protect its water quality, the City of Albuquerque sued the EPA for approv-ing Isleta’s standards (which are more stringent than off-reservation standards) in Federal District Court. InOctober 1997,the US Supreme Court upheld the tribe’s standards as approved by the lower court. The casewas decided in the favor of the Isleta Pueblo,with reasoning that has implications for Native governmentsthroughout the country. Isleta Pueblo’s standards were based on three significant use designations:use forirrigation,use for recreation,and use for religious ceremonies. The last reason is important to emphasizebecause no Native government had ever before asserted its right to religious freedom for the protection of itswaters. The Pueblo’s contention was based on the fact that tribal religious ceremonies (including bathing inthe river) were adversely impacted by the contamination of the river water by toxic discharges fromAlbuquerque’s run-off and from the municipal waste treatment facility located four miles north of the Pueblo.Although the city claimed that the tribe’s standards were arbitrary and capricious and would create an undueburden on the city to comply, the court upheld the Pueblo’s water quality standards.

With projections indicating that climate change is likely to lead to changes in the amounts and timing ofwater flows,it is possible that changes in water quality and quantity, especially during the summer, couldbecome an issue in other regions. Lower lake levels in the upper Great Lakes region and reduced flows inwestern rivers may be examples of where this issue could arise.

icant as a basis for family life only in the far northof Canada and Alaska,where some communitiesstill support themselves by a combination of sub-sistence, welfare and market economies;mixedmarket-subsistence-welfare economies also exist insome tribal communities in the conterminous US.

The values that have supported subsistenceeconomies have persisted,however, and this is amajor feature that distinguishes Native peoplesfrom other contemporary rural residents. Theethics embodied in these subsistence systems con-tinue to form the core set of values in many mod-ern Native communities, even as they are also inte-grated,in varying degrees,into mainstream marketeconomies. These values are incorporated into,and reinforced by spiritual teachings,moral princi-ples,and community and family relationships. Tothe extent that these values continue to shape con-temporary attitudes and relationships of Nativepeoples,they form a crucial part of the treasury ofvalues on which the Nation may draw in address-ing global climate change (see box,“PremisesSupporting Sustainable Subsistence Economies”).

Relying on a closed system involving primarily localresources rather than on a global network that pro-vides food and goods introduces much greater ele-ments of risk. Membership in a community pro-vides one means of limiting risk,because eachhousehold has more or less equal opportunities toshare in the community’s resources.4 Under suchcircumstances,special provisions are often made forindividuals such as the elderly and the sick whomight be unable to contribute regularly to the com-munity’s present food supply (Axelrod,1984,1997;Brams,1996). Prosperity, in this economic system,ismeasured not exclusively in terms of accumulatedpossessions – food, fields,and horses – but in termsof personal relationships. Having a large andhealthy family, respectful and aware of their obliga-tions,provides assurance of mutual support by

365


Premises Supporting Sustainable Subsistence Economies

The most important differences between market economies and the subsistence economies of Native peoplesinvolve concepts of surplus,accumulation, ownership,private property, individuality, and community.Participation in a sustainable subsistence economy demands knowledge,attitudes,behaviors,and expectationsthat differ significantly from those which work in market economies. The basic premises that have emergedand still generally prevail include recognition that:

Sustainable subsistence activities require traditional ecological knowledge acquired over generations (LaDuke,1994).

Subsistence communities are fairly, though not completely, closed economic systems. For day-to-day necessi-ties,people rely primarily on goods and services produced within their own community. Other goodsmight be acquired from external sources through trading (or, in earlier times, raiding),but sustenancedepends on work done within the community.

Most subsistence communities use a variety of food sources and methods of food production,including hunt-ing, fishing,gathering,and agriculture.

Success in food production varies from individual to individual within each community, and varies for eachindividual over time.

Food preservation technologies are often limited primarily to drying and smoking,technologies that do notsupport the storage of large amounts of foodstuffs over several years.

Food is sought in the amounts needed to sustain the community, with perhaps a small surplus for trading.Hoarding of food makes very little sense because it generally cannot be preserved,and so is actively dis-couraged.

Subsistence communities are small enough so that everyone within an individual community knows andacknowledges their relationships with everyone else. Kinship ties,in particular, are widely recognized(Akerlof, 1984;Schelling,1978,1984).

Family relationships carry with them obligations of mutual support and sharing of resources.Use of resources by the current generation must take into account the needs of future generations. An indi-

vidual might gather, harvest,or hunt enough for household and family use,but must leave enough toensure regeneration of renewable resources.

4The Lakota (western Sioux) begin each traditional religious ceremony

with the phrase “Mitakuye Owasin”which can be translated as “All myrelatives”or more freely “We are all related.” In the world-view of manyNative Peoples,this concept includes not only all of humanity, but alllife,and physical objects including rocks,the earth,stars and water.Many traditional Native spiritual leaders support the concept that alllife shares a foundation in DNA,and all of life and physical mattershares common structures at the atomic and molecular level.

specialization or over-dependence on single sourcesof food. In the past, woodlands tribes, for example,raised corn,beans,and squash,while also gatheringwild rice,berries,acorns and nuts,wild turnips andonions,and other edible and medicinal roots andplants (Vennum,1988;Ritzenthaler and Ritzenthaler,1991). These tribes also fished,and hunted largeand small animals and waterfowl. Pueblo communi-ties developed extensive systems of agriculture,butwere likewise active hunters. Inuit communitiesstill rely on caribou,whaling, fishing,and gatheringof plants and berries for balance in their diets. Foreach of these economic systems,the likelihood ofall sources of food failing simultaneously in a rela-tively quiescent climatic period is considerably lessthan the likelihood of one source declining for aparticular period (Trosper, 1999).

During the Native Peoples/Native Homelands work-shop,Native peoples from the Arctic and sub-Arcticpresented substantial evidence that their communi-ties are immediately jeopardized by changes in glob-al climate. In Alaska (see box “A Circle of Life:ALesson from Alaska”), rapid warming,and the envi-ronmental consequences it brings,started about 30years ago,and the lives of Native peoples there arealready being seriously affected (Gibson and

spreading widely the responsibilities for care andsustenance (Olson,1965;Hardin,1982;Houser andWhite Hat,1993).

Generosity, in this economy, is a highly prized virtue.When generosity is linked with the obligation forreciprocity, the development of positive human rela-tionships through sharing becomes the most effec-tive way to limit risk for any particular family orindividual. Among many Native peoples,there werehistorically – and continue to be – regular, institu-tionalized opportunities to show generosity, by giv-ing food and material gifts to members of the com-munity. During dances and special celebrations,Pueblo families open their homes to invited guestsand community members for home-cooked feasts,and give away baskets of food and cloth to honorfamily members whom are participating in cere-monies. Northern Plains’tribes hold giveaways inhonor of particular individuals – the honoree is cele-brated,not by receiving presents,but by having giftsgiven to others by the family in his or her name.Families may save for a year or more in order to pro-vide appropriate honor to a family member.

Traditional subsistence economies use a secondmethod of managing risk – the avoidance of over-

366


A Circle of Life: A Lesson from Alaska

Caleb Pungoyiwi is a Yupik Eskimo who moves back and forth between Alaska and Siberia in pursuit of wal-rus and other sea mammals. Gathering food directly from the land and the sea makes the Yupiks very carefulobservers of their surroundings. In recent years,they have noticed that the walrus are thinner, their blubberless nutritious,and the oil from walrus fat does not burn as brightly in their lamps as in times of old. At thesame time,they have noticed that there are fewer and weaker seals. The Yupik hunters have had to go fartherand farther from shore to reach the ice pack to find the newborn seals that are being fed fish from nearbywaters by their parents. Concurrently, scientists have observed that the sea ice over much of the Arctic isthinner and melting back,with the changes encompassing a broader area than that observed by the Yupiksearlier (Rothrock et al.,1999).

Both the Yupiks and the scientists have come to understand the intertwined chain of events that is occurring(Tynan and DeMaster, 1997). The retreat of the sea ice due to large-scale warming has reduced the platformthat seals and walrus have used to rest between searches for fish and mussels; weakened and less productive,they provide less sustenance for both the Yupiks and the whales. The Yupiks have also observed some killerwhales eating sea otters,an unusual shift in the whales’diet apparently brought on by the reduced number ofseals. The loss of sea otters is important because sea otters control the number of sea urchins.With fewer seaotters,there are more urchins and therefore less kelp (which the urchins eat). And with less kelp,there isreduced habitat for fish;it is fish that would normally be the major food source for the whales as well as theYupiks,the Inupiats,and other Northern peoples. As another part of this ecological continuum,the sea icequiets ocean waters during winter storms,helping protect young fish;it also accumulates nutrients that,whenthe ice melts,create a springtime algal bloom on which the fish feed at a critical stage in their development.What has occurred may seem like only a little warming in a very cold place,but as the Native traditions makeclear, everything is woven together – disruption in one place affects everything else (NPNH,2000;Moreno,1999;Huntington,2000).

Schullinger, 1998). Changes in climate,coupled withother human influences,are now becoming moreand more rapid in other regions,with projections ofmuch more rapid change in the future. Thesechanges are likely to bring much larger changes inland cover and wildlife than have occurred in thepast. As the world warms,as precipitation patternsshift,as sea level changes,as mountain glaciers andsea ice melt (IPCC,1996a),just as in Alaska,ecosys-tems elsewhere are likely to change in complex andoften unexpected ways,affecting the resources thatcan be drawn from them (IPCC,1996b).

In the Plains, warmer winter conditions are alreadyfavoring certain types of grasses,thereby changingthe mix of vegetation types. The distributions,tim-ing of migrations,and abundances of waterfowl andother birds are also changing (Sorenson et al.,1998;Price,2001;T. Root,Univ. of Michigan,unpublisheddata). These changes have been and will be affectednot only by temperature and precipitation changes,but also by changes in the timing of the ripening ofplants and crops,and other ecosystem factors onwhich Native peoples depend. How much longerthese types of changes can go on without a seriousdisruption of the services and diversity provided bythe various ecosystems that support subsistenceeconomies is not known,although there are signsthat some systems are already being seriouslystressed.

Adaptation OptionsWhile local environments have changed significantlyover past centuries, changes have often been slowenough to allow adjustment or for local environ-ments to be managed. Historically, Native peoplesactively worked to manage their environments inways that led to desired and productive results.Native peoples were not passive inhabitants of theirhomelands,simply fitting into niches convenientlyprovided by a supportive environment. Tribes useda variety of consciously developed technologies andculturally based choices to improve opportunitiesfor obtaining resources. The Paiutes,Hopi,Apaches,and Tohono O’odham,all lived in desert environ-ments,but employed significantly different methodsof land use. The Paiutes based their subsistence ona wide variety of plants, fish,and animals,and tookadvantage of whichever food supplies were mostabundant, even if this meant making short migra-tions to take advantage of each season’s particularopportunities. Boundaries between the variousPaiute bands were relatively flexible,and permittedhunting and gathering over relatively extensive areasof the Great Basin. The Hopi,in contrast,settled invillages on Black Mesa and created permanent fields

that sustained repeated harvests of corn,squash,andbeans. The Apaches,on the other hand,used fire asa technique for hunting,driving animals in a singledirection for harvest. Regular burning of huntingareas (a technique also used by the Paiutes) had theadditional effect of promoting plant growth andimproving forage for game animals (White andCronon,1988). The ancestors of the modernTohono O’odham developed a sophisticated systemof irrigation to support extensive agriculture. Manyof these techniques are no longer available or arenot likely to be adequate,however, to sustain subsis-tence economies in the event of the large changesin climate that are projected.

When changes were rapid and Native peoples werefaced with inadequate resources,several approacheswere used to adapt. Historically, many tribal com-munities could rely on migration to adapt to chang-ing resource bases in any particular local area.However, the establishment of reservations has limit-ed the option of entire tribes moving to more hos-pitable locations to seek water, cropland, forests,orcooler temperatures. While individuals can pursuethat option,it is not an option that is available tomost tribes because they are tied to where they areby land ownership and governance issues.

When Native peoples were challenged by radicalchanges in their physical environments,a secondapproach was to incorporate new technologies. Asthe Lakota (western Sioux) moved from forestsaround the Great Lakes to the Great Plains,theyfound the most suitable agricultural lands – the bot-tom lands along the Missouri River – already occu-pied by residents of large and stable villages. Withintwo generations,however, they adopted the com-plex technologies of horsemanship and the gun.This rapid adaptation provided the Lakota with bothmilitary and economic advantages,and provided thematerial foundation for a prosperous and vital cul-ture (Houser, 1995). This willingness to absorb newtechnologies,new materials,and new ways of doingthings forms a common theme in the histories ofmany Native peoples. Cloth (and the complexitiesof sewing woven textiles) replaced hides. Glassbeads from Bohemia replaced porcupine quills.Aniline dyes replaced vegetable colorings. Steelknives replaced stone or bone implements. Cottonthread replaced sinew. For the future,adopting newtechnologies is likely to be the only means for deal-ing with the disruptions to the traditional subsis-tence economies.

367


deeper life experiences. A variety of indigenousplants and animals,including migrating fish andwaterfowl,provide many tribes with sustenance,asindicated in the previous section,and essential com-ponents of many cultural traditions. As climateshifts,the optimal habitats for various plants (includ-ing medicinal plants) and fish runs are likely to shift.These changes are,in turn,likely to lead to a declin-ing presence of some plants while other plantsbecome more abundant,altering the resource baseand cultural experience for many tribal communi-ties. At deeper levels,humans’whole experience oftheir environment is likely to diverge from what hasbeen sustained through many generations via histor-ical and religious traditions. For Native peoples,externally driven climate change is likely to disruptthe long history of intimate association with theenvironment.

Central to the worldviews of Native peoples is anacknowledgement of kinship with all of creation.Through honoring and paying close attention totheir relatives,no matter how those relationshipsare defined,Native peoples have acquired and con-tinue to draw strength from unique insights aboutthe interactions of climate and the environmentalhealth of their homelands. These insights can havevery practical significance. For example,in the1970s and 80s,elders on the Rosebud SiouxReservation in South Dakota raised many questionsabout the potential implications of proposed effortsto exterminate prairie dogs in the sparsely settledwestern portions of their reservation. Although theelders expressed their objections in the language of

5. Cultural Sites, Wildlife, andNatural Resources

Ceremonial and historic sites, graves and archeologi-cal locations,special mountain and riverine environ-ments,and seasonal cycles and migrations are cen-tral parts of the cultural traditions and traditionalindigenous knowledge to Native peoples. Takentogether, atmospheric conditions and the characterof local landscapes – both the vegetation cover andthe wildlife – help to shape people’s sense of placeand how they relate to what surrounds them. WhileNative peoples have no monopoly among Americanson love of land, water, wildlife,and the sea,theirinterests start from different premises and havedeveloped over thousands of years of living on thiscontinent. As a result,the connections of Nativepeoples to their homelands differ, at fundamentallevels,from the kinds of relationships developed indensely populated suburban and urban environ-ments. These differences are frequently explained inspiritual terms,although the differences also includetraditional ecological and intellectual knowledgeand historical familiarities. These understandingsand relationships have been,and continue to be,transmitted orally and through ceremonial formsthat carry the interconnections of nature and histo-ries forward to future generations (Brody, 1982;Goodman,1990;Hiss,1991;Gallagher, 1993;Basso,1996;Bordewich,1996).

Changes in climate and in ecosystems in thedecades ahead are likely to have consequences andinfluences that are both practical and that affect

368


SONG OF THE SKY LOOM

Oh our Mother the Earth, oh our Father the SkyYour children are we, and with tired backsWe bring you the gifts that you love.Then weave for us a garment of brightness;May the warp be the white light of morning,May the weft be the red light of evening,May the fringes be the falling rain,May the border be the standing rainbow.Thus weave for us a garment of brightnessThat we may walk fittingly where birds sing,That we may walk fittingly where grass is green,Oh our Mother the Earth, oh our Father the Sky!

Found in “Songs of the Tewa”as translated by Herbert Joseph Spinden, copyright 1933.

Lakota spirituality, which values balance in natureand emphasizes the importance of each part of cre-ation,wildlife biologists have since come to recog-nize that prairie dogs are a “keystone species”thatplays a pivotal role in the maintenance of ecosys-tems on the Great Plains.5 Additional examples ofthe value of this close local knowledge are abun-dant:the exceptionally successful forestry manage-

ment of the Menominee Nation;the care of Creebands in harvesting animals to ensure that all gener-ations of game and fish survive in sufficient quanti-ties to ensure the continuity of all species.

Climate change will bring changes to the landscapesand wildlife that are important to Native peoples,changing the surroundings in ways that will change

369


Shifting Ecosystem Boundaries and the Sense of Place

Anyone who has ever walked up a mountain has experienced the fact that the climate at a particular altitudedetermines the vegetation. Because of this,mountaintops have their own particular ecosystems that are iso-lated like islands from surrounding areas. Although not unique to this region,the very large reservations ofsouthwestern New Mexico and southeastern Arizona are home to many “sky islands”that rise high above thedesert floor, some of them reaching to heights exceeding 10,000 feet. These unique regions,which are natu-ral parks in a sea of desert scrub,add greatly to the quality of life,and their biodiversity and genetic richnessis among the greatest to be found anywhere on the continent (Arias Rojo et al.,1999). These environmentsare particularly meaningful, even sacred,to many of the tribes,being environments with particular types ofplants and wildlife that are important culturally and medicinally. At the same time,they are particularly vul-nerable to climate change because of their unique settings in a largely arid region. As climate warms,ecologi-cal communities from the bottom to top will be disrupted as plant species gradually move up-slope,each attheir own rate. The existing mountain ecosystems will be forced into ever- diminishing areas of land andthose species already at the mountaintops will become extinct because they cannot migrate easily to distant,higher mountains.

Ecosystem shifts will occur not only in the vertical,but also in the horizontal dimension. For centuries,theAnishinaabeg (Ojibway or Chippewa) who live around Lake Superior and along the upper Mississippi Riverhave depended upon the natural resources of the forests,lakes,and rivers of the region (Vennum,1988).Many of the reservation locations were selected to ensure access to culturally significant resources,such asmaple sugar bushes and wild rice beds,whose locations were thought to be fixed. As drier summer condi-tions cause the western prairies to shift eastward toward the western Great Lakes,the extents of maple,birch,and wild rice habitats in the US are likely to be significantly reduced. Because Ojibway communities cannot,as a whole,move as the ecosystems that are their homes shift, climate change is likely to reduce the resourcesneeded to sustain their traditional culture and impact their economic productivity and the value of estab-lished treaty rights unless adjustments are made (e.g.,see Keller, 1989).

The wild rice that grows abundantly in shallow lake and marshy habitats of northern Wisconsin andMinnesota is likely to be adversely affected. Wild rice plays a critical role in the economic and ceremonial lifeof many tribes. The hand-harvested and processed seed is highly prized as a gourmet food and adds signifi-cant commercial value to the rural reservation economy. Federal treaties guarantee the right of theAnishinaabeg to gather wild rice in their aboriginal territories,which cover much of the states of Wisconsinand Minnesota. As the climate changes,deep or flooding waters in early spring could delay germination ofthe seed on lake or river bottoms,leading to crop failure. Lower water levels later in the summer could causethe wild rice stalks to break under the weight of the fruithead or make the rice beds inaccessible to har-vesters. Extended drought conditions could encourage greater natural competition from more shallow waterspecies. Climate change will disrupt the current balances,and,as was illustrated during the dry summer of1988,conflicts over water can pit federal river management policies against tribal treaty rights and statedemands for water.

Evidence of significant patterns of change over the past 10,000 years confirms that substantial ecosystemchanges can occur as a result of changes in climate. Presuming future changes occur to the same extent aspast changes,tribes that trace their ancestry to the wooded regions will slowly become overtaken by grass-lands,such that the entire nature of place for many Native peoples is likely to change.

5Personal communications from Lionel Bordeaux and Ronald Trosper.

to sustain the presence of particular types of usefulplants or animals for at least a while longer. Whereecosystems shift from Native land holdings to near-by non-Native lands,new areas may need to bedeveloped or acquired to allow access to traditionalfood sources. Increased involvement of Nativeexperts in resource management,particularly ofpublic lands,may improve the quality of the newenvironments as well as help to sustain traditionalplants. Where climatic and ecosystem shifts are sig-nificant,new approaches will be needed. In all ofthese situations,adapting to changing wildlife andland cover on tribal lands will be challengingbecause options for continued access by Native peo-ples to traditional ecosystem resources on neighbor-ing lands may be limited.

In planning and working to meet the changing con-ditions, experience indicates that both traditionalknowledge and contemporary scientific knowledgecan help to understand and improve the environ-ment. The relationships between Native peoplesand their environments provide significant insightsand context for the scientific findings about climatechange and its implications for human life. Buildingthe bridge will be essential, for many tribal commu-nities use such traditional understandings,devel-oped over many generations,to guide their uses oflands within their immediate political jurisdictions.Because land use decisions on reservations will haveinfluences on,and be influenced by, the health andservices provided by wider regional ecosystems,itwill be essential for Indian and non-Indian people to

human experiences. Mountain environments,edgesof ecosystems,and bird populations will be especial-ly vulnerable. For example,the “Sky Islands”in themountainous west and the prairie-forest interfacebetween the Great Plains and Great Lakes will beplaces where significant changes seem likely (seebox,“Shifting Ecosystem Boundaries and the Senseof Place”). Wildlife,which is central to the culturallife of many tribes,is likely to be significantly affect-ed over coming decades. For example,under theequilibrium climate conditions of the Canadian cli-mate scenario,models of bird distributions (Price,2000) project a gross loss of up to 27% of theneotropical migratory birds,32% of the short-dis-tance migratory birds,and 40% of the resident birdspecies in Arizona (J. Price,draft report prepared forthe Environmental Protection Agency). Becausesome extirpations (local extinctions) are likely to beoffset by immigrations,this study suggests that thenet changes (6% loss of neotropical migrants,15%loss of short-distance migrants and 30% gain in resi-dent species) are not likely to be as severe as thegross. Whether colonizing species can “replace”extirpated species in an ecological sense isunknown at this time,as are the overall rates ofchange. From the point-of-view of Native peoples,what may be as important is the degree to whichany of these changes will impact their cultures orreligions (see box,“Wildlife and Ceremonies”).

Adaptation OptionsIn some cases,improved or altered land manage-ment practices (e.g., fire management) may be able

370


Wildlife and Ceremonies

Many Native peoples use wildlife as integral parts of their cultural and religious ceremonies. Among the vari-ous Pueblo peoples (e.g.,Hopi,Zuni, Keres,Tewas,Tiwas,and Towas), religious ceremonies are the center oftheir cultural lives (Tyler 1991). Birds are seen as spiritual messengers and are completely integrated into thetraditions of these Native American communities. More than 200 species of birds have unique Native names,and more than 100 are essential to parts of the Pueblo culture. Birds mark the passing of the seasons and areconsidered to have valuable spiritual properties needed by members of these Pueblos. Among the Zuni,prayer sticks are used as offerings to the spirit realm. Each prayer stick,depending on its purpose, requires aparticular combination of feathers drawn from among 72 different species of birds (Tyler 1991). Prayer sticksserve many of the same spiritual purposes in the Zuni religion that rosary beads serve to the Catholic reli-gion.

Zuni also have separate names for the Western and Mountain Bluebirds. These species are only found on thePueblo in winter and are used as symbols of transitions for fall and spring. Among both the Hopi and theZuni, bluebirds are associated with puberty rituals surrounding the passage from girlhood into womanhood(Tyler 1991).

Because of such linkages,shifts in climate and consequent shifts in the timing or the distributions of wildlifespecies are likely to have profound impacts on the cultural and religious lives of these peoples.

work together towards understanding each others’perspectives and choices about climate change,itsimplications,and how best to adapt.

COPING AND ADAPTATIONSTRATEGIESDuring the assessment process,speakers from trib-al communities consistently attributed theendurance of Native peoples,through extremeconditions and brutal transitions,to spiritual andcultural values. Although public and ceremonialexpressions of these values differ considerablyfrom tribe to tribe,Native people identify manycommonalties across wide geographical,linguistic,and environmental distances. These values form aconnector between the past and future,bringingimportant lessons learned through tens of thou-sands of years on the continent into the frame-works for choices that will need to be made aboutthe futures of Native peoples and homelands.From the analyses to date,it is clear that respond-ing to substantial changes in climate,while popula-tions remain in fixed and permanent locations,isvery likely to require new technologies,skilled per-sonnel,and financial resources. These three neces-sities,however, are desperately scarce in many trib-al communities. Most tribal communities are limit-ed in ways of creating wealth and they continue torely heavily on transfer payments from the federalgovernment. Adjusting plans for economic andsocial development to account for climate changemay require fresh thinking in federal and tribalpolicies and budgets. For example,because landsare usually held in trust for Native peoples,it canbe very difficult to obtain a mortgage or other loansince the assets are not held personally. Severaloptions,however, are emerging from considerationof adaptation and coping options.

Enhance Education and Access toInformation and Technology

Becoming educated on issues concerning climatechange will be critical for Native peoples through-out the country. Both those who live in tribalcommunities and those who make their homeselsewhere need to develop the understanding andskills to deal with a changing climate6. This educa-tion needs to be both comprehensive and wide-spread (see Johnson,1999). It is especially impor-

tant to improve the quality of education in the sci-ences and technology in the K-12 schools and tribalcolleges that serve Native youth. It will be essentialto enlist elders and mentors within each Native com-munity (including within each culturally distinctregion) to assist in the integration of contemporaryinformation and traditional values,and of Indian stu-dents who choose to pursue university degrees andcareers in science with their tribal communities.

Promote Local Land-use andNatural Resource Planning

In 1976,with the passage of the Indian Self-Determination and Education Act,the federal govern-ment began to encourage tribal governments to takeresponsibility for developing and implementing plansfor use of tribal lands and natural resources in collab-oration with local agencies. Since that action,severaltribes have received international recognition for suc-cess in managing their local resources. TheMenominee Nation of Wisconsin regularly trains man-agers from all parts of the world because of the effec-tiveness of the tribe’s sustainable timber and forestrypractices. The Spokane Tribe has built an exemplarywater resources program. Cost-effective ways – usingexisting networks and organizations – need to bedeveloped to inform decision-makers in tribal com-munities,and to provide shared access to adequatetechnical resources. Technologies now exist that canassist tribes to make thoughtful and informed choic-es. Ways need to be found to provide informationthat will support the abilities of Native peoples andtheir leaders to make prudent choices based onappropriate knowledge and appropriate values,usingappropriate processes aimed at promoting andenhancing diversified and sustainable economies intribal communities.

Participate in Regional andNational Discussions and Decision-making

One result of the trusteeship system has been thetight concentration of the attentions of tribal govern-ments on their relationships with the Bureau ofIndian Affairs. Gradually, since the late 1960s, federalagencies have begun to recognize that the trustresponsibility to Native peoples extends through allparts of the federal government. Tribes,which onceviewed the creation of relationships with federalagencies other than the BIA risky at best and irrele-vant at worst,are working increasingly successfullyacross agency and departmental lines. Federal agen-cies are also learning how to provide appropriatekinds and levels of service to Native peoples.

371


6A variety of professional organizations provide assistance to tribes

working on natural resource issues:the Native American Fish andWildlife Society;Inter-tribal Timber Council;Indian Agriculture Council;National Tribal Environmental Council,and the American Indian HigherEducation Consortium,among others.

they seek to understand and prepare for climatechange. Having better estimates of the patterns,magnitudes,and rates of climatic changes to beexpected is essential. Accurate weather data fromtribal communities – particularly those in remoterural areas – have rarely been compiled. Little sys-tematic effort has been made by federal,tribal,orstate agencies to gather information about the micro-climatic conditions associated with the various reser-vations. One significant research priority, therefore,requires the training of members of tribal communi-ties to collect and interpret weather information,including providing for the acquisition,installation,and maintenance of appropriate instrumentation tosupport the collection and recording of these data.

A second urgent research need requires inventoriesof the uses and conditions of land and naturalresources on each reservation.Such an inventorycan integrate information from remote sensing andgeographic information systems as well as tribalinformation on water quantity and quality and first-hand personal observations and culturally basedknowledge. While Native oral histories clearly havemany insights to offer, data acquisition and sharingof information have become especially complicatedissues for Native peoples. Generations of scholarlyobjectification of Native peoples,and appropriationof knowledge,objects,and human remains have cre-ated severe problems of mistrust of scholarly inquiry,and strong resistance to the sharing of sacred orprivileged information. The legitimacy of Native con-cerns over these issues cannot be dismissed,nor canthe urgent needs for research to be conducted fromwithin Native communities. As these essential andprimary needs are satisfied,Native peoples have con-sistently demonstrated willingness to take part inbroader conversations,as teachers and students,col-leagues and leaders,with individual and culturallybased perceptions that can enrich discussions andstrengthen decisions for all parties involved.Equipping tribes with the capability to scientificallysample and analyze plant and animal populationswill enhance the opportunities for broadening tribalparticipation in such conversations.

Such inventories are needed to provide the basis forestablishing baselines of environmental conditionsand economic and cultural activities on each reserva-tion. This needs to be followed by assessments ofthe opportunities and vulnerabilities that changes inclimate might bring. These sector analyses willrequire regional projections of future changes for avariety of climate variables,so that the wide range ofpotential impacts (and opportunities) can be evaluat-ed by the various tribes. Especially important are

Although the trust relationships between the federalgovernment and Native peoples are complex,theyare not impenetrable. New relationships are essen-tial to address issues of climate change. While creat-ing relationships with agencies of state governmentshas frequently been viewed by tribes as threateningtheir sovereignty, serious discussions about climatechange – at the regional,state,and national levels –need to include informed stakeholders from everyrelevant jurisdiction. A pertinent model of interac-tion and collaboration that provides technical sup-port,advice,and assistance to tribal environmentalofficers has been developed between tribes in thenorthern Great Plains and the University of NorthDakota. Similarly, the Southwest Strategy, an initia-tive of all stakeholders in Arizona and New Mexico,serves as a useful framework for strengthening com-munication and collaboration with tribes and federalagencies. Their success in broadening participationand making knowledge available in useful wayscould provide helpful lessons for other states,tribes,and regions.

CRUCIAL UNKNOWNS ANDRESEARCH NEEDSBecause there are many hundreds of tribes,there aremany hundreds of situations facing Native peoples as

372


Figure 3: Changes in Pacific Ocean temperatures associated withthe Pacific Decadal Oscillation appear to be the cause of changesin the migration path of salmon returning to spawn (Mysak, 1986).Under normal conditions, the salmon return to spawn by passingthrough the Juan de Fuca Strait and traditional tribal harvestingareas. Under El Niño conditions, however, the salmon take a pathto the north of Vancouver Island, passing through Johnstone Straitand not passing through traditional harvesting areas (Groot andQuinn, 1987; Hsieh and Lee, 1989, McKinnel et al., 1999). As the cli-mate warms, the likelihood of the salmon taking one path versusthe other is likely to change, and with further warming, the salmonmay no longer return to their traditional streams.

The Johnstone Strait Diversion at Sockeye Salmon

predictions of changes in water availability, becausewater is the fundamental resource for agriculture,tourism,and other vital activities.

Native communities also need to understand howecosystems are likely to respond to climate changes,both large-scale and small. What will happen tomigrations of birds, waterfowl,and fish? What willhappen to forests and grasslands? What will happento the rich mix of flora and fauna on which tradi-tional cultures and subsistence economies depend?How will ecosystems function differently in thefuture? Will they provide greater or fewerresources,and in what mixtures? That large changesare possible is evident from the changes occurringin response to climate variations. For example, vari-ations in Pacific Ocean temperatures associatedwith the Pacific Decadal Oscillation have beenobserved to cause changes in the migration path ofsalmon returning to spawn,thereby affectingwhether fish pass through traditional harvestingareas (see Figure 3). Gaining a better understandingof such variations,and then of the changes that willbe brought on by climate change,will be essential.

The rates at which Native communities will be ableto respond to major changes vary widely fromgroup to group. Many Native peoples continue toview human actions and community decisions interms of generations. This perspective sometimesmeans that reaching conclusions may require moretime and discussion than is customary in market-driven societies. Other groups clearly thrive onchange and move aggressively to take advantage ofnew opportunities for education,economic develop-ment,and technological innovation. Some of thechallenges presented by changes in climate mayrequire relatively rapid social adaptation and swiftaction. Other challenges may require the capacityto endure discomfort and examine new situationsfrom a variety of perspectives until appropriateresponses can be formulated.

Education and the exchanges of information acrosscultural boundaries,activities that enhance the abili-ties of Native and non-Native peoples to cope withsystemic changes,have been occurring at the indi-vidual level for generations. The NativePeoples/Native Homelands assessment process hasstarted to advance these interactions.

Whatever changes befall Earth’s climate,there isnow fresh ground for hope that Native peoples andnon-Natives may be able to address the challengescollaboratively, as relatives. Mitakuye Owasin (weare all related)!

END NOTES1.The number of Native Americans depends on thedefinition that is used. As a result,the number ofthose counted as Native Americans can differ. Forexample,the US Bureau of the Census counts asAmerican Indian anyone who identifies himself orherself as such. As in asking about other ancestralconnections,census enumerators require no proof ofIndian identity. Thus,census data include individualswho may identify themselves culturally and socially asAmerican Indian,but who are not formally enrolled asmembers of a particular tribe. As a result,the censusproduces a comparatively high count of the numberof American Indian people in the United States (USBureau of the Census,1990). The Bureau of IndianAffairs,on the other hand,counts only individualswho are officially enrolled as members of federallyrecognized tribes. Each tribe has the right to estab-lish its own criteria for enrollment. Most tribesrequire that a certain percentage of the individual’sancestors must have been members of that tribe.Some tribes recognize only affiliation through thefather’s family. Still other tribes have residencyrequirements indicating that the individual must liveon the tribe’s reservation for a specified number ofyears. The BIA’s total (seehttp://www.doi.gov/bia/aitoday/q_and_a.html) thusyields a lower number of American Indians. As a fur-ther complication,some tribes are recognized by stategovernments,but not by the Bureau of Indian Affairs(e.g.,the Lumbee of North Carolina). Members ofthese tribes are,therefore, recognized as Indian bysome levels and agencies of government,but not byothers. Periodically, a tribe may succeed in complet-ing the BIA’s rigorous process for obtaining federalrecognition,thus increasing the number of Indianpeople recognized as such by the Department of theInterior. Further, descendants of the original inhabi-tants of the Hawaiian Islands have,using theDepartment of the Interior’s own criteria,made credi-ble claim for federal recognition as Native Americans(Bordewich,1996).

2.The presence of non-Native Americans on reserva-tion lands was largely prevented until the DawesSeveralty Act,commonly called the Allotment Act, waspassed by Congress in 1887. Prior to this law, reserva-tion lands were held corporately by an entire tribeand no particular individual held title to any particulartract of land. Furthermore,no outsiders, except gov-ernment officials and soldiers, were permitted to livewithin reservation borders. The Allotment Act,howev-er, mandated that each member of a tribe receive anindividual allotment of land. The allotments varied in

373


Brody, H., Maps and Dreams, Pantheon Books,NewYork,1982.

Brown, D., Bury My Heart at Wounded Knee:AnIndian History of the American West, Henry Holt,NewYork,487 pp.,1991.

Cornell,S., The Return of the Native:American IndianPolitical Resurgence, Oxford University Press,OxfordUnited Kingdom,1998.

Deloria,V.,Jr., Red Earth,White Lies:Native Americansand the Myth of Scientific Fact, Fulcrum Publishing,Golden Colorado,271 pp.,1997.

DOI (Department of the Interior),Lands under thejurisdiction of the Bureau of Indian Affairs,(http://www.doi.gov/bia/realty/area.html.),1996.

EIA (Energy Information Administration),Energy con-sumption and renewable energy development potentialon Indian lands,Department of Energy,Washington DC,March 2000.

Engelthaler, D. M.,et al.,Climatic and environmentalpatterns associated with hantavirus pulmonary syn-drome, Four Corners region,United States, EmergingInfectious Diseases, 5, 87-94,1999.

Gallagher,W., The Power of Place, Harper Perennial,New York,1993.

Getches, D. H,C. F.Wilkinson,and R.H.Williams,Jr.,Cases and Materials on Federal Indian Law, FourthEdition,West Group,St. Paul,Minnesota,1998.

Gibson,M.A.,and S.B.Schullinger,Answers from theIce Edge:The consequences of climate change on lifein the Bering and Chukchi seas,Arctic Network andGreenpeace USA pamphlet,(see http://www.glob-alchange.org/monitall/98apr6.htm),32 pp., June,1998.

Goodman,R., Lakota Star Knowledge, Sinte GleskaUniversity Press,Rosebud,South Dakota,1990.

Groot,C.,and T. P. Quinn,Homing migration of sockeyesalmon to the Fraser River, Fisheries Bulletin, 85, 455-469,1987.

Hardin,R., Collective Action, Johns Hopkins UniversityPress,Baltimore,Maryland,1982.

Hiss,T., The Experience of Place, Vintage Books,NewYork,1991.

size from 80 to 1,040 acres,depending on the par-ticular reservation. After each head of householdand family member had received an allotment,theremaining unassigned lands within the boundariesof each reservation could be opened to non-Indianhomesteaders.7 These settlers were granted cleartitle to the lands on which they settled if they ful-filled the normal conditions of homesteading. Landthat was conveyed in this way to homesteaders wassimply subtracted from the total lands that had beenoriginally reserved for the tribe. As a result,until thepassage of the Indian Reorganization Act in 1934,significant amounts of reservation lands passed outof Indian ownership even though they were withinthe original boundaries of reservation.

LITERATURE CITEDAkerlof, G.A., An Economic Theorist’s Book of Tales,University of Cambridge Press,Cambridge UnitedKingdom,1984.

Arias Rojo,H., J. Bredehoeft,R.Lacewell, J. Price, J.Stromberg,and G.Thomas, Sustaining and EnhancingRiparian Migratory Bird Habitat on the Upper SanPedro River, Expert Report prepared for theCommission for Environmental Cooperation,Montreal,Canada,1999.

Axelrod,R., The Evolution of Cooperation, BasicBooks,Inc.,New York,1984.

Axelrod,R., The Complexity of Cooperation, PrincetonUniversity Press,Princeton,New Jersey, 1997.

Basso,K., Wisdom Sits in Places:Landscape andLanguage Among the Western Apache, University ofNew Mexico Press,Albuquerque,New Mexico,1996.

BIA (Bureau of Indian Affairs) Website,http://www.doi.gov/bia/aitoday/q_and_a.html),2000a.

BIA (Bureau of Indian Affairs) Website,http://www.gdsc.bia.gov/images/bia_usa_chip.jpg ,2000b.

Bordewich, F. M., Killing the White Man’s Indian:Reinventing Native Americans at the End of theTwentieth Century, Anchor Books,New York,1996.

Brams,S. J.,and A. D.Taylor, Fair Division, CambridgeUniversity Press,Cambridge,United Kingdom,1996.

374


7Many tribes were able to avoid the allotment of tribal lands.The

reservations of the Menominee Nation,several Pueblo tribes and alarge portion of the Navajo Nation remain undivided and intact withintheir original borders.

Houser, S.,Mending the circle: Peer group lending formicroenterprise development in tribal communities,Rural Development Strategies, edited by David W.Sears and J. Norman Reid,pp.204-232,Nelson HallPublishers,Chicago,Illinois,1995.

Houser, S.,and A.White Hat,Bringing accountabilityhome:Changes in the management of public responsi-bility within tribal governments,commissioned by theCanadian Federal Department of Indian and NorthernDevelopment,Ottawa,1993.

Hsieh,W.W.,and W. G. Lee,A numerical model for thevariability of the northeast Pacific Ocean,CanadianSpecial Publication of Fisheries and Aquatic Sciences,108, 247-254,1989.

Huntington,H. P.,Impacts of changes in sea ice andother environmental parameters in the Arctic,in theProceedings of the Workshop on Impacts of Changesin Sea Ice and Other Environmental Parameters inthe Arctic, Girdwood,Alaska, February 15-17,2000,Marine Mammal Commission, Washingtonn,DC,inpress,2000.

IPCC (Intergovernmental Panel on Climate Change),Climate Change 1995: The Science of Climate Change,Contribution of Working Group I to the SecondAssessment Report of the Intergovernmental Panel onClimate Change, edited by J.T.,Houghton,L. G. MeiraFilho,B.A.Callander, N. Harris,A.Kattenberg,and K.Maskell,Cambridge University Press,Cambridge,UnitedKingdom,572 pp.,1996a.

IPCC (Intergovernmental Panel on Climate Change),Climate Change 1995:Impacts,Adaptations, andMitigation of Climate Change:Scientific-TechnicalAnalyses,.Contribution of Working Group II to theSecond Assessment Report of the IntergovernmentalPanel on Climate Change, edited by R.T.Watson,M.C.Zinyowera,and R.H.Moss,Cambridge University Press,Cambridge,United Kingdom,880 pp.,1996b.

Johnson,T.,World out of balance:In a prescient timeNative prophecy meets scientific prediction, NativeAMERICAS:Hemispheric Journal of Indigenous Issues,Volume XVI, Fall/Winter, 8-25,1999.

Keller, R.H.,America’s Native Sweet:Chippewa treatiesand the right to har vest maple sugar, American IndianQuarterly, Spring,117-135,1989.

LaDuke,W.,Traditional ecological knowledge and envi -ronmental futures, Colorado Journal of InternationalEnvironmental Law and Policy, 5, 127,1994.

Lawson,M.L., Dammed Indians, University ofOklahoma Press,Norman,Oklahoma,1982.

McKinnell,S.,H. J. Freeland,and S. D. Grouix,Assessingthe northern diversion of sockeye salmon returning tothe Fraser River, Fisheries Oceanography, 8, 104-114,1999.

Moreno, F.,In the Arctic,ice is life, Native AMERICAS:Hemispheric Journal of Indigenous Issues, XVI(Fall/Winter), 42-45,1999.

Mysak,L.A.,El Niño,interannual variability, and fish-eries in the northeast Pacific Ocean, Canadian Journalof Fisheries and Aquatic Sciences, 43, 464-497,1986.

NPNH (Native Peoples/Native Homelands AssessmentTeam), Report of the Native Peoples/NativeHomelands Climate Change Workshop, 28 October –1 November 1998, Albuquerque,New Mexico.NationalAeronautics and Space Administration,Washington DC,in press,2000.

Olson,M., The Logic of Collective Action, HarvardUniversity Press,Cambridge,Massachusetts,1965.

Pevar, S.L., The Rights of Indians and Tribes, secondedition,Southern Illinois University Press,Carbondale,Illinois,1992.

Price, J.,Modeling the potential impacts of climatechange on the summer distribution of Michigan’snongame birds, Michigan Birds and Natural History7, 3-13,2000.

Price, J.,Climate change,birds,and ecosystems – Whyshould we care?,in Managing for Healthy Ecosystems,Volume II:Issues and Methods, Section 3, ClimateChange and Ecosystem Health, edited by D. Rapport,C.Qualset,W. Lasley, and D. Ralston,in press,2001.

Riebsame,W., J. Robb,and D.Thoebald (editors), Atlasof the New West: Portrait of a Changing Region, W.W.Norton & Co.,New York,192 pp.,1997.

Ritzenthaler, R.E.,and P. Ritzenthaler, The WoodlandIndians of the Western Great Lakes, Waveland Press,Prospect Heights,Illinois,1991.

Rothrock, D.A.,Y.Yu,and G.A.Maykut,Thinning of theArctic sea-ice cover, Geophysical Research Letters,26(23),3469,1999.

Schelling,T. C., Micromotives and Macrobehavior, W.W. Norton & Co.,New York,1978.

375


376


ADDITIONAL BACKGROUNDREFERENCESAmbler, M., Breaking the Iron Bonds:IndianControl of Energy Development, University Press ofKansas,Lawrence,Kansas,1991.

Berkhofer, R. F.,Jr., The White Man’s Indian, A.A.Knopf, New York,1978.

Birkes, F., Sacred Ecology, Traditional EcologicalKnowledge and Resource Management, Taylor &Francis,London,England,1999.

Braroe,N.W., Indian & White, Stanford UniversityPress,Stanford,1978.

Burger,T. R. Village Journey:The Report of theAlaska Native Review Commission, Hill and Wang,New York,1985.

Cronon,W., Changes in the Land:Indians,Colonists, and the Ecology of New England, Hill &Wang,New York,1983.

Davis,T., Sustaining the Forest, the People, and theSpirit, State University of New York Press,Albany,New York,1999.

Deloria,V.,Jr.,and C. Lytle, The Nations Within:ThePast and Future of American Indian Sovereignty,Pantheon Books,New York,1984.

Diamond, J., Guns, Germs, and Steel: The Fates ofHuman Societies, W.W.Norton & Company, NewYork,1999.

Dippie,B.W., The Vanishing American:WhiteAttitudes and US Indian Policy, WesleyanUniversity Press,Middleton Connecticut,1982.

Gobert, G. G.,A.C.Abello,and J. Johnson,The individ-ual economic well-being of Native American menand women during the 1980s:A decade of movingbackward,in Changing Numbers, Changing Needs:American Indian Demography and Public Health,edited by G. Sandefur, R.R.Rindfuss,and B.Cohen,National Research Council,Washington,DC,pp.133-171,1996.

Josephy,A.M.Jr., Now That the Buffalo’s Gone,Alfred A.Knopf, New York,1982.

Schelling,T. C., Choice and Consequence, HarvardUniversity Press,Cambridge,Massachusetts,1984.

Schmaljohn,C.,and B.Hjelle,Hantaviruses:A global dis-ease problem, Emerging Infectious Diseases, 3, 95-104,1997.

Snipp,C.M., American Indians: The First of This Land,The Russell Sage Foundation,New York,1991.

Sorenson,L.,R.Goldberg,T. L.Root,and M. G.Anderson,Potential effects of global warming on waterfowl popu-lations breeding in the northern Great Plains, ClimaticChange, 40, 343-369,1998.

Spinden,H. J.,translator, Songs of the Tewa, originallypublished in 1933 by The Exposition of Indian TribalArts,Inc., republished by Sunstone Press,Santa Fe,NewMexico,1993.

Tiller,V. E.V., American Indian Reservations and TrustAreas, Economic Development Administration,UnitedStates Department of Commerce, Washington,DC,1996.

Trosper, R.L.,American Indian poverty on reservations,in Changing Numbers, Changing Needs:AmericanIndian Demography and Public Health, edited by G.Sandefur, R.R.Rindfuss,and B.Cohen,pp.172-195,National Research Council,Washington,DC,1996.

Trosper, R.L.,Traditional American Indian policy, inContemporary Native American Political Issues, edit-ed by T. R. Johnson,pp.139-63,Altamira Press,WalnutCreek,California,1999.

Tyler, H.A., Pueblo Birds and Myths, NorthlandPublishing,Flagstaff,Arizona,1991.

Tynan,C.T.,and D. P. Demaster, Observations and pre-dictions of Arctic climate change:Potential effects onmarine mammals, Arctic, 50,308-322,1997.

US Bureau of the Census,1990 Census of population,characteristics of American Indians by tribe and lan-guage,1990 CP-3-7,Washington,DC,1990.

Vennum,T., Wild Rice and the Ojibway People,Minnesota Historical Society Press,St. Paul,Minnesota,1988.

White,R.,and W. Cronon,Ecological change and Indian-White relations,in Handbook of North AmericanIndians, Volume IV, edited by W. Sturtevant,pp.417-429,Smithsonian Institution,Washington,DC,1988.

Krech,S.,III,(Ed.), Indians, Animals, and the FurTrade:A Critique of Keepers of the Game,University of Georgia Press,Athens,Georgia,1981.

Martin,C., Keepers of the Game:Indian AnimalRelationships in the Fur Trade, University ofCalifornia Press,Berkeley, California,1978.

Ortiz,A., The Tewa World:Space, Time, andBecoming in a Pueblo Society, University ofChicago Press,Chicago,Illinois,1969.

Oswalt,W. H., This Land Was Theirs, John Wiley &Sons,New York,1978.

Prucha, F. P., Americanizing the American Indians,Harvard University Press,Cambridge,Massachusetts,1973.

Prucha, F. P., The Indians in American Society: Fromthe Revolutionary War to the Present, University ofCalifornia Press,Berkeley, California,1985.

Scott,C.,Property, practice and aboriginal rightsamong Quebec Cree hunters, Hunters andGatherers 2:Property, Power, and Ideology, editedby T. Ingold,D. Riches,and J.Woodburn,pp.35-51,Berg Publishing,New York,1988.

Vogel,V., This Country Was Ours:A DocumentaryHistory of the American Indian, Harper & Row,New York,1972.

Weekes, P., The American Indian Experience,Forum Press,Arlington Heights,Illinois,1988.

377


ACKNOWLEDGMENTSMany of the materials for this chapter are based on

contributions from participants on and thoseworking with the

Native Peoples/Native Homelands—NationalWorkshop Team

Verna Teller, Isleta PuebloRobert Gough,Intertribal Council on Utility PolicySchuyler Houser, Lac Courte Oreilles Ojibwa

Community CollegeNancy Maynard, NASAFidel Moreno,Yaqui/HuicholLynn Mortensen,US Global Change Research

ProgramPatrick Spears,LakotaValerie Taliman,NavajoJanice Whitney, HETF Fiduciary

Native Peoples/Native Homelands—SouthwestAssessment Team

Stan Morain*,University of New MexicoRick Watson*,San Juan College and University of

New Mexico Diane Austin,University of ArizonaMark Bauer, Diné CollegeKarl Benedict,University of New MexicoJennifer Bondick,University of New MexicoAmy Budge,University of New MexicoLinda Colon,University of New MexicoLaura Gleasner, University of New MexicoJhon Goes In Center, Oglala Lakota NationTodd Hinckley, US Geological SurveyDoug Isely, Diné CollegeBryan Marozas,DOI Bureau of Indian AffairsLynn Mortensen,US Global Change Research

ProgramVerna Teller, Isleta PuebloCarmelita Topaha,NavajoRay Williamson,George Washington University

Additional ContributorsPatricia Anderson,University of AlaskaLynne Carter, National Assessment Coordination

OfficeSusan Marcus,US Geological SurveyJeff Price,American Bird ConservancyJames Rattling Leaf, Sinte Gleska UniversityGeorge Seielstad,University of North DakotaEileen Shea,East-West Center, HawaiiTony Socci,US Global Change Research ProgramLeigh Welling,University of North Dakota


379

CHAPTER 13

CLIMATE CHANGE AND AGRICULTURE INTHE UNITED STATESJohn Reilly1, Francesco Tubiello2, Bruce McCarl3 and Jerry Melillo4,5


Chapter Summary

Introduction


Climate Context

Previous Assessments - A Brief Overview

Key Issues

Crop Yield Changes

Economic Impacts

Resource and Environmental Effects

Potential Effects of Climate Variability on Agriculture


Appendix:Review of Previous Assessments

Literature Cited

Acknowledgments

1Massachussetts Institute of Technology,2Goddard Institute for Space Studies and Columbia University, 3Texas A&M

University, 4The Ecosystems Center, Marine Biological Laboratory, 5Coordinating author for the National AssessmentSynthesis Team

and Canadian scenarios,may be overestimated.

Under climate change simulated in the two climatescenarios,consumers benefited from lower priceswhile producers’profits declined. For the Canadianscenario,these opposite effects were nearly bal-anced, resulting in a small net effect on the nationaleconomy. The estimated $4-5 billion (in year 2000dollars unless indicated) reduction in producers’profits represents a 13-17% loss of income,whilethe savings of $3-6 billion to consumers representless than a 1% reduction in the consumers food andfiber expenditures. Under the Hadley scenario,pro-ducers’profits are reduced by up to $3 billion(10%) while consumers save $9-12 billion (in therange of 1%). The major dif ference between themodel outputs is that under the Hadley scenario,productivity increases were substantially greaterthan under the Canadian, resulting in lower foodprices to the consumers’benefit and the producers’detriment.

At the national level,the models used in thisAssessment found that irrigated agriculture’s needfor water declined approximately 5-10% for 2030and 30-40% for 2090 in the context of the two pri-mary climate scenarios,without adaptation due toincreased precipitation and shortened crop-growingperiods.

A case study of agriculture in the drainage basin ofthe Chesapeake Bay was undertaken to analyze theeffects of climate change on surface-water quality.In simulations for this Assessment,under the twoclimate scenarios for 2030,loading of excess nitro-gen into the Bay due to corn production increasedby 17-31% compared with the current situation.

Pests are currently a major problem in US agricul-ture. The Assessment investigated the relationshipbetween pesticide use and climate for crops thatrequire relatively large amounts of pesticides.Pesticide use is projected to increase for most cropsstudied and in most states under the climate scenar-ios considered. Increased need for pesticide appli-cation varied by crop – increases for corn weregenerally in the range of 10-20%; for potatoes,5-15%;and for soybeans and cotton,2-5%. The resultsfor wheat varied widely by state and climate sce-nario showing changes ranging from approximately

CHAPTER SUMMARY

It is likely that climate changes and atmosphericCO2 levels,as defined by the scenarios examined inthis Assessment,will not imperil crop production inthe US during the 21st century. The Assessmentfound that,at the national level,productivity ofmany major crops increased. Crops showing gener-ally positive results include cotton,corn for grainand silage,soybeans,sorghum,barley, sugar beets,and citrus fruits. Pastures also showed increasedproductivity. For other crops including wheat, rice,oats,hay, sugar cane,potatoes,and tomatoes,yieldsare projected to increase under some conditionsand decline under others.

Not all agricultural regions of the United States wereaffected to the same degree or in the same directionby the climates simulated in the scenarios. In gener-al the findings were that climate change favorednorthern areas. The Midwest (especially the north-ern half),West,and Pacific Northwest exhibitedlarge gains in yields for most crops in the 2030 and2090 timeframes for both of the two major climatescenarios used in this Assessment,Hadley andCanadian. Crop production changes in otherregions varied,some positive and some negative,depending on the climate scenario and time period.Yields reductions were quite large for some sites,particularly in the South and Plains States, for cli-mate scenarios with declines in precipitation andsubstantial warming in these regions.

Crop models such as those used in this Assessmenthave been used at local, regional,and global scalesto systematically assess impacts on yields and adap-tation strategies in agricultural systems,as climateand/or other factors change. The simulation resultsdepend on the general assumptions that soil nutri-ents are not limiting,and that pests,insects,diseases,and weeds,pose no threat to crop growth and yield.One important consequence of these assumptions isthat positive crop responses to elevated CO2, whichaccount for one-third to one-half of the yieldincreases simulated in the Assessment studies,should be regarded as upper limits to actualresponses in the field. One additional limitation thatapplies to this study is the models’inability to pre-dict the negative effects of excess water conditionson crop yields. Given the “wet”nature of the sce-narios employed,the positive responses projected inthis study for rainfed crops,under both the Hadley

380

Potential Consequences of Climate Variability

–15 to +15%. The increase in pesticide use resultsin slightly poorer overall economic performance,but this effect is quite small because pesticideexpenditures are in many cases a relatively smallshare of production costs.

The Assessment did not consider increased croplosses due to pests,implicitly assuming that all addi-tional losses were eliminated through increased pestcontrol measures. This could possibly result inunderestimates of losses due to pests associatedwith climate change. In addition,this Assessmentdid not consider the environmental consequencesof increased pesticide use.

Ultimately, the consequences of climate change forUS agriculture hinge on changes in climate variabili -ty and extreme events. Changes in the frequencyand intensity of droughts, flooding,and storm dam-age are likely to have significant consequences.Such events cause erosion, waterlogging,and leach-ing of animal wastes,pesticides, fertilizers,and otherchemicals into surface and groundwater.

One major source of weather variability is the ElNiño/Southern Oscillation (ENSO). ENSO effectsvary widely across the country. Better prediction ofthese events would allow farmers to plan ahead,altering their choices of which crops to plant andwhen to plant them. The value of improved fore-casts of ENSO events has been estimated at approxi-mately $500 million per year. As climate warms,ENSO is likely to be affected. Some models projectthat El Niño events and their impacts on US weatherare likely to be more intense. There is also a chancethat La Niña events and their impacts will bestronger. The potential impacts of a change in fre-quency and strength of ENSO conditions on agricul-ture were modeled. An increase in these ENSO con-ditions was found to cost US farmers on averageabout $320 million per year if forecasts of theseevents were available and farmers used them to planfor the growing season. The increase in cost wasestimated to be greater if accurate forecasts werenot available or not used.

381

Chapter 13 / Agriculture

382


INTRODUCTIONBoth weather and climate af fect virtually everyaspect of agriculture,from the production of cropsand livestock,to the transportation of agriculturalproducts to market. Agricultural crop productionis likely to be significantly af fected by the project-ed changes in climate and atmospheric CO2

(Rosenzweig and Hillel,1998). While elevated CO2

increases plant photosynthesis and thus cropyields (Kimball,1983),the projected changes intemperature and precipitation have the potentialto affect crop yields either positively or negatively.The negative effects are associated with some cli-mate changes that result in more rapid plant devel-opment,and modification of water and nutrientbudgets in the field (Long,1991).

The net effects of increased CO2 and climatechange on crop yields will ultimately depend onlocal conditions. For example,higher spring andsummer air temperatures might be beneficial tocrop production at northern temperate latitudesites,where the length of the growing seasonwould increase. However, higher temperaturesmight have negative effects during crop maturityin those regions where summer temperature andwater stress already limit crop production(Rosenzweig and Tubiello,1997).

The response of agricultural systems to future cli-mate change will additionally depend on manage-ment practices,such as the levels of water andnutrient applied. Water limitation tends toenhance the positive crop response to elevatedCO2, compared to well-watered conditions(Chaudhuri et al.,1990;Kimball et al.,1995). Theopposite is true for nitrogen limitation: well-fertil-ized crops respond more positively to CO2 thanless fertilized ones (Sionit et al.,1981;Mitchell etal.,1993).

This Assessment is intended to present our latestunderstanding of the potential impacts of climatechange on the agricultural sector. The Assessmentrelies on two sources of information:the relevantscientific literature,and new quantitative and quali-tative analyses done specifically as part of theAssessment.

Complete documentation of the work of theAgriculture Assessment Team is given in their SectorReport (Reilly et al.,2000;http://www.nacc.usgcrp.gov). This Foundation document,while a stand-alone statement,summarizes the report of theAgriculture Assessment Team.

In this document, we review the major activitiesundertaken in this Assessment. First, we present asummary of the key findings of the national andinternational assessments of climate change andagriculture that have been undertaken during thepast two decades. Second, we briefly report resultsof new simulation modeling,done for thisAssessment,that considers the consequences of twodifferent climate-change scenarios on crop yield andthe economics of agriculture in the US. Third,weset out the essential findings of new analyses onhow the impacts of climate change on agriculturemay, in turn,affect resources such as water andother aspects of the environment. And finally, wediscuss the highlights of a new analysis of climatevariability on agriculture.

The agriculture sector Assessment considered cropagriculture, grazing,livestock and environmentaleffects of agriculture. The focus here is primarily oncrop agriculture,which was studied most intenselyin the Assessment. Grain production is a major con-cern,with attention given to vegetables and fruitcrops.

The approach used to assess the effects of climatechange on crop agriculture involved an “end-to-end”analysis that linked climate-change scenarios for thefuture derived from general circulation models,withcrop models designed to consider the effects of cli -mate change and elevated atmospheric CO2 on cropyields. The outputs of the crop models were inputsto an economic model that was then used to ana-lyze the economic consequences of changed cropyields on farmers and consumers.

SOCIOECONOMIC CONTEXTThe US is a major supplier of food and fiber for theworld,accounting for more than 25% of the totalglobal trade in wheat,corn,soybeans,and cotton.

CLIMATE CHANGE AND AGRICULTURE IN THE UNITED STATES

383


scenario,increases in annual average temperature of9˚F (5˚C) by the year 2100 are common across thecentral US,with changes about half this large alongthe East and West coasts. Seasonal patterns indicatethat projected changes will be particularly large inwinter, especially at night. Large increases in tem-perature are projected over much of the South insummer. In the Hadley model scenario,the easternUS has temperature increases of 3-5˚F (2-3˚C) by2100,while the rest of the nation warms more,upto 7˚F (4˚C),depending on the region.

In both models,Alaska is projected to experiencemore intense warming than the lower 48 states,andin fact,this warming is already well underway. Incontrast,Hawaii,the other Pacific islands,and theCaribbean islands are likely to experience lesswarming than the continental US,because they areat lower latitudes and are surrounded by ocean,which warms more slowly than land.

Precipitation. At this time, climate scientists haveless confidence in climate model projections ofregional precipitation than of regional temperature.For the 21st century, the Canadian model projectsthe percentage increases in precipitation will belargest in the Southwest and California,while east ofthe Rocky Mountains,the southern half of thenation is projected to experience a decrease in pre-cipitation. The percentage decreases are projectedto be particularly large in eastern Colorado andwestern Kansas,and across an arc running formLouisiana to Virginia. Projected decreases in precipi-tation are most evident in the Great Plains duringthe summer and in the East during both winter andsummer. The increases in precipitation projected tooccur in the West,and the smaller increases in theNorthwest,are likely to occur mainly in winter.

In the Hadley model,the largest percentage increas-es in precipitation are projected to be in theSouthwest and Southern California,but the increas-es are smaller than those projected by the Canadianmodel. In the Hadley model,the entire US is pro-jected to have increases in precipitation,with theexception of small areas along the Gulf Coast and inthe Pacific Northwest. Precipitation is projected toincrease in the eastern half of the nation and insouthern California and parts of Nevada and Arizonain summer, and in every region during the winter,except the Gulf States and northern Washington andIdaho.

In both the Hadley and Canadian models,mostregions are projected to experience an increase inthe frequency of heavy precipitation events. This is

Cropland currently occupies about 400 million acres,or 17% of the total US land area. In addition, grass-lands and permanent grazing and pasturelands,occu-py almost 600 million acres,another 26% of US landarea. The value of agricultural commodities (foodand fiber) exceeds $165 billion at the farm level andover $500 billion,approaching 10% of GDP, after pro-cessing and marketing.

Economic viability and competitiveness are majorconcerns for producers trying to maintain profitabili-ty as real commodity prices have fallen by abouttwo-thirds over the last 50 years. Agricultural pro-ductivity has improved at over 1% per year since1950, resulting in a decline in both production costsand prices. This trend maintains intense pressure onindividual producers to continue to increase the pro-ductivity of their farms and to reduce costs of pro-duction. In this competitive economic environment,producers see anything that might increase costs orlimit their markets as a threat to their viability. Issuesof concern include regulatory actions that mightincrease costs,such as efforts to control the off-siteconsequences of soil erosion, agricultural chemicals,and livestock wastes; growing resistance to andrestrictions on the use of genetically-modified crops;new pests;and the development of pest resistance toexisting pest-control strategies. Future changes in cli-mate will interact with all of these factors.

CLIMATE CONTEXTThis Assessment of climate change is based on cli-mate scenarios derived from climate models devel-oped at the Canadian Centre for Climate Modellingand Analysis and the Hadley Centre in the UnitedKingdom. While the physical principles drivingthese models are similar, they differ in how they rep-resent the effects of some important processes.Therefore,these two primary models paint differentviews of the future. On average over the 21st centu-ry the Canadian model projects a greater tempera-ture increase than does the Hadley model,while theHadley model projects a much wetter climate thandoes the Canadian model. By using these two mod-els,a plausible range of future temperature condi-tions is captured,with one model being near thelower end and the other near the upper end of pro-jected temperature changes over the US. Both mod-els project much wetter conditions,compared topresent, over many agricultural areas in the US.

Temperature. Average warming in the US is project-ed to be somewhat greater than for the globe as awhole over the 21st century. In the Canadian model

competitors or potential competitors. The worstoutcome for the US would be severe climateeffects on production in most areas of the world,with particularly severe effects on US producers.Consumers would suffer from high food prices,producers would have little to sell,and agriculturalexports would dwindle. While an unlikely out-come based on newer climate scenarios,someearly scenarios that featured particularly severedrying in the mid-continental US with milder con-ditions in Russia,Canada,and the northern half ofEurope produced a moderate version of this sce-nario. The US and the world could gain most if cli-mate changes were generally beneficial to produc-tion worldwide,but particularly beneficial to USproducing areas. Consumers in the US and aroundthe world would benefit from falling prices andUS producers would also gain because the improv-ing climate would lower their production costseven more than prices fell,thus increasing theirexport competitiveness. In fact,most scenarioscome close to the middle,with relatively modesteffects on world prices. The larger gainers interms of production are the more northern areasof Canada,Russia,and Northern Europe. Tropicalareas are more likely to suffer production losses.The US as a whole straddles a set of climate zonesthat include gainers (the northern areas) and los-ers (southern areas).

Effects on producers and consumers often are inopposite directions and this is often responsiblefor the small net effect on the economy. Thisresult is a near certainty without trade,andreflects the fact that demand is not very respon-sive to price so that anything that restricts supply(e.g.,acreage reduction programs,environmentalconstraints, climate change) leads to price increas-es that more than make up for the reduced out-put. Once trade is factored in,this result dependson what happens to production abroad as dis-cussed above.

US agriculture is a competitive, adaptive, andresponsive industry and will likely adapt to cli -mate change;all assessments reviewed have fac -tored adaptation into their analyses. The finaleffect on producers and the economy after consid-eration of adaptation may be either negative orpositive. The evidence for adaptation is drawnfrom analogous situations such as the response ofproduction to changes in commodity and inputprices, regional shifts in production as economicconditions change,and the adoption of new tech-nologies and farming practices.

especially notable in the Hadley model,but theCanadian model shows the same characteristic.

PREVIOUS ASSESSMENTS –A BRIEF OVERVIEWSeveral conclusions are shared among assessmentsconducted over the past quarter century. Herethese are brief ly reviewed and a more completesynopsis of some of the important previous assess-ments is given in the Appendix.

Over the next 100 years and probably beyond,human-induced climate change as currentlymodeled will not seriously imperil overall foodand fiber production in the US, nor will it great -ly increase the aggregate cost of agricultural pro -duction. Most assessments have looked at multi -ple climate scenarios. About one-half of the sce-narios in any given assessment have shown smalllosses for the US (increased cost of production)and about one-half have shown gains for the US(decreased cost of production). However, noassessment has adequately included the potentialimpacts of extreme events,such as flooding,drought,and prolonged heat waves,and thepotential effects of increased ranges of pests,diseases,and insects. The result of includingthese factors could require a reevaluation ofthis finding.

There are likely to be strong regional productioneffects within the US, with some areas sufferingsignificant loss of comparative advantage (if notabsolute advantage) relative to other regions ofthe country. With very competitive economicmarkets,it matters little if a particular region gainsor loses absolutely in terms of yield,but ratherhow it fares relative to other regions. The south-ern region of the US is persistently found to loseboth relative to other regions and absolutely. Thelikely effects of climate change on other regionswithin the US are less certain. While warming canlengthen the growing seasons in the northern halfof the country, the full effect depends on precipi-tation,notoriously poorly projected by climatemodels.

Global market effects and trade dominate interms of net economic effect on the US economy.Just as climate’s effects on regional comparativeadvantage are important,the relevant concerns arethe overall effects on global production and pricesand how US producers fare relative to their global

384


385


KEY ISSUESHere, we briefly report results of new simulationmodeling done for this Assessment that considersthe consequences of two different climate changescenarios on crop yield and the economics of agri-culture in the US. In addition, we set out the essen-tial findings of new analyses on how the impacts ofclimate change on agriculture may, in turn,affectresources such as water and other aspects of theenvironment. And finally, we discuss the highlightsof new analyses of climate variability on agriculture.

Four key issues were identified:• Crop Yield Changes• Changes in Economic Impacts• Resource and Environmental Effects

Changing water demands for irrigationSurface water qualityIncreasing pesticide use

• Climate Variability

1. Crop Yield Changes

Approach

The agriculture-sector team investigated the effectsof climate change on US crop production,usingfuture climate scenarios generated by two climatemodels,the Hadley and Canadian models,as inputinto a family of dynamic crop-growth models. TheDSSAT family of models was used extensively in thisstudy to simulate wheat,corn,potato,soybean,sorghum, rice,and tomato (Tsuji et al.,1994). TheCENTURY model was used to simulate grasslandand hay production (Parton et al.,1994). Finally, themodel of Ben Mechlia and Carrol (1989) was usedto simulate citrus production.The models were runto simulate yields at 45 sites across the US. Thesesites were chosen using USDA national and state-level statistics to be in areas of major production.

All models employed have been used extensively toassess crop yields across the US under current con-ditions as well as under climate change (e.g.,Rosenzweig et al.,1995; Parton et al.,1994,Tubielloet al.,1999). Apart from CENTURY, which runs witha monthly time-step,all other models use dailyinputs of solar radiation,minimum and maximumtemperature,and precipitation to calculate plantphenological development from planting to harvest,photosynthesis and growth,and carbon allocation tograin or fruit. All models use a soil component tocalculate water and nitrogen movement,and arethus able to assess the effects of different manage-

ment practices on crop growth. The simulationsperformed for this study considered:1) rainfed pro-duction;and 2) optimal irrigation,defined as re-fill-ing of the soil water profile whenever water levelsfall below 50% of capacity at 30 cm depth. Fertilizerapplications were assumed to be optimal at all sites.Atmospheric concentrations of CO2 assumed in thecore analysis were as follows:350 parts per millionby volume (ppmv) for the base,445 ppmv for theyear 2030,and 660 ppmv for 2090. The crop mod-els assumed that crops such as wheat, rice,barley,oats,potatoes,and most vegetable crops,tend torespond favorably to increased CO2, with a doublingof CO2 leading to yield increases in the range of 15-20%. Other crops including corn,sorghum,sugarcane,and many tropical grasses, were assumed to beless responsive to CO2, with a doubling of the gasleading to yield increases of about 5%.

In addition to current practices at each site,simula-tions were done that included dif ferent adaptationtechniques. These consisted largely of testing theeffects of early planting,a realistic scenario at manynorthern sites under climate change;and of testingthe performance of cultivars better adapted towarmer climates,using currently available geneticstock. In general,early planting was considered forspring crops,to avoid heat and drought stress in thelate summer months,while taking advantage ofwarmer spring conditions. New, better-adapted cul-tivars were tested for winter crops,such as wheat,to increase the time to maturity (shortened underclimate change scenarios) and to increase yieldpotential.

Two other groups in the US developed additionalanalyses,independent from the core study describedabove. Specifically, researchers at the PacificNorthwest National Laboratories (PNNL) developednational-level analyses for corn,winter wheat,alfalfa,and soybean,using climate projections from theHadley model (Izarraulde et al.,1999). Anothergroup,co-located at Indiana University and PurdueUniversity, focused on corn,soybeans,and wheat,developing a regional analysis for the Midwest,including the states of Indiana,Illinois,Ohio,Wisconsin,and Michigan,using Hadley model pro-jections (Southworth et al.,2000).

In the PNNL study, the baseline climate data wereobtained from national records for the period1961–1990. The scenario runs were constructed fortwo future periods (2025–2034 and 2090–2099).The Erosion Productivity Impact Calculator (EPIC)was used to simulate the behavior of 204 farms withconsiderations of soil-climate-management combina-

386


tions under the baseline climate,the two futureperiods,and their combinations with two levels ofatmospheric CO2 concentrations (365 and 560ppmv).

In an independent study by Indiana University andPurdue University, a baseline climate was definedusing the period 1961-1990. Several future scenar-ios were analyzed for the decade of 2050,withatmospheric CO2 concentration set at 555 ppmv.Crop yields were simulated with the DSSAT modelat 10 representative farms in the Corn Belt andLake States. Adaptations studied included change ofplanting dates,as well as the use of cultivars withdifferent maturity groups. These results were notincluded in the economic modeling but provideanother source of information.

Although specific differences in time horizons,CO2

concentrations,and simulation methodologies com-plicate the comparison of these additional analysesto the work discussed herein,model findings wereoverall in general agreement with results of thecore study.

Results

Here we present the results from the models forseveral major crops. The DSSAT analyses for wheat,corn,alfalfa,and soybean, were integrated withresults from two additional independent studies.The national average changes in yields for drylandand irrigated crops with and without adaptation aregiven in Figures 1a-d. The national averages werecalculated by summing regional estimates for thecoterminous United States (Figure 2) that are speci-fied in Agriculture Sector Model (ASM). The region-al estimates were derived by using crop-model out-puts for sites in the region and harvested acreage ineach ASM region based on data from the 1992National Resource Inventory.

Yield Changes for Major Crops

A summary of the changes in simulated crop yieldsunder the Canadian and Hadley scenarios relative topresent yields is given below.

Winter wheat. Even with adaptation, rainfed pro-duction was reduced by an average of 9% in the2030 time period under the Canadian scenario.Adaptation techniques helped to counterbalanceyield losses in the Northern Plains,but not in theSouthern Plains where losses were more severe anddue to reductions in precipitation. Yields increasedan average of 23% under the Hadley scenario forthe 2030 period. Average dryland yields increasedunder both climate scenarios for the 2090 period,up to 59% in the case of the Hadley scenario.Irrigated wheat production increased under bothclimate scenarios by up to 16% on average by theend of the 21 st century when adaptation strategieswere used.

Spring wheat. Dryland production of spring wheatyields increased under both scenarios,either withor without adaptation. Adaptation techniques,including early planting and new cultivars,helpedto improve yields under both scenarios,up to 59%for the Hadley scenario in 2090. Irrigated yieldswere reduced slightly under the Canadian scenarioand increased slightly under the Hadley scenario.

Corn. Dryland corn production increased at mostsites due to increases in precipitation under bothclimate scenarios. Average yields were up bybetween 15 to 40% by the end of the 21st centuryin much of the Corn Belt region. Larger yield gainswere simulated in the Northern Great Plains and inthe Northern Lakes Region,where higher tempera-

Figure 2. Agriculture Sector Model (ASM) Regions with USDARegions Overlaid. (ASM regions follow state boundaries exceptwhere further disaggregated). The economic analysis in theAssessment is summarized for the 10 USDA regions outlined in themap. Source: Changing Climate and Changing Agriculture: Reportof the Agricultural Sector Assessment Team, 2000. See Color PlateAppendix

Dominant Land Uses, 1992

387


Figure 1a - Dryland Yields Without Adaptation

Figure 1a-d. Relative changes (% change relative to present) in crop yield for two time periods, 2030s and 2090s, under the Canadian andHadley Scenarios. 0 = no change. Under the two climate scenarios, most crops showed substantial yield increase, even without adaptation,under dryland conditions. Irrigated yields increased less or decreased. (Source: Changing Climate and Changing Agriculture: Report of theAgricultural Sector Assessment Team, 2000) See Color Plate Appendix

Figure 1b - Dryland Yields With Adaptation

Figure 1c - Irrigated Yields Without Adaptation Figure 1d - Irrigated Yields With Adaptation

tures were also beneficial to production. Irrigatedcorn production was not greatly affected at mostsites.

Potato. Potato production decreased across manysites analyzed. While major production areas in thenorthern US experienced either small increases orsmall decreases,at some other sites potato yieldsdecreased up to 50% from current levels. At thesesites,there was little room for cultivar adaptation,because the projected higher fall and winter tem-peratures negatively affected tuber formation.Adaptation of planting dates mitigated only some ofthe projected losses.

Citrus. Production largely benefited from the highertemperatures projected under all scenarios.Simulated fruit yield increased in the range of 60 to100%,while irrigation water use decreased. Croplosses due to freezing diminished by 65% in 2030,and by 80% in 2090.

Soybean. Soybean production increased at mostsites analyzed,in the range 20 to 40% for sites ofcurrent major production. Larger gains were simu-lated at northern sites where cold temperatures cur-rently limit crop growth. The Southeast sites consid -ered in this study experienced large reductionsunder the Canadian scenario. Losses were reducedby adaptation techniques involving the use of culti-vars with dif ferent maturity classes. (For regionaldetails see the Southeast chapter).

Sorghum. Sorghum production,especially withadaptation, generally increased under rainfed condi-tions,due to the increased precipitation projectedunder the two scenarios considered. Higher tem-peratures at northern sites further increased rainfedgrain yields. By contrast,irrigated production with-out adaptation was reduced almost everywhere,because of negative effects of higher temperatureson crop development and yield.

Rice. Rice production increased slightly under theHadley scenario,with the increases in the range 1-10%. Under the Canadian scenario, rice productionwas 10-20% lower than current levels at sites inCalifornia and in the Delta region.

Tomato. Without adaptation,irrigated tomato pro-duction decreased at most of the simulated sites dueto increased temperatures. Noted exceptions werethe northern locations where production is current-ly limited by low temperatures and by short grow-ing seasons. Reductions were in the 10 to 15%range under the Canadian scenario. Under the

388


Hadley scenario, reductions in tomato yields werein the 5% range. Adaptation strategies resulted inincreased yields of tomatoes under the two climatescenarios.

Even without adaptation,the weighted averageyield impact for many crops grown under drylandconditions across the entire US was positive underboth the Canadian and Hadley scenarios. In manycases,yields under the 2030 climate conditionsimproved compared with the control yields undercurrent climate,and improved further under the2090 climate conditions. These generally positiveyield results were observed for cotton,corn forgrain and silage,soybean,sorghum,barley, sugarbeet,and citrus fruit. The yield results were mixedfor other crops (wheat, rice,oats,hay, sugar cane,and potatoes) showing yield increases under someconditions and declines other conditions.

Changes in irrigated yields,particularly for thegrain crops, were more often negative or less posi-tive than dryland yields. This reflected the fact thatunder these climate scenarios precipitation increas -es were substantial. The precipitation increasesprovided no yield benefit to irrigated cropsbecause they face no water stress under currentconditions since all the water needed is providedthrough irrigation. Higher temperatures speddevelopment of crops and reduced the grain fillingperiod,thereby reducing yields. For dryland crops,the positive effect of more moisture and CO2 fertil-ization counterbalanced the negative effect of high-er temperatures.

Water demand by ir rigated crops dropped substan-tially for most crops. The faster development ofcrops due to higher temperatures reduced thegrowing period and thereby reduced waterdemand,more than offsetting increased evapotran-spiration due to higher temperatures while thecrops were growing. To a large extent,the reducedwater use thus reflected the reduced yields on irri-gated crops. Increased precipitation also reducedthe need for irrigation water.

Adaptations examined in the crop modeling studiescontributed small additional gains in yields of dry-land crops,particularly for those with large yieldincreases due to climate change. Adaptationoptions examined,including shifts in planting timesand choice of cultivars adapted to new climaticconditions. For the most part,however, these adap-tations had little additional benefit where yieldsincreased from climate change. This suggests thatadaptation may be able to partly offset changes in

comparative advantage across the US that mayresult under these scenarios. Other strategies foradaptation,such as whether or not to switch cropsor to irrigate, were options available in the eco-nomic model. The decisions to undertake thesestrategies are driven by economic considerations;that is,whether they are profitable under marketconditions simulated in the scenario. Adaptationsfor several crops were not considered because theoptions available,such as changing planting date,were not applicable to many perennial and treefruit crops. Adaptation studies were conducted foronly a subset of sites considered in the study andthese results were extrapolated to other sites.

Adaptation contributed greater yield gains for ir ri-gated crops. Shifts in planting dates were able toreduce some of the heat-related yield losses. Withhigher yields than in the no-adaptation cases, waterdemand declines were not as substantial. Again,this reflected the fact that the adaptations consid-ered extended the growing (and grain-filling peri-od) and this extension meant longer periods overwhich irrigation water was required.

The factors responsible for the positive results ofthis Assessment varied,but can generally be tracedto aspects of the climate scenarios. First,increasedprecipitation in these transient climate scenarios isan important factor contributing to the more posi-tive effects for dryland crops and explains the dif-ference between dryland and irrigated cropresults. The benefits of increased precipitation out-weighed the negative effects of higher tempera-tures for dryland crops,whereas increased precipi-tation had little yield benefits for irrigated cropsbecause water stress is not a concern for cropsalready irrigated. In fact,where the climate scenar-ios projected both higher temperatures anddecreases in precipitation,such as for the CentralPlains regions of Kansas and Oklahoma, rainfedcereal production,notably winter wheat, was nega-tively affected.

As noted for the Central Plains,not all agriculturalregions of the United States are affected to thesame degree or direction by the climates in thescenarios. In general, climate change as projectedin the two climate scenarios favored northernareas. The Midwest (especially northern areas),West,and Pacific Northwest exhibited large gainsin yields for most crops with both climate scenar-ios in the 2030 and 2090 time frames. Yieldchanges in other regions were mixed,dependingon the climate scenario and time period. Forexample in the Southeast,simulated yields for most

crops increased under the Hadley scenario in boththe 2030 and 2090 time frames. Yield estimates var-ied widely among crops under the Canadian sce-nario. Citrus yields increased slightly by 2030,anddramatically by 2090. Dryland soybean yieldsdecreased in the range of 10-30 % in about 2030,and by up to 80 % in about 2090. And rice yieldsdecreased on the order of 5 to 10 % for both timeperiods

The potential for within-region differences washighlighted in the Indiana University/PurdueUniversity study of the Midwest. In this study,decreases were found in corn yields across thesouthern portion of the region’s southern states —Indiana,Illinois,and Ohio. In addition,decreases,oronly small increases in yields, were found for soy-bean and wheat across these same southern loca-tions. In the region’s northern states, Wisconsin andMichigan,there were simulated increases in yieldfor all the crops studied,with soybean showing themost dramatic increases. In addition,a variabilityanalysis indicated that a doubling of current climatevariability in association with climate change wouldproduce the most detrimental climate conditionsfor crop growth across this region (Southworth etal.,1999).

Crop models such as those used in this Assessmenthave been used at local, regional,and global scalesto systematically assess impacts on yields and adap-tation strategies in agricultural systems,as climateand/or other factors change. The simulationresults depend on the general assumption that soilnutrients are not limiting,and that pests,insects,diseases,and weeds pose no threat to crop growthand yield (Patterson et al.,1999;Rosenzweig andHillel,1998;Rosenzweig et al.,2000;Strzepek etal.,1999;Tubiello et al.,1999;Walker et al.,1996).One important consequence of these assumptionsis that positive crop responses to elevated CO2,responsible for one-third to one-half of the yieldincreases simulated in the Assessment studies,should be regarded as upper limits to actualresponses in the field. One additional limitationthat applies to this study is the models’inability topredict the negative effects of excess water condi-tions on crop yields. Given the “wet”nature of thescenarios employed,the positive responses pro-jected in this study for rainfed crops,under boththe Hadley and Canadian scenarios,may be overes-timated.

389


390


2. Economic Impacts

Approach

The crop results were combined with impacts onwater supply, livestock,pesticide use,and shifts ininternational production to estimate impacts on theUS economy. This allowed the estimation of region-al production shifts and resource use in response tochanging relative comparative advantage amongcrops and producing regions. These changes wereestimated using a US national agricultural sectormodel (ASM) (Adams et al.,1990,1997) that islinked to a global trade model.

The ASM is based on the work of Baumes (1978)which was later modified and expanded by Burtonand Martin(1987);Adams et al.(1986);Chang et al.(1992) and Lambert et al.(1995). Conceptually,ASM is a price endogenous,mathematical program-ming model of the type described in McCarl andSpreen (1980). Constant elasticity curves are usedto represent domestic consumption and exportdemands as well as input and import supplies.Elasticities were assembled from a number ofsources including USDA through the USMP model-ing team (House,1987) and prior model versions.ASM is designed to simulate the effects of variouschanges in agricultural resource usage or resourcesavailable on agricultural prices,quantities pro-duced,consumers’and producers’ welfare, exports,imports and food processing. In calculating these

effects,the model considers production,processing,domestic consumption,imports, exports and inputprocurement.

The model distinguishes between primary and sec-ondary commodities,with primary commoditiesbeing those directly produced by the farms and sec-ondary commodities being those involving process-ing. Within ASM,the US is disaggregated into 63geographical production subregions.Each subregionpossesses different endowments of land,labor andwater as well as crop yields. Agricultural produc-tion is described by a set of regional budgets forcrops and livestock.Marketing and other costs areadded to the budgets following the proceduredescribed in Fajardo et al.(1981) such that the mar-ginal cost of each budget equals marginal revenue.ASM also contains a set of national processing budg-ets which uses crop and livestock commodities asinputs (USDA,1982). There are also import supplyfunctions from the rest of the world for a numberof commodities. The demand sector of the modelconsists of the intermediate use of all the primaryand secondary commodities,domestic consumptionuse and exports.

There are 33 primary crop and livestock commodi-ties in the model. The primary commodities depictthe majority of agricultural production,land use andeconomic value. The model incorporates process-ing of the primary commodities. There are 37 sec-ondary commodities that are processed in themodel. These commodities are chosen based ontheir linkages to agriculture. Some primary com-modities are inputs to the processing activitiesyielding these secondary commodities and certainsecondary products (feeds and by-products) are inturn inputs to production of primary commodities.Three land types (crop land,pasture land,and landfor grazing on an animal unit month basis) are spec-ified for each region. Land is available according toa regional price elastic supply schedule with arental rate as reported in USDA farm real estate sta-tistics. The labor input includes family and hiredlabor. A region-specific reservation wage and maxi-mum amount of family labor available reflect thesupply of family labor. The supply of hired laborconsists of a minimum inducement wage rate and asubsequent price elastic supply. Water comes fromsurface and pumped ground water sources. Surfacewater is available at a constant price,but pumpedwater is supplied according to a price elastic supplyschedule.

US agricultural sector models typically only dealwith aggregate exports and imports facing the total

Figure 3a and b. The economic index is change in welfareexpressed as the sum of producer and consumer surplus in billionsof dollars. There were net economic benefits for the US under mostof the scenarios examined in the Assessment. Foreign consumersalso gained from lower commodity prices on international markets.Source: Changing Climate and Changing Agriculture: Report of theAgricultural Sector Assessment Team, 2000. See Color PlateAppendix

Economic Impacts of Climage Change on USAgriculture

US without regional trading detail. The ASMincludes foreign regions,and shipment among for-eign regions modeled as 6 spatial equilibrium mod-els for the major traded commodities (Takayama andJudge,1971). To portray US regional ef fects,US mar-kets are grouped into the ten regional definitionsused by the USDA. We also added variables for ship-ment among US regions,and shipment between USregions and foreign regions. The commodities sub-ject to explicit treatment via the spatial equilibriumworld trade model components are hard red springwheat (HRSW),hard red winter wheat (HRWW),softred winter wheat (SOFT),durum wheat (DURW)),corn,soybeans and sorghum.These commodities areselected based on their importance as US exports.The rest of the world is aggregated into 28 coun-tries/regions. Transportation cost,trade quantity,price and elasticity were obtained from Fellin andFuller (1998),USDA (1987) statistical sources andthe USDA SWOPSIM model (Roningen,1986).

In the base results, climatic effects on crops andlivestock in the rest of the world were assumed tobe neutral,that is,no climate change effects on agri-culture in the rest of the world were assumed. Totest how sensitive the results were to this assump-tion three scenarios of climate impacts on agricul-ture in the rest of the world were used. These weredeveloped from previous work reported in Reilly etal.(1993;1994),and based on a global analysis usinga Hadley Centre climate scenario and a global agri-cultural model developed by Darwin et al.(1995).These climate scenarios were not completely consis-tent with the new scenarios used for the US,butprovide a good test of the sensitivity of US econom-ic results to impacts in the rest of the world.

Results

The net economic ef fect on the US economy wasgenerally positive, reflecting the generally positiveyield effects (Figures 3a,b). The exceptions weresimulations under the Canadian climate scenario in2030,particularly in the absence of adaptation.

Foreign consumers gained in all the scenarios as aresult of lower prices for US export commodities.The total effects (net effect on US producers andconsumers plus foreign gains) ranged from a $0.5billion loss to a $12.5 billion gain.

This Assessment found that producers and con-sumers were affected in opposite ways by climatechange (Figures 4a,b). Producers’incomes generallyfell due to lower prices. Producer reductionsranged from about $0.1 up to 5 billion. The largest

391


reductions were under the Canadian scenario.Under the Hadley scenario,producers suffered fromlower prices,but enjoyed considerable increases inexports such that the net effect was for only verysmall reductions. Economic gains accrued to con-sumers through lower prices in all scenarios. Gainsto consumers ranged from $2.5 to 13 billion.

Different scenarios of the effect of climate changeon agriculture abroad did not change the net impacton the US very much,but redistributed changesbetween producers and consumers. The directiondepended on the direction of effect on worldprices. Lower prices increased producer losses andadded to consumer benefits. Higher prices reducedproducer losses and consumer benefits.

Modeled projections of livestock production andprices were mixed. Increased temperatures directlyreduced productivity, but improvements in pastureand grazing and reductions in feed prices due tolower crop prices counter these losses. (For addi-tional comments on livestock,see the Great Plainschapter).

Producer versus Consumer Impacts of Climate Change

Canadian Model

Hadley Model

Consumer SurplusProducer Surplus

Consumer SurplusProducer Surplus

15

15

10

5

0

-5

-10

2030 2030w/adapt. 2030

w/pests 20902090

w/adapt.2090

w/pests

2030 2030w/adapt. 2030

w/pests 20902090

w/adapt.2090

w/pests

Figure 4a and b. In the model simulations consumers generally bene-fited from climate change while producers experienced lower incomedue to lower prices for commodities resulting from increased yieldsand supply. Source: Changing Climate and Changing Agriculture:Report of the Agricultural Sector Assessment Team, 2000. See ColorPlate Appendix

weighted index across livestock and crop produc-tion. Many of the impacts,because they occur atthe regional level,are dealt with in more detail inthe regional chapters.

3. Resource and EnvironmentalEffects

In terms of improving the coverage of potentialimpacts of climate change on agriculture,this studyhas made advances over previous assessments.Some of the advances were in the area of resourceand environmental effects (Figure 6). Details of thestudies underlying this summary are given in theAgriculture Sector Report (Reilly, et al.2000;http://www/nacc.usgcrp.gov).

Demand for land. Agriculture’s pressure on landresources generally decreased under both climatescenarios across the 21 st century. Area in croplanddecreased 5 to 10 % while area in pasture decreased10 to 15 %.

Grazing pressure. Animal unit months (AUMs) ofgrazing on western lands decreased on the order of10% under the Canadian climate scenario andincreased 5 to 10% under the Hadley climate sce-nario.

Demand for water, a national perspective. At thenational level,the models used in this Assessmentfound that irrigated agriculture’s need for waterdeclined approximately 5-10% for 2030 and 30-40%for 2090 climate conditions as represented in thetwo scenarios. At least two factors were responsiblefor this reduction in water demand for irrigation.One was increased precipitation in some agricultur-al areas. The other was that faster development ofcrops due to higher temperatures resulted in areduced growing period and thereby reduced waterdemand. In the crop modeling analyses done forthe Assessment,shortening of the growing periodreduced plant water-use enough to more than com-pensate the increased water losses from plants andsoils due to higher temperatures.

Demand for water, a regional perspective. Thecompetition for water between agriculture andother uses was explored through a case study of theEdwards Aquifer that serves the San Antonio regionof Texas. Agriculture uses of water compete withurban and industrial uses and tight economic man-agement is necessary to avoid unsustainable use ofthe resource. Aquifer discharge is through pumpingand artesian spring discharge.

Aggregate regional production changes (Figures 5a,b) were positive for all regions in both the 2030and 2090 time frames under the Hadley scenario.Adaptation measures had a small additional posi-tive effect. In contrast, aggregate productionchanges differed among regions under theCanadian scenario in both the 2030s and 2090s. Itwas positive for most northern regions,mixed forthe Northern Plains,and negative for Appalachia,the Southeast,the Delta states,and the SouthernPlains. Adaptation measures helped somewhat forthe southern regions,but the aggregate productionwas lower in these regions under both the 2030and 2090 climates considered. Aggregate produc-tion is represented in Figures 5a,b as a price-

392


Regional Production Changes Relative to Current Production

Figure 5a and b. In the model simulations productionincreased in northern regions as a result of longer growingseasons, and in western regions due to increased precipita-tion. Higher temperatures and increased drought condi-tions contributed to production declines or smaller increas-es in southern and plains regions. Source: ChangingClimate and Changing Agriculture: Report of theAgricultural Sector Assessment Team, 2000. See ColorPlate Appendix

The study found that the Canadian and Hadley sce-narios of climatic change caused a slightly negativewelfare result in the San Antonio region as awhole,but had a strong impact on the agriculturalsector.

The regional welfare loss,most of which wasincurred by agricultural producers, was estimatedto be between $2.2 and 6.8 million per year if cur-rent pumping limits are maintained.

A major reason for the current pumping limits is topreserve the artesian spring flows that are criticalto the habitat of local endangered species. Tomaintain spring flows at the currently specifiedlevel to protect endangered species,pumpingwould need to be reduced in the future with cli-mate change. The study calculated that under thetwo climate scenarios,pumping would need to bereduced by 10 to 20% below the limit currently setand this would cost an additional $0.5 to 2 millionper year. Welfare in the non-agricultural sector wasonly marginally reduced by the climatic changesimulated by the two climate scenarios. The valueof water permits rose dramatically.

The agricultural use of water is discussed in severalof the regional chapters including the Great Plains.

Surface water quality. As part of the Assessment, astudy was undertaken of the linkages between cli-mate change and nitrogen loading of ChesapeakeBay. The Chesapeake Bay is one of nation’s mostvaluable natural resources,but has been severelydegraded in recent decades. Soil erosion and nutri-ent runoff from crop and livestock productionhave played a major role in the decline of the Bay.Based on simulations done for this Assessment,under the Canadian and Hadley scenarios for the2030 period,nitrogen loading from corn produc-tion increased by 17 to 31% compared with condi-tions under current climate. Potential effects of cli-mate change on water quality in the ChesapeakeBay must be considered very uncertain becausecurrent climate models may not fully represent theeffects of extreme weather events such as floodsor heavy downpours,which can wash largeamounts of fertilizers,pesticides,and animalmanure into surface waters.

Surface water quality is also discussed in theSoutheast and West chapters.

Pesticide expenditures. The Assessment investigat-ed the relationship between pesticide use and cli-mate for crops that require relatively large amounts

of pesticide. Pesticide expenditures increased underthe climate scenarios considered for most cropsstudied and in most regions. Increases for cornwere generally in the range of 10 to 20%, for pota-toes of 5 to 15% and for soybeans and cotton of 2 to5%. The results for wheat varied widely by state andclimate scenario showing changes ranging fromapproximately –15 to +15%. These projections werebased on cross-section statistical evidence on therelationship between pesticide expenditures andtemperature and precipitation.

The increase in pesticide expenditures couldincrease environmental problems associated withpesticide use,but much depends on how pest con-trol evolves over the next several decades. Pestsdevelop resistance to control methods, requiring acontinual evolution in the chemicals and controlmethods used.

The increase in pesticide expenditures resulted inslightly poorer overall economic performance butthis effect was quite small because pesticide expen-ditures are a relatively small share of productioncosts. The approach used in the Assessment did notconsider increased crop losses due to pests,implicit-ly assuming that all additional losses were eliminat-ed through increased pest control measures. Thismay underestimate pest losses.

393


Changes in Resource Use

Figure 6. In the simulations resource use generally declinedas less crop and grazing land was needed. Use of waterand irrigated crop land declined the most because the twoclimate scenarios used favored dryland over irrigated crops(cc-Canadian, hc=Hadley). Source: Changing Climate andChanging Agriculture: Report of the Agricultural SectorAssessment Team, 2000. See Color Plate Appendix

Canadian (cc) and Hadley (hc) Climates, Without Adaptation

2030cc 2090cc 2030hc 2090hc

20

10

0

-10

-20

-30

-40

-50

-60

Irrig. Land DrylandCropland Pasture

AUMsWater

Labor

these ENSO conditions was found to cost US farm-ers about $320 million on average per year if accu-rate forecasts of these events were available andfarmers used them as they planned for the growingseason. The increase in cost was estimated to begreater if accurate forecasts were not available ornot used.

ADAPTATION STRATEGIESAdaptations such as changing planting dates andchoosing longer season varieties are likely to offsetlosses or further increase yields. Adaptive measuresare likely to be particularly critical for the Southeastbecause of the large reductions in yields projectedfor some crops under the more severe climate sce-narios examined. Breeding for response to CO2 willlikely be necessary to achieve the strong fertilizationeffect assumed in the crop studies. This is an unex-ploited opportunity and the prospects for selectingfor CO2 response are good. However, attempts tobreed for a single characteristic are often not suc-cessful,unless other traits and interactions are con-sidered. Breeding for tolerance to climatic stress hasalready been heavily exploited and varieties that dobest under ideal conditions usually also outperformother varieties under stress conditions. Breedingspecific varieties for specific conditions of climatestress is therefore less likely to encounter success.

Some adaptations to climate change and its impactscan have negative secondary effects. For example,an examination of use of water from the Edward’saquifer region around San Antonio,Texas foundincreased pressure on groundwater resources thatwould threaten endangered species dependent onspring flows supported by the aquifer. Anotherexample relates to agricultural chemical use. Anincrease in the use of pesticides and herbicides isone adaptation to increased insects, weeds,and dis-eases that could be associated with warming.Runoff of these chemicals into prairie wetlands,groundwater, and rivers and lakes could threatendrinking water supplies,coastal waters, recreationareas,and waterfowl habitat.

The wide uncertainties in climate scenarios, regionalvariation in climate effects,and interactions of envi-ronment,economics,and farm policy suggest thatthere are no simple and widely applicable adapta-tion prescriptions. Farmers will need to adaptbroadly to changing conditions in agriculture,ofwhich changing climate is only one factor. Some ofthe possible adaptations more directly related to cli-mate include:

4. Potential Effects of ClimateVariability on Agriculture

Ultimately, the consequences of climate change forUS agriculture hinge on changes in climate variabili-ty and extreme events. Agricultural systems are vul-nerable to climate extremes,with effects varyingfrom place to place because of differences in soils,production systems,and other factors. Changes inprecipitation type (rain,snow, and hail),timing,fre-quency, and intensity, along with changes in wind(windstorms,hurricanes,and tornadoes),could havesignificant consequences. Heavy precipitationevents cause erosion, waterlogging,and leaching ofanimal wastes,pesticides, fertilizers,and other chem-icals into surface water and groundwater. While allof the risks associated with these impacts are notknown,the system is known to be sensitive tochanges in extremes. The costs of adjusting to suchchanges will likely increase if the rate of climatechange is high,although early signals from a rapidlychanging climate would reduce uncertainty andencourage early adaptation.

One major source of weather variability is the ElNiño/Southern Oscillation (ENSO) phenomenon.ENSO phases are triggered by the movement ofwarm surface water eastward across the PacificOcean toward the coast of South America and itsretreat back across the Pacific,in an oscillating fash-ion with a varying periodicity. Better prediction ofthese events would allow farmers to plan ahead,planting different crops and at dif ferent times. Thevalue of improved forecasts of ENSO events hasbeen estimated at approximately $500 million peryear.

ENSO’s effects can vary from one event to the next.Predictions of the details of ENSO-driven weatherare not perfect. There are also widely varyingeffects of ENSO across the country. The tempera-ture and precipitation effects are not the same in allregions;in some regions the ENSO signal is relative-ly strong while in others it is weak,and the changesin weather have different implications for agricul-ture in different regions because climate-related pro-ductivity constraints differ among regions underneutral climate conditions.

As climate warms,ENSO is likely to be affected.Some models project that El Niño and La Niñaevents and their impacts on US weather willbecome more intense with climate change. Thepotential impacts of projected changes in frequencyand strength of ENSO conditions on agriculturewere modeled in this Assessment. An increase in394


Sowing dates and other seasonal changesPlant two crops instead of one or a spring and fallcrop with a short fallow period to avoid excessiveheat and drought in mid-summer. For already warmgrowing areas,winter cropping could possiblybecome more productive than summer cropping.

New crop varietiesThe genetic base is very broad for many crops,andbiotechnology offers new potential for introducingsalt tolerance,pest resistance,and general improve-ments in crop yield and quality.

Water supply, irrigation, and drainage systemsTechnologies and management methods exist toincrease irrigation efficiency and reduce problemsof soil degradation,but in many areas,the economicincentives to reduce wasteful practices do not exist.Increased precipitation and more intense precipita-tion will likely mean that some areas will need toincrease their use of drainage systems to avoidflooding and waterlogging of soils.

CRUCIAL UNKNOWNS ANDRESEARCH NEEDSFurther research is needed in several areas. Broadly,these include:1) integrated modeling of the agricul-tural system;2) research to improve resiliency of theagricultural system to change;and 3) several areas ofclimate-agriculture interactions that have not beenextensively investigated.

Integrated modeling of the agricultural system

• The main methodology for conducting agricul-tural impacts models has been to use detailedcrop models run at a selected set of sites and touse the output of these as input to an economicmodel. This approach has provided great insightsbut future assessments will need to integratethese models to consider interactions and feed-backs, multiple environmental stresses (tropos-pheric ozone,acid deposition,and nitrogen dep-osition),transient climate scenarios,global analy-sis,and to allow study of uncertainty wheremany climate scenarios are used. The presentapproach of teams of crop modelers runningmodels at specific sites severely limits the num-ber of sites and scenarios that can be feasibly beconsidered.

• The boundaries of the agricultural system in anintegrated model need to be expanded so thatmore of the complex interactions can be repre-sented. Changes in soils, multiple demands for

water, more detailed analysis and modeling ofpests,and the environmental consequences ofagriculture and changes in climate are areas thatneed to be incorporated into one integratedmodeling framework. Agricultural systems arehighly interactive with economic managementchoices that are affected by climate change.Separate models and separate analyses cannotcapture these interactions.

Resiliency and adaptation

• Specific research on adaptation of agriculture toclimate change at the time scale of decades tocenturies should not be the centerpiece of anagricultural research strategy. Decision-making inagriculture mostly involves time horizons of oneto five years,and long-term climate predictionsare not very helpful for this purpose. Instead,effort should be directed toward understandingsuccessful farming strategies and where adapta-tions to many changes are needed to managerisk.

• There is also great need for research to improveshort-term and intermediate term (i.e.,seasonal)weather predictions and on how to make betteruse of these predictions.

New areas of research

• Experimentation and modeling of the interac-tions of multiple environmental changes oncrops (changing temperature,CO2 levels,ozone,soil conditions,moisture,etc.) are needed.Experimental evidence is needed under realisticfield conditions such as Free Air Carbon DioxideEnrichment (FACE) experiments for CO2 enrich-ment.

• Much more work on agricultural pests and theirresponse to climate change is needed.

• Economic analyses need to better study thedynamics of adjustment to changing conditions.

• Climate-agriculture-environment interactions areperhaps one of the more important vulnerabili-ties,but the existing research is extremely limit-ed. Soil, water quality, and air quality should beincluded in a comprehensive study of interac-tions.

• Agricultural modeling must be more closely inte-grated with climate modeling and modelers todevelop better techniques for assessing theimpacts of climate variability. This requires sig-nificant advances in climate predictions to betterrepresent changes in variability as well as assess-ment of and improvements in the performanceof crop models under extreme conditions.

395


The study assigned subjective probabilities to thescenarios,attempted to project ranges of crop yieldimprovement in the absence of climate change,andcompared climate-induced changes to normal vari-ability in crop yields and uncertainty in future pro-jections of yield. A summary point highlighted thedifficulty in ultimately detecting any crop yieldchanges due to climate given the year-to-year vari-ability and the difficulty in disentangling climateeffects from the effects of new varieties and otherchanging technology that would inevitably be intro-duced over the 25-year period.

1988-1989:US EPAUS EPA (Smith and Tirpak,1989) evaluated theimpacts of climate change on US agriculture as partof an overall assessment of climate impacts on theUS. The agricultural results were published inAdams,et al.1990. The study evaluated warmingand changes in precipitation based on doubled CO2

equilibrium climate scenarios from 3 widely knownGeneral Circulation Models (GCMs),with increasedaverage global surface warming of 4.0 to 5.2ºC (7.2to 9.4ºF). In many ways the most comprehensiveassessment to date,it included studies of possiblechanges in pests and interactions with irrigationwater supply in a study of California. The mainstudy on crop yields used site studies and a set ofcrop models to estimate crop yield impacts. Thesewere simulated through an economic model.Economic results were based on imposition of cli-mate change on the agricultural economy in 1985.Grain crops were studied in most detail,with a sim-pler approach for simulating impacts on othercrops. Impacts on other parts of the world werenot considered. The basic conclusions summarizedin the Smith and Tirpak report were:

• Yields could be reduced,although the combinedeffects of climate and CO2 would depend on theseverity of climate change.

• Productivity may shift northward.• The national supply of agricultural commodities

may be sufficient to meet domestic needs,butexports may be reduced.

• Farmers would likely change many of their prac-tices.

• Ranges of agricultural pests may extend north-ward.

• Shifts in agriculture may harm the environmentin some areas.

1988-1990: Intergovernmental Panel on ClimateChange (IPCC), first assessment reportIn the first assessment report of theIntergovernmental Panel on Climate Change (IPCC),(Parry 1990a and in greater detail, Parry, 1990b)

APPENDIX – REVIEW OFPREVIOUS ASSESSMENTSHere we provide a short summary of the majorassessments of the potential consequences of cli-mate change for agriculture. The summary does notinclude a detailed scientific literature review thatforms the foundation for past assessments as well asthis one. A new set of reviews on climate changeimpacts on crops,livestock,pests,and soils,as wellas discussion of global and regional impacts,hasbeen published in a special edition of the journal,Climatic Change, Climate Change:Impacts OnAgriculture (Reilly, 1999). The 5 articles included inthe edition contain over 500 citations,providing adetailed guide to the scientific literature relating cli-mate change and agriculture.

1976-1983:National Defense University A National Defense University (Johnson,1983) proj-ect produced a series of reports with the 1983report providing the final report on agriculture,inte-grating yield and economic effects. It focused onthe world grain economy in the year 2000,consider-ing both warming and cooling of up to approxi-mately 1ºC (1.8ºF) for large warming or cooling and0.5ºC (0.9ºF) for moderate changes for the US,withassociated precipitation changes on the order of +/-0-2%. These estimates varied somewhat by region.The base year for comparison purposes was 1975.It relied on an expert opinion survey for yieldeffects,using the results to create a model of crop-yield response to temperature and precipitation formajor world grain regions. There was no explicitaccount of potential interactions of pests, changes insoils,or of livestock or crops such as fruits and veg-etables. No direct effects of CO2 on plant growthwere considered as the study remained agnosticabout the source of the climate change (e.g.,whether due to natural variability or human-induced). Economic effects were assessed using amodel of world grain markets.

Crop yields in the US were estimated to fall by 1.6to 2.3% due to moderate and large warming and toincrease by very small amounts (less than 3%) withlarge cooling and even smaller amounts with moder-ate cooling. Warming was estimated to increasecrop yields in the (then) USSR,China,Canada,andEastern Europe,with cooling decreasing crop pro-duction in these areas. Most other regions wereestimated to gain from cooling and suffer yield loss-es from warming. The net effect was a very smallchange in world production and on world prices.

396


North American agriculture was brief ly addressed.The assessment was based mainly on a literaturereview and, for regional effects,on expert judge-ment. North American/US results mainly summa-rized the earlier EPA study. Some of the main contri-butions of the report were to identify the multiplepathways of effects on agriculture including effectsof elevated CO2, shifts of climatic extremes, reducedsoil water availability, changes in precipitation pat-terns such as the monsoons,and sea-level rise. Italso identified various consequences for farmingincluding changes in trade,area farmed,irrigation,fertilizer use,control of pests and diseases,soildrainage and control of erosion, farming infrastruc-ture,and interaction with farm policies. The overallconclusion of the report was that:“on balance,theevidence suggests that in the face of estimatedchanges of climate, food production at the globallevel could be maintained at essentially the samelevel as would have occurred without climatechange;however, the cost of achieving this wasunclear.”

As an offshoot of this effort,the Economic ResearchService of USDA (Kane et al.,1991 and subsequent-ly, as Kane et al.,1992,and Tobey et al.,1992) pub-lished an assessment of impacts on world produc-tion and trade,including specifically the US. Thestudy was based on sensitivity to broad generaliza-tions about the global pattern of climate change asportrayed in doubled CO2 equilibrium climate sce-narios,illustrating the importance of trade effects. A“moderate impacts scenario”brought together avariety of crop model results based on doubled CO2

equilibrium climate scenarios and the expert judge-ments for other regions that were the basis for theIPCC. In this scenario,the world impacts were verysmall (a gain of $1.5 billion 1986 US$). The US wasa very small net gainer ($0.2 billion) with China,Russia,Australia,and Argentina also benefiting whileother regions lost. On average,commodity priceswere estimated to fall by 4%,although corn and soy-bean prices rose by 9-10%.

1990-1992: US DOE, Missouri, Iowa, Nebraska,Kansas (MINK) Study In the Missouri,Iowa,Nebraska,Kansas (MINK)Study (Rosenberg,1993;Easterling et al.,1993) thedust bowl of the 1930s was used as an analogue cli-mate for global change for the four-state region.Unique aspects of the study included considerationof water, agriculture, forestry, and energy impacts,and projection of regional economy and crop vari-ety development to the year 2030. Crop responsewas modeled using crop models, river flow usinghistorical records,and economic impacts using an

input-output model for the region. Despite the factthat the region was “highly dependent”on agricul-ture compared with many areas of the country, thesimulated impacts had relatively small ef fects on theregional economy. Climate change losses in termsof yields were on the order of 10 to 15%. With CO2

fertilization effects,most of the losses were eliminat-ed. Climate impacts were simulated for currentcrops as well as “enhanced”varieties with improvedharvest index,photosynthetic efficiency, pest man-agement,leaf area,and harvest efficiency. Theseenhanced varieties were intended to represent pos-sible productivity changes from 1990 to 2030 andincreased yield on the order of 70%. The percentagelosses due to climate change did not differ substan-tially between the “enhanced”and current varieties.

1992:Council on Agricultural Science andTechnology (CAST) ReportThe Council on Agricultural Science and Technology(CAST, 1992) report,commissioned by the USDepartment of Agriculture did not attempt any spe-cific quantitative assessments of climate changeimpacts, focusing instead on approaches for prepar-ing US agriculture for climate change. It focused ona portfolio approach, recognizing that predictionwith certainty was not possible.

1992-1993:Office of Technology Assessment study The Office of Technology Assessment (OTA,1993)study, similar to the CAST study for agriculture,focused on steps that could prepare the US for cli-mate change rather than estimates of the impacts.The study’s overall conclusions for agriculture werethat the long-term productivity and competitivenessof US agriculture were at risk and that market-drivenresponses may alter the regional distribution andintensity of farming. It found institutional impedi-ments to adaptation, recognized that uncertaintymade it hard for farmers to respond,and saw poten-tial environmental restrictions and water shortagesas limits to adaptation. It also noted that decliningFederal interest in agricultural research and educa-tion could impede adaptation. The study recom-mended removal of institutional impediments toadaptation (in commodity programs,disaster assis-tance,and water-marketing restrictions),improve-ment of knowledge and responsiveness of farmersto speed adaptation,and support for both generalagricultural research and research targeted towardspecific constraints and risks.

1992-1994: US EPA Global AssessmentA global assessment (Rosenzweig and Parry, 1994;Rosenzweig et al.,1995) of climate impacts onworld food prospects expanded the method used inthe US EPA study for the United States to the entire

397


the studies listed above. The overall conclusionsincluded a summary of the direct and indirecteffects of climate and increased ambient CO2,regional and global production effects,and vulnera-bility and adaptation. With regard to direct and indi-rect effects:

The results of a large number of experiments toresolve the effect of elevated CO2 concentrations oncrops have confirmed a beneficial effect. The meanvalue yield response of C3 crops (most crops exceptmaize,sugar cane,millet,and sorghum) to doubledCO2 is +30% although measured responses rangefrom -10 to +80%.

• Changes in soils,e.g.,loss of soil organic matter,leaching of soil nutrients,salinization,and ero-sion,are a likely consequence of climate changefor some soils in some climatic zones. Croppingpractices including crop rotation,conservationtillage,and improved nutrient management are,technically, quite effective in combating orreversing deleterious effects.

• Livestock production will be affected by changesin grain prices, changes in the prevalence anddistribution of livestock pests,and changes ingrazing and pasture productivity, as well as thedirect effects of weather. Heat stress in particu-lar may lead to significant detrimental effects onproduction and reproduction of some livestockspecies.

• The risk of losses due to weeds,insects,and dis-eases is likely to increase.

With regard to regional and global productioneffects:

• Crop yields and productivity changes will varyconsiderably across regions. Thus,the pattern ofagricultural production is likely to change in anumber of regions,with some areas experiencingsignificantly lower crop yields and other areasexperiencing higher yields.

• Global agricultural production can be maintainedrelative to base production under climate changeas expressed by GCMs under doubled CO2 equi-librium climate scenarios.

• Based on global agricultural studies using dou-bled CO2 equilibrium GCM scenarios,lower lati-tude and lower income countries are likely to bemore negatively affected.

With regard to vulnerability and adaptation:

• Vulnerability to climate change depends on phys-ical and biological response,but also on socioe-

world. It was based on the same suite of crop andclimate models and applied these to many sitesaround the world. It used a global model of worldagriculture and the world economy that simulatesthe evolving economy through to 2060,assumed tobe the period when the doubled CO2-equilibriumclimates applied. The global temperature changeswere +4.0 to +5.2ºC (7.2 to 9.4ºF). Scenarios withthe CO2 fertilization effect and modest adaptationshowed global cereal production losses of 0 to5.2%. In these scenarios,developed countriesshowed cereal production increases of 3.8 to 14.2%,while the developing countries showed losses of 9.2to 12.5%. The study concluded that in the develop-ing world there was a significant increase in thenumber of people at risk because of climate change.The study also considered dif ferent assumptionsabout yield increases due to technology improve-ment,trade policy, and economic growth. These dif-ferent assumptions and scenarios had equal or moreimportant consequences for the number of peopleat risk of hunger.

Other researchers simulated yield effects estimatedin this study through economic models, focusing onimplications for the US (Adams et al.,1995) andworld trade (Reilly et al.1993;1994). Adams et al.(1995) estimated economic welfare gains for the USof approximately $4 and 11 billion (1990 US$) fortwo climate scenarios and a loss of $16 billion forthe other scenario,under conditions reflectingincreased export demands and a CO2 fertilizationeffect (550 ppmv CO2). The study found thatincreased exports from the US,in response to highcommodity prices resulting from decreased globalagricultural production,led to benefits to US pro-ducers of approximately the same magnitude as thewelfare losses to US consumers from high prices.Reilly et al.(1993;1994) found welfare gains to theUS of $0.3 billion (1990 US$) under one GCM sce-nario and up to $0.6 to $0.8 billion losses in theother scenarios when simulating productionchanges for all regions of the world through a trademodel. They also found widely varying effects onproducers and consumers,with producers’effectsranging from a $5 billion loss to a $16 billion gain,echoing the general findings of Adams et al.(1995).In particular, Reilly et al.1994 showed that in manycases,more severe yield effects produced economicgain to producers because world prices rose.

1994-1995:IPCC, Second Assessment ReportThe Second Assessment Report of the IPCC includ -ed an assessment of the impacts of climate changeon agriculture (Reilly et al.,1995). As an assessmentbased on existing literature,it summarized most of

398


conomic characteristics. Low-income popula-tions depending on isolated agricultural systems,particularly dryland systems in semi-arid and aridregions,are especially vulnerable to hunger andsevere hardship. Many of these at-risk popula-tions are found in Sub-Saharan Africa,South andSoutheast Asia,some Pacific island countries,andtropical Latin America.

• Historically, farming systems have responded to agrowing population and have adapted to chang-ing economic conditions,technology, andresource availability. It is uncertain whether therate of change of climate and required adaptationwould add significantly to the disruption likelydue to future changes in economic conditions,population,technology, and resource availability.

• While adaptation to climate change is likely;theextent depends on the affordability of adaptivemeasures,access to technology, and biophysicalconstraints such as water resource availability,soil characteristics, genetic diversity for cropbreeding,and topography. Many current agricul-tural and resource policies are likely to discour-age effective adaptation and are a source of cur-rent land degradation and resource misuse.

• National studies have shown incremental addi-tional costs of agricultural production under cli-mate change that could create a serious burdenfor some developing countries.

• Material in the 1995 IPCC Working Group IIreport was reorganized by region with someupdated material in a subsequent special report.Included among the chapters was a report onNorth America (Shriner and Street,1998).

1995-1996. The Economic Research Service of theUSDAThe Economic Research Service of the USDA(Schimmelpfennig et al.,1996) provided a reviewand comparison of studies that it had conductedand/or funded,contrasting them with previous stud-ies. The assessment used the same doubled CO2

equilibrium scenarios of many previous studies(global average surface temperature increases of 2.5to 5.2ºC or 4.5 to 9.4ºF). Two of the main newanalyses reviewed in the study used cross-sectionevidence to evaluate climate impacts on production.One approach was a direct statistical estimate of theimpacts on land values for the US (Mendelsohn etal.,1994),while the other (Darwin et al.,1995) usedevidence on crop production and growing seasonlength in a model of world agriculture and theworld economy. Both imposed climate change on

the agricultural sector as it existed in the base year ofthe studies (e.g.,1990). A major result of theapproaches based on cross-section evidence was thatimpacts of climate were far less negative for the USand world than had previously been estimated withcrop modeling studies. While these studies showedsimilar economic effects as previous studies,theyincluded no direct ef fects of CO2 on crops,which inprevious studies had been a major factor behind rela-tively small effects. Hence,if the direct effects of CO2

on crop yields would have been included,the resultwould have been significant benefits. The more posi-tive results were attributed to adaptations implicit incross-section evidence that had been incompletelyfactored in to previous analyses. The report also con-tained a crop modeling study (Kaiser et al.,1993)with a complete farm-level economic model thatmore completely simulated adaptation responses. Italso showed more adaptation than previous studies.A summary of this review was subsequently pub-lished as Lewandrowski and Schimmelpfennig(1999).

1998-1999: Pew Center Assessment As part of a series on various aspects of climatechange aimed at increasing public understanding,thePew Center on Global Climate Change completed areport on agriculture (Adams et al.,1999). The reportseries is based on reviews and synthesis of the exist-ing literature. The major conclusions were:

• Crops and livestock are sensitive to climatechanges in both positive and negative ways.

• The emerging consensus from modeling studies isthat the net effects on US agriculture associatedwith doubling of CO2 may be small;however,regional changes may be significant (i.e.,there willbe some regions that gain and some that lose).Beyond a doubling of CO2, the negative effectswould be more pronounced,both in the US andglobally.

• Consideration of adaptation and human responseis critical to an accurate and credible assessment.

• Better climate change forecasts are a key toimproved assessments.

• Agriculture is a sector that can adapt,but changesin the incidence and severity of pests,diseases,soil erosion,tropospheric ozone, variability, andextreme events have not been factored in to mostof the existing assessments.

399


Chaudhuri, U. N.,M .B.Kirkam,and E.T. Kanemasu,Root growth of winter wheat under elevated carbondioxide and drought, Crop Science, 30, 853-857,1990.

Darwin,R.,M.Tsigas, J. Lewandrowski,and A.Raneses,World agriculture and climate change: Economic adap-tations, Agricultural Economic Report No. 703, USDepartment of Agriculture,Natural Resources andEnvironmental Division,Economic Research Service,Washington,DC,1995.

Easterling,W.E.,III, P. R.Crosson, N. J. Rosenberg,M.McKenney, L.A.Katz,and K.Lemon,Agriculturalimpacts of and responses to climate change in theMissouri-Iowa-Nebraska-Kansas (MINK) region, ClimateChange, 24, 23-61,1993.

Fajardo, D.,B.A.McCarl,and R.L.Thompson,A Multicommodity Analysis of Trade Policy Effect:The Case ofNicaraguan Agriculture. American Journal ofAgricultural Economics 63:23-31,1981.

Fellin,L.,and S.Fuller, Effects of Privatization Mexico’sRailroad System on US-Mexico Overland Grain/OilseedTrade, Transportation Research Forum Vol 37,N1:46-64,1998.

House,R.M.,USMP Regional Agricultural Model,UnitedStates Department of Agriculture,Economic ResearchService,National Economics Division,Draft,1987.

Izarraulde,C.R.,R.A.Brown,and N. J. Rosenberg,USregional agricultural production in 2030 and 2095:Response to CO2 fertilization and Hadley climatemodel (HadCM2) projections of greenhouse-forced cli-matic change,PNNL-12252, Pacific Northwest NationalLaboratories,Richland,Washington,1999.

Johnson, D. G., The World Grain Economy and ClimateChange to the Year 2000:Implications for Policy,National Defense University Press,Washington,DC,50pp.,1983.

Kaiser, H.M.,S. J. Riha, D. S.Wilks,D .G. Rossier, and R.Sampath,A farm-level analysis of economic and agro-nomic impacts of gradual warming, American Journalof Agricultural Economics, 75, 387-398,1993.

Kane,S., J. Reilly, and J.Tobey, Climate change:Economic implications for world agriculture,AER No.647,US Department of Agriculture,Natural Resourcesand Environmental Division,Economic ResearchService,Washington,DC,1991.

LITERATURE CITEDAdams,R.M.,C.Rosenzweig,R.M. Peart, J.T. Richie,B.A.McCarl, J. D. Glyer, R.B.Curry, J.W. Jones,K. J. Boote,andL.H.Allen,Global climate change and US agriculture,Nature, 345, 219-224,1990.

Adams,R.M.,S.A.Hamilton,and B.A.McCarl,The Benefitsof Air Pollution Control:The Case of Ozone and USAgriculture, American Journal of AgriculturalEconomics, 68:886-894,1986.

Adams,R.M.,R.A.Fleming,C.C.Chang,B.A.McCarl,andC.Rosenzweig,A reassessment of the economic effectsof global climate change on US agriculture, ClimaticChange, 30, 147-167,1995.

Adams, D.,R. J.Alig, B.A.McCarl, J. M.Callaway, and S.Winnett,The forest and agricultural sector optimizationmodel: Model structure and applications,USDA ForestService Research Paper PNW-RP-495,USDA ForestService, Pacific Northwest Experiment Station, Portland,Oregon,1997.

Adams,R.M.,B.H.Hurd,and J. Reilly,A review ofimpacts to US agricultural resources, Pew Center onGlobal Climate Change,Arlington, VA,36 pp.,1999a.

Baumes,H.,A Partial Equilibrium Sector Model of USAgriculture Open to Trade:A Domestic Agricultural andAgricultural Trade Policy Analysis,Ph.D. thesis,PurdueUniversity, 1978.

Ben Mechlia, N.,and J. J. Carrol,Agroclimatic modelingfor the simulation of phenology, yield,and quality ofcrop production: 1.Citrus response formulation,International Journal of Biometeorology, 33, 36-51,1989.

Burton,R.O.,and M.A.Martin,Restrictions on HerbicideUse:An Analysis of Economic Impacts on USAgriculture.” North Central Journal of AgriculturalEconomics, 9:181-194,1987.

CAST (Council for Agricultural Science and Technology),Preparing US agriculture for global climate change, TaskForce Report No. 119, CAST,Ames,Iowa,96 pp.,1992.

Chang,C.C.,B.A.McCarl, J.W. Mjelde,and J.W.Richardson,Sectoral Implications of Farm ProgramModifications, American Journal of AgriculturalEconomics, 74:38-49,1992.

400


Kane,S., J. Reilly, and J.Tobey,An empirical study of theeconomic effects of climate change on world agricul-ture, Climatic Change, 21, 17-35,1992.

Kimball,B.A.,Carbon dioxide and agricultural yield: Anassemblage and analysis of 430 prior observations,Agronomy Journal, 75, 779-786,1983.

Kimball,B.A., P. J. Pinter, Jr.,R.L.Garcia,R.L.LaMorte, G.W.Wall,D .J. Hunsaker, G.Wechsung, F.Wechsung,and T.Kartschall,Productivity and water use of wheat underfree-air CO2 enrichment, Global Change Biology, 1,429-442,1995.

Lambert, D.K.,B.A.McCarl, Q. He,M.S.Kaylen,W.Rosenthal,C.C.Chang,and W.I.Nayda,Uncertain Yieldsin Sectoral Welfare Analysis:An Application to GlobalWarming, Journal of Agricultural and AppliedEconomics, 423-435,1995.

Lewandrowski, J. and D. Schimmelpfennig,Agriculturaladaptation to climate change:issues of long-run sustain-ability, Land Economics, 75(1),39-57,1999.

Long,S. P.,Modification of the response of photosyn-thetic productivity to rising temperature by atmospher-ic CO2 concentrations: Has its importance been under -estimated?, Plant, Cell and Environment, 14(8),729-739,1991.

McCarl,B.A.and T.H.Spreen,Price EndogenousMathematical Programming As a Tool For SectorAnalysis, American Journal of AgriculturalEconomics, 62:87-102,1980.

Mendelsohn,R.,W. D. Nordhaus,and D. Shaw,Theimpact of global warming on agriculture: A Ricardiananalysis, The American Economic Review, 84(4),753-771,1994.

Mitchell,R.A.C.,V.J. Mitchell,S.P. Driscoll, J. Franklin,andD.W. Lawlor, Effects of increased CO2 concentrationand temperature on growth and yield of winter wheatat two levels of nitrogen application, Plant, Cell, andEnvironment, 16, 521-529,1993.

Office of Technology Assessment, Preparing for anUncertain Climate, Office of Technology Assessment,US Congress,Washington,DC,1993.

Parry, M.L.,Agriculture and forestry, in ClimateChange: The IPCC Impacts Assessment, edited by W. J.McG.Tegarrt, G.W. Sheldon,and D. C.Griffiths,UNIntergovernmental Panel on Climate Change, AustralianGovernment Printing Office,Canberra,1990a.

Parry, M.L., Climate Change and World Agriculture,Earthscan Publications Ltd.,London,England,1990b.

Parton,W. J., D. S.Schimel,and D. S.Ojima,Environmental change in grasslands: Assessment usingmodels, Climate Change, 28(1-2),111-141,1994.

Patterson D.T., J. K.Westbrook,R. J.V. Joyce,and P. D.Lindgren,Weeds,insects,and diseases, ClimaticChange, 43(4),711-727,1999.

Reilly, J. (Guest Editor),Climate Change,Impacts onAgriculture, Climatic Change, 43(4),645-793,1999.

Reilly, J.,et al.,Changing climate and changing agricul-ture:Report of the Agricultural Sector AssessmentTeam, available at:http://www.nacc.usgcrp.gov/sec-tors/agriculture/working-papers.html,in review, 2000.

Reilly, J., N. Hohmann,and S.Kane,Climate change andagriculture: Global and regional ef fects using an eco-nomic model of international trade,MIT-CEEPR 93-012WP, Center for Energy and Environmental PolicyResearch,Massachusetts Institute of Technology,Cambridge,Massachusetts,August 1993.

Reilly, J., N. Hohmann,and S.Kane,Climate change andagricultural trade: who benefits,who loses?, GlobalEnvironmental Change, 4(1),24-36,1994.

Reilly, J. et al.,Agriculture in a changing climate:Impacts and adaptations,in Climate Change 1995:Impacts,Adaptations, and Mitigation of ClimateChange, edited by R. T.Watson,M.C.Zinyowera,and R.H.Moss,Cambridge University Press,Cambridge,United Kingdom,pp.427-469,1995.

Roningen,V.,A Static World Policy Simulation (SWOP-SIM) Modeling Framework,Staff Rep.No.AGE860265,US Department of Agriculture,Economic ResearchService,Washington DC,1986.

Rosenberg, N.J.,1993.Towards an integrated assess-ment of climate change:the MINK study. ClimaticChange., 24: 1-175.

Rosenzweig,C.,and D. Hillel, Climate Change and theGlobal Harvest: Potential Impacts on the GreenhouseEffect on Agriculture, Oxford University Press,NewYork,1998.

Rosenzweig,C.,A.Iglesias,X.B.Yang, P. R.Epstein,andE.Chivian,Climate change and US agriculture: Theimpacts of warming and extreme weather events onproductivity, plant diseases,and pests,Center for Health

401


Southworth, J., J. C.Randolph,M.Habeck, O. C.Doering,R.A.Pfeifer, G. Rao,and J. Johnston,Consequences offuture climate change and changing climate variabilityon corn yields in the Midwest US, Agriculture,Ecosystems, and Environment, in press,2000.

Strzepek,K.M., D. C.Major, C.Rosenzweig,A.Iglesias, D.N.Yates,A.Holt,and D. Hillel,New methods of model-ing water availability for agriculture under climatechange: The US Cornbelt, Journal of the AmericanWater Resources Association, 35, 1639-1655,1999.

Takayama,T. and G.G. Judge, Spatial and TemporalPrice and Allocation Models, North-Holland PublishingCompany, 1971.

Tobey, J., J. Reilly, and S.Kane,Economic implications ofglobal climate change for world agriculture, Journal ofAgricultural and Resource Economics, 17(1),195-204,1992.

Tsuji, G.Y.,G. Uehara,and S.Balas (Eds.),DSSAT v3,University of Hawaii,Honolulu,Hawaii,1994.

Tubiello, F. N.,C.Rosenzweig,B.A.Kimball, P. J. Pinter,Jr.,G.W.Wall, D. J. Hunsaker, R.L.Lamorte,and R.L.Garcia,Testing CERES-wheat with FACE data: CO2 andwater interactions, Agronomy Journal, 91, 247-255,1999.

US Department of Agriculture, Agricultural Statistics,US Government Printing Office,Washington, D.C.,1987.

US Department of Agriculture,FEDS Budgets,EconomicResearch Service,1982.

Walker, B.,and W. Steffen (Eds.), Global Change andTerrestrial Ecosystems, IGBP Book Series,CambridgeUniversity Press,Cambridge,United Kingdom,620 pp.,1996.

and the Global Environment,Harvard University,Cambridge,Massachusetts,pp.46,2000.

Rosenzweig,C.,and M.L. Parry, Potential impact of cli-mate change on world food supply, Nature, 367,133-138,1994.

Rosenzweig,C., J.T. Ritchie, J.W. Jones, G.Y.Tsuji,and P.Hildebrand,Climate change and agriculture: Analysis ofpotential international impacts, American Society ofAgronomy Special Publication No. 59, AmericanSociety of Agronomy, Madison,Wisconsin,382 pp.,1995.

Rosenzweig,C.,and F. N.Tubiello,Impacts of future cli-mate change on Mediterranean agriculture: Currentmethodologies and future directions, Climate Change,1, 219-232,1997.

Schimmelpfennig, D., J. Lewandrowski, J. Reilly, M.Tsigas,and I. Parry,. Agricultural adaptation to climatechange: issues of long-run sustainability, AgriculturalEconomics Report No. 740, Economic ResearchService,USDA,Washington DC, June 1996.

Shriner, D. S.,and R.B.Street,North America,in TheRegional Impacts of Climate Change: An Assessmentof Vulnerability, edited by R.T.Watson,M.C.Zinyowera,and R.H.Moss,Cambridge University Press,Cambridge,United Kingdom,pp.253-330,1998.

Sionit, N., D.A.Mortensen,B.R.Strain,and H.Hellmers,Growth response of wheat to CO2 enrichment and dif-ferent levels of mineral nutrition, Agronomy Journal,73, 1023-1027,1981.

Smith, J. B.,and D.Tirpak (Eds.),The potential effects ofglobal climate change on the United States,EPA-230-05-89-050,US Environmental Protection Agency,Washington,DC,1989.

402


ACKNOWLEDGMENTSMany of the materials for this chapterare based on contributions from participants on and

those working with the

Agriculture Sector Assessment TeamJohn Reilly*,Massachusetts Institute of TechnologyJeff Graham*,US Department of Agriculture

(through Sept.1999)James Hrubovcak*,US Department of Agriculture

(from October 1999)David G.Abler, Pennsylvania State UniversityRobert A.Brown,Battelle-Pacific Northwest National

LaboratoryRoy Darwin,US Department of AgricultureSteven Hollinger, University of IllinoisCesar Izaurralde,Battelle-Pacific Northwest National

LaboratoryShrikant Jagtap,University of Florida-GainesvilleJames Jones,University of Florida-GainesvilleJohn Kimble,US Department of AgricultureBruce McCarl,Texas A&M UniversityLinda Mearns,National Center for Atmospheric

ResearchDennis Ojima,Colorado State UniversityEldor A.Paul,Michigan State UniversityKeith Paustian,Colorado State UniversitySusan Riha,Cornell UniversityNorm Rosenberg,Battelle-Pacific Northwest

National LaboratoryCynthia Rosenzweig, NASA-Goddard Institute for

Space StudiesFrancesco Tubiello, NASA-Goddard Institute for

Space Studies


403


405

CHAPTER 14

1Arizona Department of Water Resources; 2Coordinating author for the National Assessment Synthesis Team; 3USGeological Survey, 4Pacific Institute for Studies in Development,Environment and Security.

POTENTIAL CONSEQUENCES OF CLIMATEVARIABILITY AND CHANGE FOR THEWATER RESOURCES OF THE UNITEDSTATESKatharine Jacobs1,2, D. Briane Adams3, and Peter Gleick4


Chapter Summary

Background

Socioeconomic and Institutional Context


Key Issues

Competition for Water Supplies

Surface Water Quality

Groundwater Quantity and Quality

Heavy Precipitation,Floods,and Drought

Ecosystem Vulnerabilities



Literature Cited

Acknowledgments


• Increases in global temperatures have beenaccompanied by more precipitation in the midand high latitudes.

• Precipitation has increased an average of 10%across the US,with much of the increase attrib-uted to heavy precipitation events.

• Nationally, streamflow has increased about threetimes more than the increase in precipitation.Regionally, the higher streamflows haveincreased in many areas,but not in the Westwhere snowmelt dominates peak flows.

• Reductions in areal extent of snowpack in thewestern mountains have been observed,alongwith substantial retreat of glaciers.

• In snowpack-dominated streams,a shift has beenobserved in the timing of the peak runoff to ear-lier in the season.

• No significant increases in the frequency ofdroughts or winter-type storms have beenobserved on a national basis.


• Historic trends towards increased precipitationare very likely to continue.

• It is possible that there will be an increase ininterannual variability, resulting in more severedroughts in some years.

• The Canadian and Hadley climate models used inthe Assessment generally do not agree on precip-itation impacts,with the exception of showingan increase in precipitation in the Southwest.

• Increases in temperature, even in the context ofincreases in precipitation,are likely to result insignificant loss of soil moisture in the NorthernGreat Plains.

• Snowpack is very likely to be reduced even inthe context of higher precipitation.

• If the number of high intensity storm eventsincreases, flushing of contaminants into water-sheds is likely to increase,causing episodic waterquality problems.

• Quality and quantity impacts are very likely to beregionally specific.

• Surprises are likely, since many water-relatedimpacts cannot be predicted.

CHAPTER SUMMARYContext

Water supply conditions in all regions and sectors inthe US are likely to be affected by climate change,either through increased demands associated withhigher temperatures,or changes in precipitation andrunoff patterns. Water sector concerns includeeffects on ecosystems,particularly aquatic systemssuch as lakes,streams, wetlands,and estuaries.Although competition for water supplies is extreme-ly intense,particularly in the western US,substantialability to adjust to changing demands for waterexists in the current water management system. Itis not known whether the effects of climate changewill require dramatic changes in infrastructure tocontrol flooding and provide reliable water suppliesduring drought. However, it is known that precipita-tion and temperature changes are already increasingrunoff volumes and changing seasonal availability ofwater supply, and that these changes are likely to bemore dramatic in the future.


406

Key Findings

• More pressure on surface water supplies is likelyto come from population shifts and changes inwater right allocations to accommodate endan-gered species and the water rights of NativeAmericans. Although wetter conditions in theSouthwest may alleviate some of these stresses,stress is likely to increase in the Northern GreatPlains and in snowpack-dependent watersheds.

• Groundwater supplies are already over-drafted inmany parts of the country, and pressure ongroundwater supplies is likely to increase to off-set changes in surface water supply availability.However, long-term increases in precipitationwill possibly increase recharge rates in someareas.

• It is likely that aquatic and riparian ecosystemsmay be damaged even in the context of higherprecipitation,due to higher air temperatures andreduced summer flows. It is also probable thatchanges in water temperature in lakes andstreams will affect species composition.

• Water managers have multiple opportunities toreduce future risks by incorporating “no-regrets”changes into their operating strategies that areappropriate regardless of climate change.

• Institutions governing water rights are generallyvery inflexible,and are likely to prove to beobstacles to adaptation.

• Improvements are needed in monitoring effortsto identify key impacts related to water quantityand quality, biological conditions of key habitats,snowpack conditions,and groundwater supplies.

Chapter 14 / Water Resources

407

Increases in greenhouse gas concentrations are verylikely to affect global temperature and lead tochanges in the amount,timing,and geographic dis-tribution of rain,snowfall,and runoff. Changes arealso likely in the timing,intensity, and duration ofextreme events such as floods and droughts. Suchchanges possibly will have greater impacts on theregions and sectors than changes in average temper-ature or precipitation. Higher demand for water isprobable in areas where increased temperatureresults in higher evapotranspiration. Although inmost regions it is possible that increased streamflowwould relieve current stress,some regions are likelyto experience greater difficulty in meeting theirwater supply needs if precipitation increases do notoffset increases in evapotranspiration. Key variablesin determining likely impacts and responses includechanges in soil moisture and cloud cover, seasonalityof precipitation,and the response of vegetation tochanges in moisture,temperature,and increased car-bon dioxide availability.

While many of the most significant impacts in theagricultural, forestry, ecosystem,energy, and humanhealth sectors relate to the basic issue of wateravailability, it is likely that there will be some seriousimpacts on water quality as well. There is a directrelationship between quantity of flows and dilutionof pollutants in surface water;higher runoff is likelyto improve water quality, but increased intensity ofrainfall will probably result in increased erosion andflushing of contaminants into watersheds. Higherwater temperature will affect the ecology of wet-lands,lakes,and streams. Much less research hasbeen done on impact-related issues.

The primary water resource issue for the US is thedistribution of supply and demand,not the totalquantity of water available. The nature of water con-cerns varies by region across the country. For muchof the western US, water resources are often sepa-rated both by time and distance from waterdemands. As a result,substantial infrastructure hasbeen developed to store and transport water sup-plies (for example,from the Colorado River andNorthern California to the Southwest and SouthernCalifornia). There are more than 80,000 dams and

POTENTIAL CONSEQUENCES OF CLIMATEVARIABILITY AND CHANGE FOR THE WATERRESOURCES OF THE UNITED STATES

BACKGROUNDWater-related concerns are central to this NationalAssessment because the hydrologic (water) cycle isa fundamental component of climate and becausewater plays a role in every sector and region in theUS. Despite many remaining uncertainties,a signifi-cant amount of research has been done on the con-nections between climate change and waterresources in the US;a searchable bibliography ofalmost 900 scientific articles is available athttp://www.pacinst.org.

The US has a wide variety of tools,institutions andmethods for coping with water resource problems,and many of these will be useful for addressing theimpacts of climate changes. Water managers alreadydeal with climate variability; reservoirs are designedwith some flexibility for extreme high and lowflows;techniques and technologies are available formanaging water demands. But global climatechange raises some unresolved concerns for thewater sector. What will be the economic costs ofcoping with climate changes imposed on top ofexisting variability? Are existing institutions suffi-ciently flexible to handle the additional stresses?What might be the nature of unexpected climate“surprises” for the water sector? And will watermanagers be willing and able to prepare in advancefor conditions different from those they are normally faced with?


408

Figure 1: The Central Arizona Project brings Colorado River water330 miles uphill to Tucson and Phoenix, Arizona. Source: K. Jacobs


409

Inadequate water and wastewater infrastructure,common along the Mexican border and in somerural areas,leads to high risk for health problems.Because the public has very high expectationsregarding water quality, and perceptions of healthrisks are not always accurate,health issues may havea high profile and require particular attention.

Health-related issues that have been linked tochanges in the hydrologic cycle include potentialfor increases in water-borne pathogens such asCryptosporidium and vector-borne diseases such asencephalitis,as well as outbreaks in marinepathogens associated with red tide (Bernard et al.,1999). Hantavirus,a disease spread by deer mice,has also been linked to extreme climate variabilityassociated with the El Niño Southern Oscillation(ENSO). With higher rainfall, rodent populationstend to increase,which increases the chance ofhuman contact and disease.

Virtually all indices of vulnerability relative to waterhave identified the over-appropriation of westernstreams and rivers and over-drafting of groundwatersupplies as key issues. Gleick (1990) identified indicators of water resource vulnerability for the USand found that the most vulnerable regions werethe high irrigation areas along the eastern drainageof the Rocky Mountains,the Central Valley of California,and Southern California. The overallindex prepared by Hurd et al.(1999a) indicates thatthe most vulnerable watersheds are in the West,Southwest,and Great Plains (see Figure 2).

reservoirs in the US,and millions of miles of canals,pipes,and tunnels (Schilling et al.,1987,see Figure1). Although this infrastructure is sophisticated andhas allowed the development of urban and agricul-tural areas,it is also a source of vulnerability to cli-mate change,partially because it has been designedbased on the assumption that future conditions willbe similar to the historically observed climate. Someargue that there is substantial robustness built intothe system that provides some margin of safety, butfailure to re-evaluate these assumptions and identifykey vulnerabilities may prove to be costly in thefuture.

Water supply issues in the eastern US relate to aginginfrastructure and inadequate storage capacity duringtimes of drought. Flood control issues and environ-mental impacts of structural solutions are also of con-cern. In general,local surface water and groundwa-ter supplies are available for domestic and industrialuse without major water transfers between basins,but excess reservoir capacity to respond to droughtis quite limited. In some areas,such as New YorkCity, reservoir function is threatened by upstreamdevelopment. New York and other cities also haveserious problems on a regular basis with water mainbreaks causing flooding and other damage.

The initial charge of this Assessment included identi-fying areas of existing stress and vulnerability andevaluating new problems that climate change maybring. This has necessarily resulted in an identifica-tion of negative effects,though certain aspects of cli-mate change are likely to improve conditions insome areas of the US. Even in the absence of climatechange,adapting to existing stresses (such as aginginfrastructure,inadequate water supplies for areas ofrapid growth,etc.) and increased pressures frompopulation dynamics would be expensive. Frederickand Schwarz (1999a) estimate that the annualizedwater-related costs associated with the demands ofan increasing population are likely to approach $13.8billion by 2030. The impacts of climate change onthese costs depend on the nature of the changes.The estimated costs include investments in newwater supplies and conservation measures as well asthe impacts on streamflows and irrigated lands. Thecosts would be much higher if climate change wereto significantly decrease water availability (as underthe Canadian climate model scenario) or increase themagnitude and timing of extreme events. This isbecause the current infrastructure and managementpractices are designed based on the historical climateconditions.

Current Climate Vulnerability Map, Water Supply,Distribution and Consumptive Use

RelativeVulnerability

Figure 2: Assessed vulnerability based on current climate andwater resource conditions, based on data describing the following:share of streamflow withdrawn for use, streamflow variability,evapotranspiration rate, groundwater overdraft, industrial usesavings potential, and water trading potential. Source: Hurd, B.J.,N. Leary, R. Jones and J. Smith. (1999a).See Color Plate Appendix

Hydrologic Unit Boundary

State Boundary

Low Vulnerability <1.5

Medium Vulnerability ≥1.5

High Vulnerability >2.2

The state to state variations among water rights sys-tems results in substantial complexity. In general,water rights in eastern states are not likely to be eas-ily quantifiable,which limits management options.The prior appropriation doctrine of the western USis relatively inflexible in dealing with changing envi-ronmental and societal needs (see box,“MajorDoctrines for Surface Water and Groundwater”).

Some innovative institutions are developing inresponse to particular problems. For example,the“temporary” water banks in California to respond todrought and in Arizona to respond to long-term sup-ply reliability issues offer some protection to exist-ing water rights while providing much-needed flexi-bility. Water banks generally provide opportunitiesfor short-term transfers of agricultural water sup-plies to municipal end users on a willing buyer/will-ing seller basis. In the case of the Arizona WaterBanking Authority, excess Colorado River water isbeing stored underground through recharge proj-ects to offset future shortfalls in supply. This oppor-tunity is expected to be available on an interstatebasis among the Lower Colorado Basin states in thenear future. Similar types of contingency planningbetween jurisdictions and water rights holderscould prove beneficial in responding to short-termemergencies. Longer-term changes in climatic con-ditions that would require permanent changes tolegal systems could be more problematic.

Many have argued that an open market in waterrights would help resolve conflict and increase effi-ciency because water would flow to the highest andbest use based on willingness to pay (NationalResearch Council,1992;Western Water PolicyReview Advisory Commission,1998). It is widelyacknowledged that market-related solutions mayrelieve some water supply problems,especially inthe West. However, water marketing is an imperfectsolution. Of particular concern are third partyimpacts in water transfers,and overall equity issues.Water markets are developing in many states,butthey are generally regulated markets in order to pro-tect the public interest. Mechanisms exist to identi-fy economic values for non-market goods and servic-es,but water rights for non-market values such asecosystems,aesthetics,and recreation have difficultycompeting with major economic forces. There isalso a risk that disproportionate burdens will beplaced on the social groups that can least affordthem (such as rural farming communities,NativeAmericans,and communities along the Mexican bor-der with inadequate infrastructure) (Dellapenna,1999b;Gomez and Steding,1998).

A rise in average temperature, even in the context ofhigher precipitation,is most likely to impact aquaticsystems,including riparian habitat,and freshwaterand estuarine wetlands. In some cases,this isbecause expected changes in precipitation do notoffset increased evapotranspiration,though seasonaland regional impacts are likely to vary. Certaincoastal systems,prairie potholes (small ponds andlakes formed by glacial deposits),and Arctic andalpine ecosystems are thought to be especially vul-nerable. Stresses within the contiguous US are likelyto come from changes in the distribution of precipi-tation as well as increases in its intensity.

SOCIOECONOMIC ANDINSTITUTIONAL CONTEXTPopulation pressures,including shifts towards west-ern and coastal urban areas,land use practices,andclimate change are all likely to increase stress onwater supply systems. The need to reserve water forinstream uses,endangered species protection, recre-ation,and American Indian water rights settlementsalso places new demands on a water rights systemthat in many parts of the country is already seriouslystressed. As society changes,its value system alsoevolves. Placing more value on protection of fishand wildlife habitat and recreational values is likelyto force institutional change at the same time thatnew stresses are appearing due to climate variabilityand change. Although there is substantial uncertain -ty in the projections of changes in runoff that arederived from the climate models,socioeconomicconditions are even less predictable.

There is a need for more flexible institutionalarrangements and more effective ways of makingwater policy decisions in order to adapt to changingconditions (not just changes in climate,but multipleexisting stresses). The legal framework for waterrights varies from state to state,with nearly infinitepermutations at the local level. The one characteris-tic that is typical of most institutions related towater is inability to respond efficiently to changingsocioeconomic and environmental conditions. Thisis primarily because institutions tend to reflect exist-ing water right holders’interests,and substantialinvestments are made based on expectations regard-ing availability of supplies. Devising new legal andrelated institutions that can introduce the necessaryflexibility into water management without destabi-lizing investors’ expectations,while at the same timeincorporating public values (ecological, recreational,aesthetic,etc.) is a significant challenge.


410


411

Dellapenna (1999a) says true (unregulated) marketshave seldom existed for water rights and there aregood reasons for believing that they seldom will.This is because water, like air, is viewed as a “publicgood,” which means that people cannot realisticallybe excluded from using it,at least on a subsistencebasis. There is reluctance to pay the full cost ofwater, including the replacement cost;people aregenerally charged only the cost for capturing anddistributing water. Key factors in developing aworkable market are whether the market willenable consumers to meet their needs and whethergovernment regulation and assistance at the marginscan correct for market failures.

Numerous institutional issues related to respondingto potential climate change stem from the fact thatvarious agencies and levels of government handlewater quality and water quantity issues. Water quali-ty regulation derives primarily from federal authori-ties such as the Clean Water Act and the Safe

Drinking Water Act. Water quantity regulation(through rights,allocations,and permits) is primarilyhandled by the states. An illustration of this prob-lem is found in a study by Eheart et al.(1999),which evaluated the impact of reduced precipita-tion in the Midwest on ability to meet federal dis-charge water quality standards. They found that a25% reduction in precipitation could reduce thecritical dilution flow that determines dischargeimpact on water quality by 63%. They concludedthat this has implications for the process of settingTotal Maximum Daily Loads (TMDL) for Non-PointDischarge Elimination Permits under the CleanWater Act. Section 303(d) requires states to identifywaters that do not meet water quality standards andto establish plans to achieve the TMDL standards.Eheart et al.(1999) noted that the present regulato-ry scheme in many midwestern states is not sophis-ticated enough to take into account the interplaybetween water quality and quantity.

Major Doctrines for Surface Water and Groundwater

Surface WaterRiparian doctrine – Authorization to use water in a stream or other water body is based on ownership of theadjacent land. Each landowner may make reasonable use of water in the stream but must not interfere withits reasonable use by other riparian landowners. The riparian doctrine prevails in the 31 humid states east ofthe 100th meridian.

Prior appropriation doctrine – Users who demonstrate earlier use of water from a particular source acquirerights over all later users of water from the same source. When shortages occur, those first in time to divertand apply the water to beneficial use have priority. New diversions,or changes in the point of diversion orplace or purpose of use, must not cause harm to existing appropriators. The prior appropriation doctrine pre-vails in the 19 western states.

GroundwaterAbsolute ownership – Groundwater belongs to the overlying landowner, with no restrictions on use and noliability for causing harm to other existing users. Texas is the sole absolute ownership state.

Reasonable use doctrine – Groundwater rights are incident to land ownership. Owners of overlying land areentitled to use groundwater only to the extent that the uses are reasonable and do not unreasonably interferewith other users. Most eastern states and California subscribe to this doctrine. Some states,such as Arizona,have modified the reasonable use doctrine by requiring state permits to use groundwater in certain high useareas.

Prior appropriation permit system – Groundwater rights are determined by the rule of priority, which pro-vides that prior uses of groundwater have the best legal rights. States administer permit systems to determinethe extent to which new groundwater uses will be allowed to interfere with existing uses. Most westernstates employ some form of permit system.

Sources:US Army Corps of Engineers, Volume III,Summary of Water Rights – State Laws and AdministrativeProcedure report prepared for US Army, Institute for Water Resources, by Apogee Research,Inc., June 1992;and US Geological Survey, National Water Summary 1988-89-Hydrologic Events and Floods and Droughts,Water Supply Paper 2375 (Washington,DC,US Government Printing Office,1991).


412

The institutions that have been successful in manag-ing water resources tend to use an integratedapproach to management,and to incorporate natu-ral watershed boundaries rather than politicalboundaries for management areas. Relatively inno-vative water management districts have beenformed in many states to address specific resourceconditions (see box,“Tennessee Valley Authority –Integrated Water Resources Management”).

CLIMATE VARIABILITY ANDCHANGEHistoric trends show that the surface temperatureof the Earth has increased by about 1˚F (just over0.6˚C) over the 20th century, with 1998 the warmestyear on record. Higher temperatures have resultedin reductions in snow cover and sea ice extent.

70.0

60.0

50.0

40.0

30.0

20.0

10.0

0.0

y = -0.1167x + 44.308

Figure 4: In some western watersheds, runoff timing appears to beshifting from spring to winter, suggesting a change in snowfall andsnowmelt dynamics. Source: Gleick, P.H. and E.L. Chalecki. (1999)JAWRA. Dec. pp. 1429-1442.

Sacramento River Runoff

The Tennessee Valley Authority –Integrated Water Resources Management

A well-known example of integrated water resources management is the Tennessee Valley Authority (TVA),which operates in a seven-state area in the southeastern US. Founded in 1933,TVA pioneered the concept of"unified river basin development" within the Tennessee River Basin,integrating water resources development,social and economic development,power production,and natural resources conservation.

TVA’s water management programs focus on the operation of a large, multipurpose reservoir system thatincludes more than 50 dams and reservoirs. The system is operated as an integrated unit to provide for navi-gation, flood control, hydropower generation,summer recreation levels,and minimum flows for the mainte-nance of water quality and aquatic habitat. In one of the most flood-prone areas in the US,TVA has historical-ly taken a dual approach to flood management that combines reservoir system control with a floodplain man-agement program to encourage appropriate shoreline development. Environmental concerns are integratedinto reservoir operations,while TVA’s Watershed Teams work at grassroots levels to motivate local action tocontrol non-point source pollution. TVA alsomaintains web-sites and special telephone sys-tems to facilitate public access to streamflowdata,dam release information,and other systeminformation.

TVA has a sophisticated streamflow and rainfalldata collection and monitoring system,coupledwith a state-of-the-art simulation and optimizationmodeling system. Monitoring and forecastingoccur on a continuous,24-hour basis.Additionally,TVA utilizes 10-day and seasonalweather forecasts to guide reservoir planning.TVA is engaged in joint work with NOAA to bet-ter utilize seasonal forecast information. Thesecapabilities will assist TVA in adapting to climatechange and variability.

Figure 3: TVA has more than 50 dams in seven states. Source:Tennessee Valley Authority


413

Understanding historic changes,or projecting futurechanges in streamflow conditions will require moreevaluation of the complex role of changing precipi-tation and temperature patterns as well as the roleof land-use change on streamflow. These unresolvedissues further reinforce the importance of maintain-

Trend analysis shows that since WWII,there hasbeen significant retreat of snow cover in spring(Groisman,1999). At the same time,there appear tobe shifts in the seasonality of runoff in some west-ern rivers consistent with what would be expectedfrom changes in snowfall and snowmelt dynamicsdue to warming (Gleick and Chalecki,1999;Union of Concerned Scientists,1999) (see Figure 4).

Increases in global temperatures have been accom-panied by more precipitation in the mid and highlatitudes (based on Northern Hemisphere land-sur-face records) and increases in atmospheric watervapor in many regions of North America and Asiawhere data are adequate for analysis (IPCC,1996).Karl et al.(1996) show that the meteorologicaldrought indices suggest that there have been morewet spells,but no significant changes in drought ona national basis. The precipitation increase in theUS has been attributed primarily to an increase inthe heaviest precipitation events (Karl and Knight,1998;Groisman et al.,1999;Karl et al.,1996). Thesechanges are statistically significant and most appar-ent during the spring,summer, and autumn monthsin the contiguous US. Based on recent work by Linsand Slack (1999),the warm season precipitationincreases may be responsible for increases instreamflow in the low to moderate range (i.e.theflow values that are most commonly observed dur-ing the summer and early autumn months). Usingdischarge data from a national network of streamgages for the period 1944-1993,Lins and Slack(1999) found statistically significant increases in theannual median streamflow at 29% of the streamgages nationwide and decreases at only 1% of thestream gages. Most trends were even more positivefor the lower streamflow quantiles. Fewer signifi-cant trends were observed in high streamflows.Only 9% of the gages, for example,had significanttrends in the annual maximum streamflow and,ofthese,more showed decreases than increases.Groisman et al.(2000) show that high streamflow inthe mountainous West has not changed despiteincreases in heavy precipitation events. They attrib-ute this to a trend toward reduced snow coverextent leading to a lower and earlier peak in theannual cycle of runoff. In the East and South how-ever, increasing trends in high and very high stream-flow are shown to relate to increases in heavy andvery heavy precipitation events. In fact,there is anamplification of the trends in precipitation acrossthe highest precipitation and streamflow rates by afactor of about three. It is well-recognized (Karland Reibsame,1989) that small changes in precipita-tion can be amplified into large changes in stream-flow (see Figure 5).

Observed Changes In Streamflow and Precipitation (1939-99)

0 20 40 60 80 100Light/Low Moderate Heavy/High

22.5

20

17.5

15

12.5

10

7.5

5

2.5

0

-2.5

Projected 21st Century Change in US Daily Precipitation

Canadian Climate ModelHadley Climate Model

Figure 5: The graph shows changes in the intensity of precipitationand streamflow, displayed in 5% increments, during the period1939-99 based on over 150 unregulated streams across the US withnearby precipitation measurements. As the graph demonstrates,the largest changes have been the significant increases in theheaviest precipitation events and the highest streamflows. Notethat changes in streamflow follow changes in precipitation, but areamplified by about a factor of 3. Source: Groisman, et.al. (2001).

Figure 6: These projections from the Hadley and Canadian modelsshow the changes in precipitation over the 21st century. Each mod-els’ projected change in the lightest 5% of precipitation events isrepresented by the far left bar and the change in the heaviest 5% bythe far right bar. As the graph illustrates, both models project sig-nificant increases in heavy rain events with smaller increases ordecreases in light rain events. Source: National Climatic DataCenter.See Color Plate Appendix

ing adequate nationwide networks of precipitationand streamflow gages to help describe and predictchanges in average streamflow and,more important-ly, streamflow variability.

In addition to a trend analysis of climatic conditionsover the past 100 years,this Assessment has evaluat-ed scenarios from two General Circulation Models(GCMs),one from the Canadian Centre for ClimateModelling and Analysis (henceforth referred to asthe Canadian model),and the second,the “HadCM2”model from the Hadley Centre for ClimatePrediction and Research of the MeteorologicalOffice of the United Kingdom (henceforth referredto as the Hadley model). The Canadian and Hadleymodels used in this Assessment both project signifi-cant warming (5-9˚F or 3-5˚C) in most parts of theUS by 2090. However, with the exception of thesouthwestern US,where both models show a largeincrease in precipitation in the future,especially inwinter (Felzer and Heard,1999),the changes in pre-cipitation predicted by the two models are striking-ly different. In general,the Hadley model suggestsmuch wetter conditions than the Canadian model.When comparing output of multiple GCMs,theHadley model increases in precipitation are themost extreme. The precipitation increase inSouthern California and Arizona is related to increas-es in sea surface temperatures in the eastern Pacificand southward shifts in the jet stream that looselyresemble the El Niño pattern. Differences between


414

the two models are explained in part by differencesin the land surface models relating to soil moistureand moisture availability in summer (Felzer andHeard,1999) (see Figure 6 and figures on precipita-tion change in the Climate and West Chapters).

Precipitation is a key climatic variable,but it is diffi-cult to predict changes at the local level becausethey are affected by land surface features that are atsmaller scales than the GCM outputs (Felzer andHeard,1999). Precipitation itself is a sub-grid-scaleprocess,meaning that clouds and convection occuron scales smaller than GCM grids. Both modelsshow increases in heavy precipitation events andincreased storminess over the eastern Pacific,off theWest Coast of the US (Lambert,1995;Carnell andSenior, 1998; Felzer and Heard,1999). However, pre-cipitation patterns will vary regionally. An importantissue for improving the utility of GCM output is theability to downscale the models to a regional orwatershed level where the information can be mostuseful to water managers.

Differences in temperature and moisture levels overland and sea are crucial in determining precipitationlevels along the coasts. Over oceans, warming leadsto increased evaporation and more precipitationbecause of the limitless supply of water. In contrast,because of the limited moisture holding capacity ofthe land, warming may cause drying and less precip-itation. Globally, the models show decreased stormfrequency, with increases in intensity. Over the US,the models do not produce a consistent projectionregarding storm frequency (Lambert,1995;Carnelland Senior, 1998; Felzer and Heard,1999). Notrends have been identified in North America-widestorminess or in storm frequency variability in theperiod 1885-1996 (Hayden,1999a).

Wolock and McCabe (1999a) have used a water-bal-ance model and output from the two GCMs to esti-mate the effects of climate change on mean annualrunoff for the major water resource regions of theUS. The model includes the concepts of climaticwater supply and demand,seasonality in climaticwater supply and demand,and soil-moisture storage.Inputs to the model are monthly precipitation and potential evapotranspiration,which is calculatedfrom monthly temperature using the Hamon equa-tion (Hamon,1961). To evaluate the model’s reliabil-ity to estimate mean annual runoff for the 18 water-resources regions in the coterminous US,VEMAP-gridded monthly climate data for 1951-80 were usedin conjunction with the water-balance model to esti-mate mean annual runoff. These estimated runoffdata were compared with measured data for the

Figure 7: The estimated percent changes in average annual runoffbased on the Canadian and Hadley models are not well correlated.The Canadian model predicts declines in runoff in all regions exceptCalifornia, while the Hadley model projects increases in mostregions, particularly in the Southwest. The models differ in precipi-tation predictions in part due to underlying model construction.Source: Wolock, D.M. and G.J. McCabe, 1999a. See Color PlateAppendix

Projected Changes in Average Annual Runoff Based on Two GCMs

Hamlet and Lettenmaier also evaluated the impacton various water management objectives of the pro-jected changes in streamflow (see Figure 9). From awater supply perspective,Hamlet and Lettenmaierfound that on average,comparing the base casewith output from four transient GCMs negativelyaffected four water resources objectives:non-firmhydropower production,irrigation,instream flow forfish,and recreation at Lake Roosevelt. The Hadleymodel also showed negative impacts on flood con-trol and navigation,due to the significantly wetterconditions in that model. Hamlet and Lettenmaiernoted that an adaptive strategy would be to shift the

same period. The water-balance model reasonablysimulated measured mean annual runoff for most ofthe water-resources regions. In general,the resultsfrom these two GCMs are not well correlated,andproject different changes in mean annualrunoff (see Figure 7). The difficulty of projectingcombined effects of changes in precipitation,tem-perature,and seasonality of events make projectionsof impacts based on GCM output uncertain.

On the other hand,both large-scale climate modelsand regional hydrologic models agree that ifchanges in temperature of the magnitude identifiedin the climate models occur, substantial changes inthe amount of precipitation that falls as snow versusrain and earlier melting of snowpack are very likelyto result in changes in the runoff regime (Frederickand Gleick,1999;Hamlet and Lettenmaier, 1999;McCabe and Wolock,1999;Leung and Wigmosta,1999). Snowpack is very likely to be reduced evenin the context of higher precipitation because ofthe warming trend (see Figure 8). The effects ofchanges in the timing and volume of runoff willprobably be felt in most sectors and regions that aresnowpack-dependent (Gleick,1998),althoughchanges in runoff regimes will probably be highlyregionally specific. For example,Leung andWigmosta (1999) assessed the effects of climatechange from the NCAR Community Climate Model(downscaled through a regional climate model) onthe American River and the Middle Fork Flatheadwatersheds in the Pacific Northwest. There was a61% reduction in snowpack on the American River,accompanied by a major shift in streamflow. On theMiddle Fork Flathead there was an 18% reductionand no major shift in runoff. In both watersheds,there was a higher frequency of extreme low andhigh flow events.

Hamlet and Lettenmaier (1999) used four GCMs toevaluate runoff implications of climate change forvarious watersheds along the Columbia River.Altered streamflow information was simulated andused to drive a reservoir model to evaluate impactson water management. Relatively large increases inwinter runoff volumes and reductions in wintersnowpack resulted in all cases. The March snowwater equivalent averaged 75-85% of the base casefor 2025,and 55 to 65% of the base by 2045. Theearlier snowmelt,coupled with higher tempera-tures, reduced runoff volumes in spring,summer,and early fall. The researchers found that whilehigher temperatures increase the potential evapo-transpiration, reduced soil moisture in the summeris likely to ultimately limit the actual evapotranspira-tion.


415

Canadian Model

Hadley Model

Southern Rocky MountainsPacific Northwest

Sierra NevadaCentral Rocky Mountains

Figure 8: Percentage change from the 1961-90 baseline in the April1 snowpack in four areas of the western US as simulated for the21st century by the Canadian and Hadley models. April 1 snowpackis important because it stores water that is released into streamsand reservoirs later in the spring and summer. The sharp reduc-tions are due to rising temperatures and an increasing fraction ofwinter precipitation falling as rain rather than snow. The largestchanges occur in the most southern mountain ranges and thoseclosest to the warming ocean waters. Source: McCabe, G.J. andD.M. Wolock. 1999. See Color Plate Appendix

Percentage Change in Snowpack


416

hydropower production period in the Columbia tothe summer, using stored winter flows. However, animportant consideration is that while re-operation ofthe reservoirs can improve conditions within onemanagement objective,impacts to one or moreobjectives cannot be avoided unless the total systemdemands are reduced. This is difficult to accomplishas the regional population increases.

As has been noted by many researchers attemptingto model regional impacts,inadequate spatial resolu-tion of climate models to capture the topographicfeatures is a key problem in downscaling from theglobal models to local hydrologic features. Theresulting disparity between GCMs in precipitationpredictions must be addressed before water man-agers will be confident of likely outcomes.However,major advances have been made in theuse of regional climate models to drive hydrologicmodels in the Pacific Northwest (Leung and Ghan,1999;Leung and Wigmosta,1999;Georgakakos et al.,1999). In the Southwest,the Regional ClimateSystem Model has been used since 1995 for 48-hourprecipitation and streamflow predictions with goodsuccess,including during the 1997-1998 El Niño sea-son. This is one of the first global-to-mesoscale-to-watershed-basin scale predictions of this type(Miller et al.,1999).

A key variable in predicting water supply conditionsis the impact of CO2. Higher CO2 tends to stimulateplant growth, resulting in feedback effects. Thewater requirements to support more biomass couldpossibly reduce the runoff associated with a givenlevel of precipitation. However, higher CO2 levelsincrease stomatal resistance to water vapor trans-port,which could decrease water use of plants(Frederick and Gleick,1999). Under arid conditions,an increase in biomass from elevated CO2 is likelyto reduce runoff to streams,thereby leaving morewater on the landscape. Under conditions of ample water for plant growth,elevated CO2 causes partialstomatal closure with a consequent decrease in tran-spiration per unit of leaf area. However, leaf arealikely will be increased,so it is difficult to predictoverall impacts on water use. The CO2 effect hasbeen observed in several ecosystems,and varies byspecies (Kimball,1983).

Natural variability in climate has been traced to anumber of phenomena related to ocean tempera-tures and changes in global circulation patterns.Some of the resulting weather patterns can now bepredicted with some accuracy, such as those associ-ated with the El Niño Southern Oscillation (ENSO).The ability to develop forecasts useful to water managers based on these patterns is increasing,allowing for adaptive responses. It is not yet clearhow these patterns will be affected by globalchanges in climate. Both models show more intensestorms,but they do not agree on changes in stormfrequency (Felzer and Heard,1999).

Changes in Reliability of Columbia RiverWater Resources Objectives

1009590858075706560

System Objective

Base Case

HC 2025

HC 2045

HC 2095

MPI 2025

MPI 2045

Figure 9: Four major objectives are impacted by low summerstreamflow and reservoir storage: non-firm energy production; irri-gation; instream flow; and recreation at Lake Roosevelt. Source:Hamlet, A.F. and D.P. Lettenmaier, 1999. See Color Plate Appendix

Figure 10: Rough estimate of how much snowlines in the PacificNorthwest are likely to shift by 2050, assuming about 4ºF warming.Source: R. Leung, Pacific Northwest National Laboratory.

Snow Level

Substantial water quality and temperature changescould result from changes in flow regimes. Itshould be noted that climate change could eitherincrease or decrease the availability of water. Whilethe hydrologic implications of the Canadian modelproject modest reductions in water supplies (<25%)in some regions,the Hadley model projects relative-ly large increases in water availability (25-50%) inmost regions of the US. However, there are signifi-cant regions of precipitation decrease throughoutthe US in both seasons in the Canadian model andin summer in the Hadley model. The increases inprecipitation are greater than the decreases princi-pally because of the large projected increase in theSouthwest (Felzer and Heard,1999).

KEY ISSUES Five key issues have been identified:

• Competition for water supplies• Surface water quality• Groundwater quality and quantity• Heavy precipitation, floods,and droughts• Ecosystem vulnerabilities

1. Competition for Water Supplies

Water supplyChanges in water supply availability for economicactivities and environmental uses are likely to beaffected by changes in average temperature and pre-cipitation as well as by altered frequency of extremeevents such as floods and droughts. There is generalconsensus among climate modelers that a warmerworld is very likely to lead to more precipitation atmid and high latitudes as well as an increase inheavy precipitation events in these areas. More pre-cipitation will typically lead to more runoff, but insome regions,higher temperatures and increases inevapotranspiration rates may possibly counteractthis effect. Several modeling studies for the westernUS show that precipitation rates would need toincrease by as much as 10-15% just to maintainrunoff at historical levels because of increased evap-otranspiration (Gleick and Chalecki,1999).

Changes in the timing of water supply availabilityare also very likely to occur. Surface water suppliesthat are dependent on snowmelt are likely to beaffected by changes in the amount of precipitationthat falls as rain versus snow, changes in snowpackvolume,and earlier melting due to warmer tempera-tures. There is a strong consensus amongresearchers that there is very likely to be a shift inthe peak volume and timing of runoff for water-

sheds that are affected by winter snowpack, result-ing in earlier spring runoff, higher winter flows andlower summer flows (Frederick and Gleick,1999).


417

Columbia Basin SnowExtent

(Washington & Oregon)

Figure 11: Complete loss of snow cover is projected at lower ele-vations. These maps are generated by downscaling output fromglobal to regional climate models. Output shown from these mod-els relates to the Columbia Basin; no projections are included forthe blank areas outside the basin. Source: Mote, et.al. (1999)Impacts of climate variability and change in the Pacific Northwest,University of Washington. See Color Plate Appendix

Current

20252045

Figure 12: Relative to present flows (dashed), the wetter wintersand drier summers simulated by climate models are very likely toshift peak streamflow earlier in the year, increasing the risk of late-summer shortages. Though the Columbia system is only moderate-ly sensitive to climate change, allocation conflicts and a cumber-some network of interlocking authorities restrict its ability to adapt,producing substantial vulnerability to these shortages. Source:Hamlet, A.F. and D.P. Lettenmaier. 1999.

Projected Streamflow Effects from Climate Changein the Pacific Northwest

An important area of vulnerability as a result of cli-mate change is associated with summer water sup-ply from Pacific Northwest (PNW) snow melt andtransient snow river basins that are at moderate ele-vations. Under current climate conditions,summerstreamflows in these moderate elevation basins arestrongly affected by snow accumulation,whichfunctions as winter storage. As the temperaturerises,snow lines move up in elevation (see Figure10) and overall snow extent is reduced (see Figure11). As a result,maximum streamflows tend to beshifted towards the winter, with correspondingreductions in summer streamflow volumes (seeFigure 12). High-elevation basins,which are belowfreezing for much of the winter season,are lessaffected by the changes in temperature,and the tim-ing changes are less pronounced. For regional scale

watersheds like the Columbia River Basin,whichintegrate both of these responses,the effects areintermediate. The lower summer flows that resultfrom the shift in the hydrograph are likely to exacerbate existing conflicts between summerwater supply for human use (e.g.,irrigation east ofthe Cascades and municipal use west of theCascades),and maintenance of summer instreamflow for ecological purposes (such as protectingsalmon habitat). In the Pacific Northwest RegionalWorkshop it was concluded based on the 1995 MaxPlanck Institute climate model scenario that themost significant vulnerability to climate change isthe potential for declining summer water supply inthe context of rising demand (Pacific NorthwestRegional Report,1999).

Since spring runoff events are likely to be earlier,reservoir management will need to become moresophisticated in managed watersheds. For exam-ple,optimized dynamic reservoir operation ruleswill likely become more appropriate than tradition-al rule curves. Relying more on medium and long-term predictions of weather will likely maximizesupply and minimize risk of flooding (Georgakakoset al.,1999).

Depending on the degree to which river systemsare managed, water supply effects can be damp-ened by storage and release regimes. However, astudy of potential impacts on the Colorado Riverunder current “Law of the River”operating proce-dures indicates that even small decreases in averagerunoff could lead to a dramatic decrease in powergeneration and reservoir levels (Gleick and Nash,1993) as the system tries to maintain committeddeliveries of water. Many storage systems,like theColorado,can readily handle year-to-year variabilitybut may have more difficulty with long-termchange.

Water demandWater withdrawals increased faster than populationgrowth for most of this century and reached 341billion gallons per day in 1995. However, since1975 water use has been decreasing on a per capitabasis,and total withdrawals have declined 9% sincetheir peak in 1980 (Solley et al.,1998,see Figure13). Per capita consumptive use is expected tocontinue to decline in some areas,due primarily toreductions in irrigated acreage,improvements inwater use efficiency, recycling and reuse,and use ofnew technologies. Brown (1999) developed water-use forecasts to the year 2040 under several scenar-ios. Total withdrawals would increase only 7% by2040 with a 41% increase in population under the


418

Figure 13: Although US population has continued to increase,withdrawals have declined on a per capita basis. Reductions aredue to increased efficiency and recycling in some sectors, and areduction in acreage of irrigated agriculture. Source: Solley, W.B.,R.R. Pierce, and H. A.Perlman, 1998.

Water Withdrawals and Population Trends

Consumptive Water Use by Sector

Figure 14: Agricultural water use is the highest consumptive usesector. Source: Data from Solley, W.B., R.R. Pierce, and H.A.Perlman, 1998. See Color Plate Appendix

cooling at electric generating plants may decreasebecause of increased pressure to divert more waterfor other uses. Climate change could possibly affectnavigation by changing water levels in rivers, reser-voirs,and lakes (e.g.,Great Lakes,Mississippi River,and Missouri River;see Midwest chapter),as well asby changing the frequency of floods and droughts.

In the Pacific Northwest, hydropower and endan-gered species preservation are in increasing conflictbecause minimum streamflows must be maintainedfor habitat or water must be diverted away fromturbines to protect migrating fish. Changes in sea-sonal runoff, even with no change in precipitation,could represent a very serious additional complica-tion. Outflows from some power plants containwaste heat,which affects water temperature in thearea of discharge. Although these discharges areregulated, changes in demand for electric power arelikely to affect aquatic habitat. Precipitationchanges in specific regions,such as the PacificNorthwest and portions of the eastern US,willaffect hydropower capacity in the future.

Water issues for Native American communities areparticularly critical,in part because of geographicand legal limitations and competition for resources.Significant concerns exist related to fisheries andaquatic habitat,especially with regard to Native sub-sistence economies. These concerns are particular-ly important in the Pacific Northwest. BecauseIndian reservations are found throughout the US,water issues vary substantially by geographicregion. In most cases,the tribal culture is tied to aspecific place and traditional survival strategies.

middle population projection. However, even withreduced per capita use,urban demand is increasingin major metropolitan areas along the coasts and inthe Southwest,due to population increases.Agricultural irrigators will likely continue to havecompetition from municipal users for available sup-plies. Under drought conditions,competition forwater between the agricultural and urban users islikely to intensify (see Figure 14).

Increased water use efficiency is believed to be akey solution to the increasing stress on water sup-plies. It is widely thought that potential exists toreduce total demand for water without affectingservices or quality of life. However, as more andmore waste is taken out of the system,futuredemand is less easily reduced in response to droughtor short-term delivery problems. This “hardening ofdemand”is widely recognized by water managers.Conservation investments generally need to berenewed over time,since the effectiveness of manyprograms declines over time (including the impactsof conservation pricing and the effectiveness of lowwater use plumbing devices).

Changes in average temperature,precipitation andsoil moisture caused by climate changes are likely toaffect demand in most sectors,especially in the agri-culture, forestry, and municipal sectors. Increasedtemperatures and decreased soil moisture are verylikely to increase irrigation water needs for somecrops. There is a clear linkage between weather pat-terns and water demand in these sectors (see Figure15).

In 1995,irrigation accounted for 81% of total con-sumptive freshwater use and 39% of total waterwithdrawals in the US (Solley et al.,1998). Total useof water in agriculture has been declining since1980,with the exception of the Southeast,where a39% increase in irrigated acreage of row crops wasidentified between 1970 and 1990 (IrrigationJournal,1996). McCabe and Wolock (1992) used anirrigation model to demonstrate that increases inmean annual water use in agriculture are more likelyto result from increases in temperature than fromdecreases in precipitation. This finding may beimportant because runoff is also affected byincreased temperatures.

Hydropower and navigation are not consumptiveuses,but they are affected by both the volume andthe timing of streamflows. Demand for electricity isvery likely to increase with higher temperatures dueto corresponding demands for summer air condi-tioning,but the water available for hydropower and


419

Municipal Water Use(gal/person/day)

Agricultural Water Use (AF)

Total Municipal Water Use(AF)

Reference ETminus Rainfall (n)

Figure 15: Water demand in the agricultural and municipal wateruse sectors correlates strongly with evapotranspiration rates.Source: Arizona Deptartment of Water Resources. See Color PlateAppendix

Evapotranspiration and Water Use in Tucson

The viability of the Hopi reservation, for example,islinked to the availability of water on their reserva-tion. This results in added vulnerability to climatechange impacts.

Many American Indian communities possess senior,but unexercised water rights. As these rights areput to use,new stresses are very likely to be intro-duced in the affected watersheds. Quantifying andlitigating the water rights claims of Indian communi-ties is a major ongoing issue in many western states.

2. Surface Water Quality

Major improvements have been made in the qualityof surface water in the US,largely attributable to thesuccess of the Clean Water Act in reducing industrialpollution and discharge of sewage. In 1994,83% ofthe rivers and 87% of the lakes were consideredsuitable sources for drinking water supply (all sur-

face water must be treated before use),while 95% ofthe rivers and 82% of the lakes were suitable for fishhabitat (US Dept of the Interior, 1997). Remainingwater quality problems were attributed primarily tonon-point sources of pollution,such as nutrients,bacteria,and siltation deriving from agriculture,andurban runoff (US Dept.of the Interior, 1997).

Water quality issues associated with potential climate change impacts are more subtle than supplyissues and include potential impacts on humanhealth and ecosystem function, changes in salinityassociated with changes in stream flow, and changesin sediment regimes. Non-point source pollutionand agricultural byproducts are likely to becomemore problematic depending on the change in pre-cipitation patterns;an increase in extreme precipita-tion events is considered likely (IPCC,1996),whichmay increase risk of contamination. A balancingeffect is that with some exceptions,higher precipita-tion will result in lower concentrations of organicand inorganic constituents in surface water, due todilution. Water quality is greatly influenced by flowvariability, and some significant water quality prob-lems are episodic,e.g.,episodic acidification fromsnowmelt,and algal blooms due to nutrient increases (Mulholland and Sale,1998;Meyer et al.,1999). Increasing salinity related to ir rigation returnflows (water returning to streams and aquifers afteruse by agriculture) and greater diversions of surfacewater are ongoing issues,especially in the West.Changes in streamflow associated with increasedprecipitation are likely to reduce salinity levels,espe-cially in winter, while lower flows and higher tem-peratures could exacerbate this problem in summer.Flooding associated with more intense precipitationcan also affect water quality by overloading stormand wastewater systems,and damage sewage treat-ment facilities,mine tailing impoundments,or land-fills,thereby increasing risk of contamination.

Many of Santa Barbara’s beaches were closed in1998 due to high bacterial counts from the intenseEl Niño storm runoff. More winter runoff is likely tobring larger sediment flows to coastal waters,whilelower summer streamflow is likely to increase salini-ty and impact estuarine species (see Figure 16).

A combination of increased precipitation andwarmer, drier summers could increase fire hazard insome ecosystems. Sedimentation,landslides,andmudslides frequently follow removal of vegetationby fire (see Figure 17).

Drinking water supplies are very likely to be directlyaffected by sea-level rise in coastal areas,both


420

Figure 16: Sediment flow off Santa Barbara caused by El Niñostorm runoff. Source: Mertes, L., The Plumes and Blooms Project,ICESS/UCSB. See Color Plate Appendix

Figure 17: Coastal mudslide on Highway 50, California followingvegetation removal and heavy rainfall. Source: Eplett, R.A.,California Governor’s Office of Emergency Services.

of groundwater to water supply in many regions,effective institutions to manage groundwater are theexception, rather than the rule (Dellapenna,1999;National Research Council,1997).

Groundwater is managed through a different mechanism in virtually every state,though there are

through saltwater intrusion into groundwateraquifers and movement of the freshwater/saltwaterinterface further upstream in river basins. In thecase of New York City, if the salt front moves furtherup the Hudson River, it will threaten emergencywater supply intakes (Northeast Regional Report,inpreparation). Periodic storm surges can also affectwater quality and these are likely to be exacerbatedby rising sea levels in a warming climate.

It is likely that climate change will affect lake, reser-voir, and stream temperature through direct energytransfer from the atmosphere and changes in damoperations. Increased temperatures in surface waterare likely to eliminate some species (such as salmonand trout) that are already near their habitat temper-ature threshold (see Figure 18). Higher tempera-tures result in reduced dissolved oxygen in water,which is a measure of ecosystem condition. Hurd etal.(1999) used dissolved oxygen stressed water-sheds as an indicator of ecosystem vulnerability, andfound that the most vulnerable regions are inWisconsin,northern Illinois,southern Appalachia,South Carolina,and large portions of east Texas,Arkansas,Louisiana,and Florida (see Figure 19).Changes in temperature regimes are also likely toaffect ice cover, and mixing and stratification ofwater in lakes and reservoirs,conditions that are keyto nutrient balance and habitat value (Meyer et al.,1999).

3. Groundwater Quantity andQuality

Groundwater is the source of about 37% of irrigationwater withdrawals (Solley et al.,1998),and suppliesdrinking water to about 130 million Americans(USGS,1998). Though groundwater supplies are lesssusceptible to variations in climate than surfacewater, they may be more af fected by long-termtrends. More frequent or prolonged droughts arelikely to increase pressure on groundwater supplies,which commonly serve as a buf fer during shortagesof surface water supplies. Depletion of groundwateris significant on the High Plains,the Southwest,partsof the Southeast,and in the Chicago area (USGS,1998). Groundwater overdraft can cause substantiallong-term effects,because in some areas the availablegroundwater supply is essentially nonrenewable orbecause land compaction prevents groundwaterrecharge (see Figure 20). Where the rate of rechargeof groundwater aquifers is slower than use,long-termgroundwater pumping becomes unsustainable.However, increases in precipitation are likely over asignificant portion of the US,and many groundwateraquifers are likely to benefit. Despite the importance


421

Summer Stream TemperaturesSteamboat Creek, Oregon

Figure 18: Simulated summer stream temperatures under presentday climate (blue) and simulated temperatures under about a twicecurrent CO2 climate (red). The dashed line at 24 ºC (75 ºF) on the"water temperature" axis indicates the summer temperature toler-ance of juvenile steelhead trout. Under doubled CO2, the modelsuggests that the length of time within the year when the tempera-ture tolerance limit is exceeded is more than twice as long as undersimulated present-day climate conditions. Shaded area surround-ing the doubled CO2 temperature curve indicates an estimate ofuncertainty. Source: US Geological Survey Circular 1153, Robert S.Thompson, et al. See Color Plate Appendix

Hydrologic Unit Boundary

State Boundary

Low Vulnerability <1.5

Medium Vulnerability ≥1.5

High Vulnerability >2.2

RelativeVulnerability

Figure 19: Instream Use, Water Quality, and Ecosystem SupportAssessed Vulnerability based on current climate and waterresource conditions, based on data describing the following: floodrisk population, navigation impacts, ecosystem tolerance to coldand heat, dissolved oxygen stress, low streamflow conditions, andnumber of aquatic species at risk. Source: Hurd, B.J., N. Leary, R.Jones and J. Smith, 1999a. See Color Plate Appendix

Current Climate Vulnerability Map, Instream Use,Water Quality and Ecosystem Suppor t


422

three basic systems of groundwater rights (see box,“Major Doctrines for Surface Water andGroundwater”).

Groundwater maintains the base flow for manystreams and rivers,and lowering groundwater levelsmay reduce the seasonal flows and alter the temper-ature regimes that are required to support criticalhabitat,especially wetlands. Although conjunctivemanagement (using groundwater and surface waterin combination to meet demand) is frequently citedas a solution to water supply problems,thisapproach is only sustainable if the groundwater sup-plies are periodically recharged using surplus surface water or other alternative supplies (seeFigure 21). In order to ensure a sustainable supply,aquifers may need to be artificially recharged,whichinvolves consideration of multiple issues includingchanges in water quality in the aquifer. In some

cases,such as portions of the Ogallala Aquifer in theGreat Plains, groundwater supplies have alreadybeen over utilized and no source of renewable sup-plies is available.

Groundwater in storage is affected by seasonality,volume,and persistence of surface water inflows,and discharges from groundwater to surface water.Groundwater/surface water interactions are poorlyunderstood in most areas. Changes in precipitationand temperature may have long-term ef fects onaquifers that are relatively subtle and difficult toidentify. Groundwater is frequently found in hor i-zontal layers within an aquifer, separated by relative-ly impermeable layers of silt and clay or rock. Thewater quality is af fected by the substrate that thewater flows in,and the amount of time it has beenin storage. In some areas,only the surface zone iscontaminated by human activities. An understand-ing of the aquifer’s geology, including such charac-teristics as location of impermeable layers and thedirection of water flow, is necessary in designingappropriate management options.

Industrial pollution is the largest groundwater quali-ty issue in urban areas. Common problems includesolvents and petrochemicals. Recently released datafrom the USGS National Water Quality Assessmentindicate that volatile organic compounds (VOCs)were detected in 47% of urban wells tested between1985 and 1995. The most common VOCs foundwere the fuel additive MTBE (methyl tertiary butylether) and various solvents such as tetra-chloroethene,trichloroethene and trichloromethane(Squillace et al.,1999). Contamination of drinkingwater supplies presents serious challenges to watermanagers,especially in large urban areas. Cleanupof contaminated aquifers is extremely expensive andin some cases is not practical.

Agricultural chemicals and wastewater treatmentbyproducts such as nitrates also affect groundwaterquality in some areas.Continued degradation isanticipated in some metropolitan areas. It is unclearwhether climate change will have a significanteffect on groundwater contamination.

Increased pressure on groundwater supplies hasresulted from the Safe Drinking Water Act (SDWA)regulations. The Surface Water Treatment Rule nowrequires filtration of all surface sources. As a result,many small surface water systems are now uneco-nomical,and have either combined to form largersystems or switched to groundwater, for which fil-tration is not required. O’Connor et al.(1999) sur-

Figure 20: Land subsidence fissure,, caused by over-pumping ofgroundwater, can result in earth fissures such as this near Eloy,Arizona. Source: K. Jacobs.

Figure 21: Artificial recharge in Santa Ana riverbed. Artificialgroundwater recharge in the Santa Ana Riverbed, Orange CountyWater District, California. Source: Orange County Water District.


423

veyed 506 community water system managers inthe Pennsylvania portion of the Susquehanna RiverBasin,and found that half of all the communitywater systems in Centre County switched to ground-water or regionalized their surface water systems inresponse to SDWA regulations.

Another key groundwater quality concern is saltwa-ter intrusion in coastal aquifers. As pumping ofgroundwater increases to serve municipal demandalong the coast,freshwater recharge in coastal areasis reduced,and sea level rises, groundwater aquifersare increasingly affected by infiltration of seawater.Seawater intrusion is already a major issue inFlorida,the Gulf Coast,southern California,LongIsland,Cape Cod,and island communities. The global sea level is estimated to have risen 4 to 8inches (10-20 cm) over the past 100 years (Gornitz,1995);in some areas,the relative sea-level rise hasbeen greater because land surface elevations aresinking in some regions of the coast. Furtherincreases in sea level are very likely to accelerateintrusion of salinity into aquifers and affect coastalecosystems. The adaptation strategies for dealingwith this problem — importation of alternativesources of supply, desalination,and artificialrecharge — can be extremely expensive.

Because surface water and groundwater supplies areinterconnected and transportation of water acrosshydrologic and political boundaries is common,issues related to surface water/groundwater interac-tions will possibly be exacerbated by climatechange. Even in the context of higher average pre-cipitation,increased temperatures and changes inseasonality of runoff are likely to reduce stream-flows during the warm season,a key period forecosystem maintenance. Increased urbanization,which generally results in increased channelizationof streambeds and higher runoff rates is likely toreduce opportunities for recharge of groundwaternear areas of high groundwater demand. Watertransfers may increase pressures on areas of originfrom the perspective of water supply and ecosystemhealth. Particularly in arid and semi-arid regions,effects on surface water or groundwater resourcesresulting from climate change are likely to impactriparian systems that support a high percentage ofbiological diversity. In many cases,higher precipita-tion is likely to have a positive effect on groundwa-ter levels and riparian habitat.

4. Heavy Precipitation, Floods andDroughts

Floods,especially those related to flash floods fromintense short-duration heavy rains,are likely toincrease in magnitude or frequency in manyregions. Changes in seasonality of flood flows arevery likely to occur in those areas affected by ahigher proportion of rain to snow (yielding earlierpeak flows of shorter duration). Intensity ofdroughts is also likely to increase in some areas dueto higher air temperatures causing greater evapora-tion and water use by plants.

There are two types of socioeconomic costs relatedto floods and droughts:the costs of building andmanaging the infrastructure to avoid damages,andthe costs associated with damages that are not

Cumulative Number of Large Dams Built in the US

Average Volume of US Reservoirs Built

Figure 22: The number of large dams built in the US has declinedin recent decades, data yearly. Source: US Army Corps ofEngineers. 1996. National Inventory of Dams.

Figure 23: The average volume of reservoirs built in each five-yearperiod since 1960 has declined, data five year interval. Source: USArmy Corps of Engineers. 1996. National Inventory of Dams.


Impacts of Potential Climate Change on Aquatic Ecosystem Functioning and Health

(adapted from Meyer et al.,1999)

Region Potential Climate Ecosystem Considerations Effect

Great Lakes/ Warmer, more Altered mixing regimes in lakes (e.g.,longer summer stratification.Precambrian precipitation,but drier Changes in DOC concentration, changes in thermocline depth and Shield soils possible,depending productivity.

on the magnitude of Decreased habitat for cold and cool water fishes,increased habitatprecipitation increase. for warm water species.

Alteration of water supply to wetlands will affect composition of plant communities and carbon storage as peat.

Arctic and Much warmer, Loss or reduction of deltaic lakes.sub-Arctic increases in Reduction in area covered by permafrost,leading to drainage of North America precipitation lakes and wetlands,land slumping,sedimentation of rivers.

Increased primary productivity,but perhaps not enough to com-pensate for increased metabolic demands in predatory fish.

Shift in carbon balance of peatlands.

Rocky Warmer Changes in timberline would affect stream food webs.Mountains Increased fragmentation of cold-water fish habitat.

Fishless alpine lakes sensitive to changes in nutrient loading and sedimentation.

Current anthropogenic changes are threatening aquatic ecosystems.

Pacific Coast Warmer, less snow but Increases in productivity in alpine lakes.Mountains and more winter rain,less Increased meromixis and decreased productivity in saline lakes.western Great summer soil moisture Altered runoff regimes and increased sediment loads leading to Basin decreases in channel stability and negative impact on eco-

nomically important fish species.

Basin and More precipitation, Aquatic ecosystems highly sensitive to changes in quantity and Range,Arid warmer, overall timing of stream flow.Southwest wetter conditions Intense competition for water with rapidly expanding human

populations.

Great Plains Warmer with less Historical pattern of extensive droughts.soil moisture Reduced water level and extent of open water in prairie pothole

lakes with negative ef fects on waterfowl.Increasing warming and salinity in northern and western surface

waters threatening endemic species.Reduction in channel area in ephemeral streams.

Mid-Atlantic Warmer and and New Potentially less episodic acidification during snowmelt.England somewhat drier Possible increase in bioaccumulation of contaminants.

Bog ecosystems appear particularly vulnerable.Current context:stresses from dense human populations and a

long history of land use alterations.

Southeast Warmer with possible Increases in rates of primary productivity and nutrient cycling in precipitation increases lakes and streams.and greater clustering More extensive summer deoxygenation in rivers and reservoirs.of storms Loss of habitat for cold-water species like brook trout,which are

at their southern limit.Drying of wetland soils.Northward expansion of nuisance tropical exotic species.Increased construction of water supply reservoirs.424

avoided,including ecosystemimpacts. About $100 billion hasbeen spent by the federal gov-ernment since 1936 in the USfor the construction,operationand maintenance of flood con-trol features, yet damages asso-ciated with floods continue torise (Frederick and Schwarz,1999b). Flood damage esti-mates by state are provided bythe National Center forAtmospheric Research and theUS Army Corps of Engineers. The 1993 flood in theMississippi and Missouri Rivers caused record dam-ages of over $23 billion. These are only the of ficialdamage estimates,and do not take into account totalsocial costs. The 1999 North Carolina flood, result-ing from Hurricane Floyd,offers a recent example ofthe massive dislocations and multi-billion dollarcosts that often accompany such events. Dams andlevees have also saved billions of dollars of invest-ment,but these facilities,together with insuranceprograms,encourage development in floodplains,thereby indirectly contributing to damages(Frederick and Schwarz,1999b). In addition,struc-tural flood control features have high environmentalcosts. Climate change may affect flood frequencyand amplitude,with numerous implications formaintenance and construction of infrastructure andfor emergency management. Erosion and depositionrates in rivers and streams are likely to changeunder different precipitation regimes. The reductionin reservoir construction along with the buildup ofsediment in reservoirs will affect the resilience ofwater supply systems and their ability to handleflood flows (Frederick and Schwarz,1999b) (seeFigures 22 and 23).

Flood risks are ultimately a function of many factors,including populations exposed to floods,the natureand extent of structures within river floodplains andin coastal areas subject to storm surges,the frequen-cy and intensity of hydrologic events,and kinds ofprotection and warning available. To the extent thateach of these factors can be addressed economicallyand in a timely and integrated manner, future dam-ages can be limited. Wetlands restoration in man-aged watersheds can reduce the impact of stormwater runoff to waterways by slowing down orabsorbing excess water. Providing wetland protec-tion including buffer areas beyond the wetlandboundary is a viable method of avoiding flood dam-age or the cost of flood protection.

There are many different kinds of droughts,fromshort-term localized reductions in water availabilityto long-term and widespread shortages. Severedroughts have had widespread and devastatingeffects,particularly on agriculture. The drought ofthe 1930s affected 70% of the US and caused sub-stantial economic dislocations (Woodhouse andOverpeck,1999). Prolonged droughts affect all sec-tors of the economy, and may be especially devastat-ing for ecosystems. An evaluation of the paleoclimat-ic record indicates that droughts of Dust Bowl severi-ty are not unprecedented,at least at a regional level.Agricultural interests in the Great Plains region areparticularly concerned about the increased likeli-hood of drought with global warming (see Figure24). In the Canadian model,severe and extreme


425

Palmer Drought Severity Index ChangeCanadian Model 21st Centur yHadley Model 21st Century

>10

8

6

4

2

0

-2

-4

-6

-8

-10

Figure 24: The Palmer Drought Severity Index (PDSI) is a common-ly used measure of drought severity taking into account differencesin temperature, precipitation, and capacity of soils to hold water.These maps show projected changes in the PDSI over the 21st cen-tury, based on the Canadian and Hadley climate scenarios. A PDSIof -4 indicates extreme drought conditions. The most intensedroughts are in the -6 to -10 range, similar to the major drought ofthe 1930s. By the end of the century, the Canadian scenario proj-ects that extreme drought will be a common occurrence over muchof the nation, while the Hadley model projects much more moderateconditions. Source: Felzer, B. UCAR. — See Color Plate Appendix

Palmer Drought Severity Index Change

Figure 25: Prairie potholes are considered to be especially vulnera-ble to drying conditions in the northern Great Plains.

species will probably be vulnerable because theyhave little potential for migration (Kusler andBurkett,1999). Minor changes in maximum andminimum temperatures and seasonality of precipita-tion can have significant impacts on wetland habitat(Kusler and Burkett,1999). Wetlands that are direct-ly dependent on precipitation are likely to be morevulnerable to climate change than those that aredependent on groundwater outflows due to the sig-nificant buffering capacity of regional aquifer sys-tems (Winter, 1999). Riparian habitats are of greatconcern in part because 55% of threatened andendangered species are dependent on them(Herrmann et al.,1999).

Species live in the larger context of ecosystems andhave differing environmental needs. A change thatis devastating to one species may encourage theexpansion of another to fill that niche in the system.It is not possible to determine a single optimumenvironmental condition for all species in theecosystem (Meyer et al.,1999). Extreme conditionssuch as floods,droughts,and fire are critical to sus-taining certain ecosystems. Hydrologic conditionsaffect nutrient cycling and availability in streamsand lakes,which affects productivity. Ecologicalresponses to changes in flow regime will depend onthe regime to which it is adapted;a system that ishistorically variable can be severely disrupted by sta-bilizing the hydrologic regime (which happenswhen dams are used to regulate flow, as on theColorado River).

ADAPTATION STRATEGIES

Water management has become more complex asvalues related to water supply and demand haveshifted and regulations have proliferated,both pre-scribing and proscribing many solutions. Most ofthe options available for responding to the impactsof climate change and variability on water resourcesare alternatives that are already used to respond toexisting challenges. However, it should be notedthat optimizing water resource management underincreasing constraints (including regulatory con-straints) narrows the options available. In addition,responding to climate changes may require a broad-er set of information than is usually available todecision-makers.

Current water management practices and infrastruc-ture throughout the country are designed to addressproblems caused by existing climatic variability. Ingeneral,engineering approaches to system designrely on historic data,assuming that future climatic

drought becomes the norm for the Great Plains,butin the Hadley model such drying is not evident.Better information is needed on changes in droughtrisks from climate changes.

Droughts also affect the ability of waterways to sup-port transportation,particularly on the Great Lakesand major river systems like the Mississippi andMissouri. Climate change is likely to affect the vol-ume and timing of streamflows,the amount of sedi-ment carried and deposited in shipping channels,and the extent of ice blockage in the northernwaterways (Hurd,et al.,1999).

5. Ecosystem Vulnerabilities

Climate changes are very likely to have a wide vari-ety of effects on ecosystems. Other human-inducedchanges (such as impacts of changing land use onwater quantity and quality, sediment load,and com-petition from exotic species) are expected to be ofgreater magnitude in most parts of the country thanclimate change. However, climate change may addanother layer of stress to natural systems that havelost much of their resiliency. From an ecologicalperspective,the Arctic,Great Lakes,and Great Plains(especially Prairie Potholes,see Figure 25) regionsappear most vulnerable (see summary table adaptedfrom Meyer et al.,1999). Aquatic and riparianecosystems in the arid Southwest are also vulnera-ble to changing precipitation and runoff regimes,but the nature of predicted climate change in thatregion may alleviate existing stresses (Meyer et al.,1999),see table on page 424.

Evidence of the current warming trend can befound in Alaska,where the area of sea ice is shrink-ing,glaciers are melting,and land that has been sup-ported by permafrost for centuries is transitioningto a new ecological regime. Although the impact onriver flows in Alaska has not been sufficiently stud-ied,it is clear that major changes are occurringwhich have already affected species compositionand subsistence hunting and fishing. Changes in cli-matic conditions in Alaska are very likely to be evenmore dramatic in the future,as albedo (reflectivity)is reduced by reduced ice and snow cover and evap-oration rates increase in the summer (Felzer andHeard,1999).

Impacts on lakes and wetlands from climate changeare likely to include changes in water temperature,sedimentation and flushing rates,length of icecover, amount of mixing of stratified layers,and theinflow of nutrients and other chemicals. Montaneand alpine wetlands with temperature-sensitive


426

Key Climate Messages for Water Managers

• Climate is not static and assumptions made about the future based on the climate of the recent past may beinaccurate. Water managers should factor in the potential for climate change when designing major new infra-structure. Assumptions about the probability, frequency, and magnitude of extreme events should be carefully re-evaluated.

• There is substantial stress on the water sector even in the absence of climate change. There are numerouswatersheds that are already over-appropriated,and new stresses are coming from population dynamics,land usechanges,and changes in international economies. In some areas,the new demands associated with instreamflow needs for habitat protection and Indian water rights settlements may cause major shifts in water supplyand water rights. Climate change may pose additional stresses and could result in thresholds being reached earli-er than currently anticipated.

• Waiting for relative certainty about the nature of climate change before taking steps to reduce risks in water sup-ply management may prove far more costly than taking proactive steps now. (The suggested risk-reducing or “noregrets”steps are those that would have other beneficial effects and so are appropriate regardless of climatechange.)

• The types of changes encountered in the future may not be gradual in nature. Non-linearities and surprisesshould be expected, even if they cannot be predicted.

• The problems that are likely to result from climate change are intergenerational. Decisions made today will com-mit future generations to certain outcomes. It is important to evaluate benefits of projects over long timeframes,and develop an educated citizenry.

Other Key Considerations for Water Managers

• The water delivery, wastewater, and flood control infrastructure,particularly in the eastern US is aging andsometimes inadequately maintained and therefore vulnerable. The likely additional stresses that may result fromclimate change should encourage upgrading of key infrastructure to limit vulnerability to extreme events.

• As has been observed by many, the days of building large dams and expensive supply-side solutions are nearlyover. More innovative solutions will be required in the future. Managers will need to prepare contingency plansto face water quality and supply challenges regardless of changes in climate. Promising options include conser-vation and efficiency improvements, water banking, water transfers,conjunctive use of surface and groundwatersystems,and cooperative arrangements with other jurisdictions and communities.

• There are currently multiple disincentives to efficient utilization of water supplies. Subsidies and failure toreflect the full value of water supplies affect water pricing in virtually every sector. Americans view water as a“public good,” believing supplies should be cheap,plentiful,and contain virtually no health risk factors.Agricultural water use is generally the most highly subsidized,but there are few municipal water suppliers thatassign a value based on the replacement cost. As stresses increase on the water sector, water costs will definitelyincrease. Equity issues should be fully evaluated.

• Policies related to floodplain management and insurance currently encourage risky behavior such as rebuildingin floodplains and low-lying coastal areas after floods and storm surges. At both a national and local level,encouraging people to move away from high-risk areas would be beneficial. Incorporating wetland protectionin buffer areas beyond current wetland boundaries would of fer additional resilience to cope with potentialflooding from more intense storms.

• Catastrophic events such as floods and fires are required to sustain some ecosystems over the long term.Management of these ecosystems should allow for continued benefits from these events.

• Hydrologists have developed valuable new models of watershed and regional-level hydrology that are ready foruse. Effective use of mid- and long-range forecasts can improve the management of water resources and wouldbe a significant step in developing the flexibility and resilience needed to cope with climate change.

• A key component of either conservation programs or improved water rights administration is metering or meas-uring of individual uses. This is a basic step to understanding water use and in educating consumers about prop-er water management. Many large cities in the US (including Fresno and Sacramento,California) currently donot measure water deliveries;most agricultural water use,especially groundwater use,is also not properly moni-tored.

• Improved management opportunities are available when watersheds are managed as a hydrologic unit.• A key question that should be considered by policy makers is how much risk is acceptable to the public. It is

not reasonable to manage for or expect no risk or zero damages from natural hazards.


427

Works Association,1997). The goal was to identifyopportunities for reducing the risks associated withfuture climatic changes.

Various potential water management adaptationshave been suggested to respond to either existingstresses or new stresses associated with climatechange. They include the following:

• Increase ability to shift water within andbetween sectors (including agriculture to urban);this could increase flexibility but may requirechanges to institutional structures.

• Use pricing and market mechanisms proactivelyto increase efficiency of water use.

• Incorporate potential changes in demand andsupply in long term planning and infrastructuredesign.

• Create incentives or requirements to move peo-ple and structures out of floodplains.

• Identify ways to manage all available supplies,including groundwater, surface water and efflu-ent,in a sustainable manner.

• Restore and maintain watersheds as an integratedstrategy for managing both water quality andwater quantity. For example, restoring water-sheds that have been damaged by urbanization,forestry, or grazing can reduce sediment loads,limit flooding, reduce water temperature,andreduce nutrient loads in runoff.

• Reuse municipal wastewater, improve manage-ment of urban stormwater runoff, and promotecollection of rainwater for local use to enhanceurban water supplies.

• Increase the use of forecasting tools for watermanagement. Some weather patterns,such as ElNiños,can now be predicted with some accura-cy and can help reduce damages associated withextreme weather events.

• Enhance monitoring ef forts to improve data forweather, climate,and hydrologic modeling to aidunderstanding of water-related impacts and man-agement options.

Communication Strategies Information on the impacts of climate change isonly helpful if it is usable by water managers,landowners,emergency response teams,and otherdecision-makers. They need to understand the rangeand probability of potential outcomes. This willrequire timely, detailed information at the scaleneeded to address local conditions. In addition,since most adaptation strategies incur costs,whether they are in response to existing or newstresses,it will be important to communicate therisks,costs,and opportunities. The need for better

conditions won’t deviate significantly from thoseexperienced in the recent past. Adaptation strate-gies for dealing with climate change range from rela-tively inexpensive options such as revising operat-ing criteria for existing systems,to re-evaluatingbasic engineering assumptions in facility construc-tion,to building new infrastructure with substantialcapital costs. Strategies also include water conserva-tion technologies and policies,use of reclaimedwastewater and other alternative supplies,andimproved mechanisms for water transfers.Insufficient work has been done to evaluate thecosts and benefits of alternative adaptation strate-gies. However, improved management of existingsystems would certainly be valuable in managingchanges in the ranges projected in this Assessment.

Improved efficiency of water use is likely to resultfrom both regulatory requirements and higher watercosts,which are expected outcomes of growingdemands for clean and adequate water supplies. Amove towards marginal-cost pricing (where costsreflect the price of the next available supply) andmore extensive water markets may develop.

Water managers are currently not adequatelyengaged in the process of evaluating the risks of cli-mate change. There is much debate about whetherthis is because they understand the nature of therisks,but have concluded that the current tools theyhave are sufficient,or whether they simply do nothave the information they need in order to respondmore appropriately (Stakhiv and Schilling,1998;Frederick and Gleick,1999). Results of a recent sur-vey of western water managers (Baldwin et al.,1999) indicate that water managers who routinelydeal with variability that is an order of magnitudegreater than predicted climate changes see little rea-son to respond. Perhaps this is because they do notunderstand that potential climatic changes may beimposed on top of existing variability. However,availability of information and a greater understand-ing of the issues are likely to affect managementpractices. For example,if managers knew that thereis a high probability that the magnitude of floodevents is likely to increase in their region even if fre-quency remains the same,this information could beincorporated into planning activities. Even withoutthis kind of information,however, some waterorganizations are beginning to push for common-sense actions by water managers. The AmericanWater Works Association published recommenda-tions for water managers calling for a re-examina-tion of design assumptions,operating rules,and con-tingency planning for a wider range of climatic con-ditions than traditionally used (American Water


428

information argues for improved monitoring and mod-eling to link climate information with hydrologicimpacts.

There is significant concern that results from GCMswill be misunderstood. While they do not predict thefuture,they are useful tools for exploring future scenar-ios. GCM outputs would be more useful to water man-agers and researchers if the models’underlying con-cepts were better known. It has been suggested thatconverting the outputs into information about weathersystems,storm tracks,and likely weather events wouldbe helpful.

Balancing water supply and water quality issues whilemaintaining natural ecosystems and quality of life evenin the absence of any climate change is daunting.Adding the overlay of potential climate change increas-es the difficulty of achieving these goals. One generalconclusion of work done to date is that humans havemany options for adapting water supply and demandsystems to climate change,while unmanaged ecosys-tems may be more vulnerable to imposed changes.Human adaptation,however, is very likely to come atsubstantial economic and social cost. In addition,itshould be understood that some impacts are likely tobe unpredictable or unavoidable because of the verynature of atmospheric and climatic dynamics.

In many parts of the US,the water supply and demandpicture is a complex web of imported and local sup-plies,interconnected physical infrastructure,and over-lapping institutions and jurisdictions. Difficulties indownscaling climate models to a useful level for deci-sion-makers continue to limit the utility of the informa-tion produced to date. Depending on the geography ofparticular regions,local predictions of changes in cli-mate may be nearly impossible in the near term. Forexample,in much of the Southwest,quantitative knowl-edge of the current hydrologic cycle is quite limiteddue to the large temporal and spatial variability in pre-cipitation, runoff, recharge, evaporation,and plantwater use within basins. Much of the uncertainty iscaused by the high degree of diversity in the basin andrange topography. Likewise,when predictions of localwater supply conditions are aggregated into regions,the resulting picture may be meaningless. Althoughgeneralizations are necessary in order to communicatemajor concepts,they may have little value whenapplied to particular circumstances without appropri-ate caveats.

There are,however, some major messages that need toreach water managers. Despite the inability to provideclear, detailed information about many regionalimpacts,large-scale climatic changes are likely to occur.

These changes will very likely affect water supplyand demand in ways that may not be anticipated bycurrent water managers,and these changes may beimposed on top of existing climatic variability andhydrologic risks. While many different alternativesfor coping with impacts on the nation’s waterresources are available, we do not yet understandhow effective these will be,how expensive they willbe,or what surprises and unavoidable impacts mayoccur.

CRUCIAL UNKNOWNS ANDRESEARCH NEEDS

Strategic Monitoring Needs

• Sophisticated analysis of climate change or itsimpacts requires continuous data sets providedthrough environmental monitoring. Monitoringshould be enhanced from a strategic perspectivein order to integrate key unknowns,particularlygroundwater conditions,surface water quality, andbiological factors in key habitats. Existing pro-grams are not adequately integrated,and there arecritical gaps in both space and time. Recentdecreases in funding for stream gages and waterquality sampling activities are especially problem-atic. Monitoring provides important services forsociety, such as improved predictive capability forweather events and reservoir management.

• Additional data on snowpack,depth, extent,snowwater equivalent,etc., would be helpful to scien-tists and water managers whose supplies aredependent on snowmelt.

• Tools are needed to interpret water quality dataand make data readily accessible to decision-mak-ers.

• Engineering/Management Research Needs• Improved understanding of the demand side of

the water resource equation is needed.• More quantitative evaluation of costs and effec-

tiveness of adaptation strategies is needed.• Better analyses are needed of the ability of exist-

ing infrastructure’s capability to adapt. Howmuch flexibility is there in existing systems todeal with variability? Further investigations intothe impact of increased precipitation, flooding,and changes in water levels on the nation’s infra-structure are needed. For instance,these climate-related changes may adversely affect air and watertransportation.

• Design criteria (e.g., for 100-year floods) shouldbe reevaluated to reduce risk to infrastructure inthe context of climate change.


429

rately simulate inter-annual climate variability(persistence of extreme events) and probabilityof extreme events. This could subsequently belinked to economic costs and potential manage-ment decisions,such as land use restrictions.

• There is a need to focus on groundwater implica-tions of climate change. Groundwater rechargerates are controlled by many factors that arepoorly understood. The response of deep andshallow aquifers to historic drought should beevaluated,as well as stream/aquifer interactions,the extent of interactions between aquifers,andimpacts on riparian habitat.

• Water quality changes that result from existingclimatic variability, and the impacts of extremeevents on ecosystems,need further evaluation.

• Research is needed on highlighting thresholds ofchange in natural ecosystems,and key areas ofvulnerability including impacts of flooding anddrought. There are likely to be time lags asecosystems respond to change,but these havenot been identified or modeled so the indicatorsare not well understood.

• Biotic responses are not being accounted for ade-quately in modeling efforts,particularly feed-backs associated with changes in land cover,stomatal resistance due to increased CO2, etc.

• Better integration is needed of human and eco-logical risk assessment relative to assessments ofclimate change. Risk factors and willingness topay for damages caused by climate changeshould be evaluated as inputs to decision mak-ing.

• There is a need to improve communicationbetween scientists and water managers. For pur-poses of technology transfer, the value and ade-quacy of integrated climate, hydrologic,and man-agement systems should be demonstrated in pro-totype applications. Demonstration projectscould engage managers of surface water supplysystems in applications of reservoir managementfor their own systems. This could prove helpfulin building relationships between modelers andreal-world managers.

• More flexible institutional and legal arrange-ments should be instituted that facilitate the abil-ity to respond to changing conditions.

• Research is needed to compare and evaluateinnovative floodplain management strategies atthe local,state,and federal levels to improveresiliency to climate change.

Climate Research/Modeling

• Projecting future changes in streamflow condi-tions will require more evaluation of the com-plex role of changing precipitation and tempera-ture patterns as well as the role of land-usechange on streamflow.

• There is a need to continue to refine existingGCMs,and improve model validation and com-parison. Runoff modeling could be improved ifdifferences between the models were betterunderstood. Output should be tailored to usersneeds. Key areas for model development includebetter physically based parameterizations forgroundwater/surface water interactions,atmos-pheric feedbacks,and variability of precipitationand land surface characteristics at a watershedscale.

• Additional research is needed to explore the cur-rent causes of climate variability, such as the ElNiño/Southern Oscillation and Pacific DecadalOscillation. This will enable evaluation ofimpacts if such conditions become more persist-ent in the future.

• Existing models can be used to explore the vul-nerabilities of various regions to changes in cli-mate. These evaluations should lead to improvedunderstanding of critical changes in evapotran-spiration and runoff regimes.

• Increased data and analysis of paleoclimaticrecords will provide substantial insight into thenature and range of climate and hydrologic vari-ability (e.g.,the incidence of droughts andfloods).

Integrated Assessment Research

• Improved tools are needed for translating climatechanges into water resource impacts and issuesof public interest. For example,to be useful inriver basin management,GCMs must more accu-


430

LITERATURE CITEDAmerican Water Works Association (AWWA),Climatechange and water resources, Committee report of theAWWA Public Advisory Forum, Journal of theAmerican Water Works Association, 89(11),107-110,1997.

Anonymous,1995 Irrigation survey reflects steadygrowth, Irrigation Journal, Arlington Heights,IL,1996.

Baldwin,C.K., U. Lall,and F. H.Wagner,Time scales ofclimate variability important to Western water man-agers and their views on climate change,inProceedings of the AWRA Specialty Conference onPotential Consequences of Climate Variability andChange to Water Resources of the United States,American Water Resources Association,Herndon,VA,pp.23-26,1999.

Bernard,S.M., J. Rose,B.Sherman,M.McGeehin, J.Scheraga,and J. Patz,Water, health and climate:Assessing the potential consequences of climatechange and variability on waterborne disease risk,inProceedings of the AWRA Specialty Conference onPotential Consequences of Climate Variability andChange to Water Resources of the United States,American Water Resources Association,Herndon,VA,pp.409-413,1999.

Brown,T. C., Past and future fresh water use in theUnited States, General Technical Report RMRS-GTR-39,US Department of Agriculture, Forest Service,RockyMountain Research Station, Fort Collins,Colorado,1999.

Carnell,R.E.,and C.A.Senior, Changes in mid-latitudevariability due to increasing greenhouse gases and sul-phate aerosols, Climate Dynamics, 4, pp.369-383,1998.

Chalecki,E.L.,and P. H.Gleick,A framework of orderedclimate effects on water resources:A comprehensivebibliography, Journal of the American WaterResources Association, 35(6),pp.1657-1665,1999.URL:www.pacinst.org/CCBib.html

Dellapenna, J.W.,Adapting the law of water manage-ment to global climate change and other hydropoliticalstresses, Journal of the American Water ResourcesAssociation, 35(6),pp.1301-1326,1999a.

Dellapenna, J.W.,Legal responses to global climatechange, Proceedings of the AWRA SpecialtyConference on Potential Consequences of Climate

Variability and Change to Water Resources of theUnited States, American Water Resources Association,Herndon,VA,pp.13-16,1999b.

Eheart, J.W.,A. J.Wildermuth,and E.E.Herricks,Theeffects of climate change and irrigation on criterionlow streamflows used for determining total maximumdaily loads, Journal of the American Water ResourcesAssociation, 35(6) pp.1365-1372,1999.

Felzer, B.,and P. Heard,Precipitation differencesamongst GCMs used for the US National Assessment,Journal of the American Water Resources Association,35(6),pp.1327-1340.

Frederick,K. D.,and P. H.Gleick,Water and global cli -mate change: Potential impacts on US water resources,Pew Center on Global Climate Change, Washington,DC,1999.

Frederick,K. D.,and G. E.Schwarz,Socioeconomicimpacts of climate change on US water supplies,Journal of the American Water Resources Association,35(6),pp.1563-1584,1999a.

Frederick,K. D.,and G. E.Schwarz,Socioeconomicimpacts of climate variability and change on US watersupplies,unpublished manuscript,1999b.

Georgakakos,A. P.,H.Yao,and K. P. Georgakakos,Vulnerability of river basin management to climate vari-ability and change, Proceedings of the AWRA SpecialtyConference on Potential Consequences of ClimateVariability and Change to Water Resources of theUnited States, American Water Resources Association,Herndon,VA,pp.49-56,1999.

Gleick, P. H.et al., Water:The Potential Consequencesof Climate Variability and Change. Report of theNational Water Assessment Group for the U.S.GlobalChange Research Program. [Pacific Institute,USGS,Department of the Interior],Oakland,California.151pp.,2000.

Gleick, P. H.,Vulnerability of water systems,in ClimateChange and the US Water Resources, edited by P. E.Waggoner, pp.223-240, John Wiley and Sons,New York,1990.

Gleick, P. H.,and L.Nash,The societal and environmen-tal costs of the continuing California drought, PacificInstitute for Studies in Development,Environment,andSecurity, Berkeley, California,1993.


431

Herrmann,R.,L.Scherbarth,and R.Stottlemyer, Save thewolves,lose the watersheds:Environmental aspects ofwater use in the United States, Journal of theAmerican Water Resources Association, 35(6),1999.

Hurd,B. J., N. Leary, R. Jones,and J. Smith,Relativeregional vulnerability of water resources to climatechange, Journal of the American Water ResourcesAssociation, 35(6),pp.1399-1410,1999a.

Hurd,B,R. Jones, N. Leary and J. Smith,A regionalassessment and database of water resource vulnerabili-ty to climate change, Proceedings of the AWRASpecialty Conference on Potential Consequences ofClimate Variability and Change to Water Resources ofthe United States, American Water ResourcesAssociation,Herndon,VA,pp.45-48,1999b.

Jehl, D.,Tampa Bay looks to the sea to quench its thirst,The New York Times, March 12,2000.

Intergovernmental Panel on Climate Change (IPCC),Climate Change 1995:Impacts,Adaptations andMitigation of Climate Change:Scientific-TechnicalAnalyses, Cambridge University Press,New York,1996.

Jacobs,K.L.,Water sector summary comments forinclusion in the synthesis report from the AWRASpecialty Conference on Potential Consequences ofClimate Variability and Change to Water Resources ofthe United States,Atlanta,GA,May 10-12,1999.

Karl,T. R.,R.W. Knight, D. R.Easterling,and R. G. Quayle,Indices of climate change for the United States,Bulletin of the American Meteorological Society, 77,pp.279-292,1996.

Karl,T. R.,and R.W. Knight,Secular trends of precipita-tion amount,frequency, and intensity in the UnitedStates, Bulletin of the American MeteorologicalSociety, 79(2),pp.231-241,1998.

Karl,T. R.,and R.W. Riebsame,The impact of decadalfluctuations in mean precipitation and temperature onrunoff: A sensitivity study over the United States,Climatic Change, 15, 423-447,1989.

Kimball,B.A.,Carbon dioxide and agricultural yield: Anassemblage and analysis of 430 prior observations,Agronomy Journal, 75, 779-788,1983.

Kusler, J.,and V. Burkett,Climate Change in WetlandsAreas Part I: Potential Impacts and Interactions,Acclimations,6,1999. URL:www.nacc.usgcrp.gov/newsletter/1999.06/wet.html

Gleick, P. H.,and B.Chalecki,The impacts of climaticchanges for water resources of the Colorado andSacramento-San Joaquin River basins, Journal of theAmerican Water Resources Association, 35(6),1429-1441,1999.

Gomez,S.,and A.Steding,California water transfers: Anevaluation of the economic framework and a spatialanalysis of the potential impacts, Pacific Institute forStudies in Development,Environment,and Security,Oakland,California,57 pp.,1998.

Gornitz,V.,Sea-level rise: A review of recent past andnear-future trends, Earth Surface Processes andLandforms, 20, pp.7-20,1995.

Groisman, P.Ya,et al.,Changes in the probability ofheavy precipitation: Important indicators of climatechange, Climatic Change, 42, 243-283,1999.

Groisman, P.Ya.,Trends in the precipitation and snowcover records across the United States, Proceedings ofthe AWRA Specialty Conference on PotentialConsequences of Climate Variability and Change toWater Resources of the United States, American WaterResources Association,Herndon, VA,pp.89-92,1999.

Groisman, P.Ya.,R.W. Knight,and T. R.Karl,Mean pre-cipitation and high streamflow in the contiguousUnited States: Trends in the twentieth century, Bulletinof the American Meteorology Society, in press,2001.

Hamlet,A. F.,and D. P. Lettenmaier, Effects of climatechange on hydrology and water resources objectives inthe Columbia River basin, Journal of the AmericanWater Resources Association, 35(6),pp.1597-1625,1999.

Hatch, U.,S. Jagtap, J. Jones,and M.Lamb, Potentialeffects of climate change on agricultural water use inthe Southeast US, Journal of the American WaterResources Association, 35(6), pp.1561-1563,1999.

Hayden,B. P.,Extratropical storms: Past,present,andfuture, Proceedings of the AWRA Specialty Conferenceon Potential Consequences of Climate Variability andChange to Water Resources of the United States,American Water Resources Association,Herndon,VA,pp.93-96,1999a.

Hayden,B. P.,Climate change and extratropical stormi-ness in the United States: An Assessment, Journal ofthe American Water Resources Association, 35(6),pp.1387-1399,1999b.


432

Lambert,S. J.,The effect of enhanced greenhousewarming on winter cyclone frequencies and strengths,Journal of Climate, 8(5),pp.1447-1452,1995.

Lettenmaier, D. P.,E. F.Wood,and J. R.Wallis,Hydro-cli-matological trends in the continental United States1948-1988, Journal of Climate, 7, pp.586-607,1994.

Leung,R.L.,and S. J. Ghan, Pacific Northwest climatesensitivity simulated by a regional climate model drivenby a GCM, Part I:Control simulations, Journal ofClimate, 12, pp.2010-2030,1999a.

Leung,R.L.,and S. J. Ghan, Pacific Northwest climatesensitivity simulated by a regional climate model drivenby a GCM, Part II: 2 X CO2 simulations, Journal ofClimate, 12, pp.2031-2053,1999b.

Leung,R.L.,and M.S.Wigmosta, Potential climatechange impacts on mountain water resources in thePacific Northwest, Journal of the American WaterResources Association, 35(6), pp.1463-1473,1999.

Lins,H. F.,and J. R.Michaels,Increasing US streamflowlinked to greenhouse forcing, Eos Transactions, 75, pp.281-283,1994.

Lins,H. F.,and J. R.Slack, Streamflow trends in theUnited States, Geophysical Research Letters, 26(2),pp.227-230,1999.

Matalas, N. C.,Note on the assumption of hydrologicstationarity, Water Resources Update No. 112, TheUniversities Council on Water Resources,Carbondale,IL,pp.64-72,1988.

McCabe, G. J.,and D. M.Wolock,Sensitivity of irrigationdemand in a humid-temperate region to hypotheticalclimate change, Climatic Change, 37, pp.89-101,1992.

McCabe, G. J.,and D. M.Wolock,General-Circulation-Model simulations of future snowpack in the WesternUnited States, Journal of the American WaterResources Association, 35(6),pp.1473-1484,1999.

Mekis,E.,and W. Hogg,Rehabilitation and analysis ofCanadian daily precipitation time series, Conference onApplied Climatology, Reno,Nevada,pp.20-24,October1997.

Meyer, J. L.,M. J. Sale, P. J. Mulholland,and N. L. Poff,Impacts of climate change on aquatic ecosystem func-tioning and health, Journal of the American WaterResources Association, 35(6), pp.1373-1386,1999.

Miller, N. L., J. Kim,and R.K.Hartman,Downscaled cli-mate and streamflow study of the southwestern UnitedStates, Journal of the American Water ResourcesAssociation, 35(6),pp.1525-1538,1999.

Mulholland, P. J.,and M. J. Sale,Impacts of climatechange on water resources: Findings of the IPCCregional assessment of vulnerability for North America,Water Resources Update No. 112, The UniversitiesCouncil on Water Resources,Carbondale,IL,pp.10-16,1998.

National Research Council, Water Transfers in theWest: Efficiency, Equity, and the Environment,National Academy Press,Washington,DC,1992.

National Research Council, Valuing Groundwater:Economic Concepts and Approaches, NationalAcademy Press,Washington,DC,1997.

O’Connor, R.E.,B.Yarnal,R.Neff, R.Bord, N.Wiefek,C.Reenock,R.Shudak,C.L. Jacoy, P. Pascale,and C. G.Knight,Weather and climate extremes, climate change,and planning: Views of community water system man-agers in Pennsylvania’s Susquehanna River basin,Journal of the American Water Resources Association,35(6),1999.

Pacific Northwest Regional Report, Impacts ofClimate Variability and Change, Pacific Northwest,The JISAO/SMA Climate Impacts Group,University ofWashington,Seattle,WA,2000.

Schilling,K.E.,Water resources: The state of the infra-structure, report to the National Council on PublicWorks Improvement,Washington,DC,May 1987.

Snover,A.,E.Miles,and B.Henry, Regional workshop onthe impacts of global climate change on the PacificNorthwest final report,University of Washington,Seattle, July 1997.

Solley,W. B.,R.R.Pierce,and H.A. Perlman,Estimateduse of water in the United States in 1995, USGeological Survey Circular 1200, Denver, Colorado,1998.

Solley,W. B.,and H.A. Perlman, Factors affecting wateruse patterns in the United States,1950-1995,with pro-jections to 2040, Proceedings of the AWRA SpecialtyConference on Potential Consequences of ClimateVariability and Change to Water Resources of theUnited States, American Water Resources Association,Herndon,VA,pp.65-67,1999.


433

Vandemoer, C.,Impacts of climate change on AmericanIndian water rights and resources, Proceedings of theAWRA Specialty Conference on PotentialConsequences of Climate Variability and Change toWater Resources of the United States, American WaterResources Association,Herndon, VA,pp.17-21,1999.

Wang,B., D. C.Trabant, G.Weller, and P.Anderson, Effectof climate variability and change on the waterresources of Alaska, Proceedings of the AWRASpecialty Conference on Potential Consequences ofClimate Variability and Change to Water Resources ofthe United States, American Water ResourcesAssociation,Herndon,VA,pp.273-276,1999.

Western Water Policy Review Advisory Commission,Water in the West: Challenge for the Next Century,National Technical Information Service,Springfield,Virginia,1998.

Winter,T. C.,The vulnerability of wetlands to climatechange: A hydrologic landscape perspective, Journalof the American Water Resources Association, 35(6),1999.

Wolock, D. M.,and G. J. McCabe,Explaining spatial vari-ability in mean annual runoff in the conterminousUnited States, Climate Research, 11, 149-159,1999a.

Wolock, D. M.,and G. J. McCabe,Simulated effects of cli-mate change on mean and annual runoff in the conter-minous United States, Proceedings of the AWRASpecialty Conference on Potential Consequences ofClimate Variability and Change to Water Resources ofthe United States, American Water ResourcesAssociation,Herndon,VA,pp.161-164, 1999b.

Woodhouse,C.A.,and J.T. Overpeck,A Paleoclimaticrecord of drought for the past 2,000 years, Proceedingsof the AWRA Specialty Conference on PotentialConsequences of Climate Variability and Change toWater Resources of the United States, American WaterResources Association,Herndon, VA,pp.59-64,1999.

Squillace, P. J.,M. J. Moran,W.W. Lapham,C.V. Price,R.M.Clawges,and J. S.Zagorski,Volatile Organic Compoundsin Untreated Ambient Groundwater of the UnitedStates,1985-1995, Environmental Science andTechnology, pp.4176-4187,December 1999.

Stakhiv, E.,and K.Schilling,What can water managersdo about global warming?, Update Water Resources,No.112,The Universities Council on Water Resources.pp.33-40,1998.

Union of Concerned Scientists and the EcologicalSociety of America,Confronting Climate Change inCalifornia: Ecological Impacts on the Golden State,62 pp., UCS Publications,Cambridge,November 1999.

US Army Corps of Engineers,National study of watermanagement during drought: The report to Congress,IWR Report 94-NDS-12, September 1996.

US Army Corps of Engineers,National inventory ofdams,1996.

US Environmental Protection Agency, National waterquality inventory: 1996 report to Congress, EPA841-R-97-008, Washington,DC,1998.

US Water Resources Council,The Nation’s waterresources 1975-2000,second National WaterAssessment,US Government Printing Office,Washington,DC,1978.

US Department of the Interior, Geological Survey,National water summary 1983:Hydrologic events andissues, Water Supply Paper 2250, 1984.

US Department of the Interior, Climate change:Impacts on water resources,15 pp., June 30,1997.

US Department of the Interior, Geological Survey,Strategic directions for the US Geological SurveyGroundwater Resources Program:A report to Congress,November 30,1998.

US Department of the Interior, Geological Survey,National Assessment of Water Quality,www.water.usgs.gov/lookup/get?nawqa.html US National Assessment,draft report of the WaterSector, 2000.http://www.nacc.usgcrp.gov/sectors/water/draft-report/full-report.html


434


Water Resources Sector Assessment TeamD.Briane Adams*,US Geological SurveyPeter Gleick*, Pacific Institute for Studies in

Development,Environment,and SecurityBeth Chalecki, Pacific Institute for Studies in

Development,Environment,and SecurityJoseph Dellapenna,Villanova UniversityTed Engman, NASA Goddard Space Flight CenterKenneth D. Frederick,Resources for the FutureAris P. Georgakakos,Georgia Institute of TechnologyGerald Hansler, Delaware River Basin Commission

(retired)Lauren Hay, US Geological SurveyBruce P. Hayden,National Science FoundationBlair Henry,The Northwest Council on Climate

Change

Steven Hostetler, US Geological SurveyKatharine Jacobs,Arizona Department of Water

ResourcesSheldon Kamieniecki,University of Southern

CaliforniaRobert D. Kuzelka,University of Nebraska-LincolnDennis Lettenmaier, University of WashingtonGregory McCabe,US Geological SurveyJudy Meyer, University of GeorgiaTimothy Miller, US Geological SurveyPaul C.“Chris”Milly, NOAA Geophysical Fluid

Dynamics LaboratoryNorman Rosenberg,Battelle-Pacific Northwest

National LaboratoryMichael J. Sale,Oak Ridge National LaboratoryJohn Schaake,National Oceanic and Atmospheric

AdministrationGregory Schwarz,US Geological SurveySusan S.Seacrest,The Groundwater FoundationEugene Z.Stakhiv, US Army Corps of EngineersDavid Wolock,US Geological Survey



435

CHAPTER 15

POTENTIAL CONSEQUENCES OF CLIMATEVARIABILITY AND CHANGE FOR HUMANHEALTH IN THE UNITED STATESJonathan A.Patz1, Michael A.McGeehin2, Susan M. Bernard1, Kristie L.Ebi3, Paul R.Epstein4,Anne Grambsch5, Duane J. Gubler6, Paul Reiter7, Isabelle Romieu2, Joan B.Rose8, Jonathan M.Samet9, Juli Trtanj10, with Thomas F. Cecich11,12

Contents of this Chapter Preface

Chapter Summary

Introduction

Context

Climate and Human Health

Key Issues

Temperature-related Illnesses and Deaths

Health Effects Related to Extreme Weather Events

Air Pollution-related Health Effects

Water- and Food-borne Diseases

Insect-,Tick-,and Rodent-borne Diseases

Additional Issues



Literature Cited

1Department of Environmental Health Sciences, Johns Hopkins University School of Hygiene and Public Health;2National Center for Environmental Health, U.S.Centers for Disease Control and Prevention; 3EPRI; 4Center for

Health and the Global Environment, Harvard Medical School; 5Office of Research and Development, U.S.

Environmental Protection Agency; 6Division of Vector-Borne Diseases, U.S.Centers for Disease Control and

Prevention; 7Division of Vector-Borne Diseases, U.S.Centers for Disease Control and Prevention,Dengue Branch;8Department of Marine Sciences,University of South Florida; 9Department of Epidemiology, Johns Hopkins

University School of Hygiene and Public Health; 10Office of Global Programs,National Oceanic and Atmospheric

Administration, 11Glaxo Wellcome,Inc., 12Liaison for the National Assessment Synthesis Team. 437

Extreme weather events-related health effects Health impacts from weather disasters range fromacute trauma and drowning,to more medium- andlong-term effects,such as conditions of unsafewater, and post traumatic stress disorder (PTSD).The health impacts of extreme weather eventssuch as floods and storms hinge on the vulnerabili-ties of the natural environment and the local popu-lation,as well as on their capacity to recover.Thelocation of development in high-risk areas increas-es a community’s vulnerability to extreme weatherevents.Adverse health outcomes in the US are lowcompared with global figures partly because of themany federal,state,and local government agenciesand non governmental organizations (NGOs)engaged in disaster planning,early warning,andresponse.

Air pollution-related health effectsQuantitative studies of the potential ef fects of cli-mate change on air quality have primarily focusedon the impact of increased temperature and ultra-violet radiation on ozone formation.In general,these few studies find that ozone concentrationsincrease as temperatures rise.The specific type ofchange (i.e.,local, regional,or global),the directionof change in a particular location (i.e.,positive ornegative),and the magnitude of change in overallair quality (i.e., for all of the criteria air pollutants)that may be attributable to climate change,howev-er, are not known.Additionally, climate change mayalter the distribution and types of airborne aller-gens.

Emissions scenarios are central to assessing futureair quality in addition to assessing the effect ofaltered weather on specific air pollutant formationand/or transport.Integrated air quality modelingstudies will be necessary to assess more quantita-tively the potential health impacts of air qualitychanges associated with global climate change.

Water- and food-borne diseases Weather influences the transport of microbialagents,via rainfall runoff over contaminatedsources.Temperature also influences the occur-rence of bacterial agents,toxic algal blooms (redtides),and survival of viral pathogens that causeshellfish poisoning. Management of sewage andother wastes,and watershed protection are impor-tant to reducing health risks. Federal and state reg-ulations protect much of the US population.Nonetheless,if climate variability increases,current

PREFACEProjections of the extent and direction of the poten-tial health impacts of climate variability and changeare extremely difficult to make because of the manyconfounding and poorly understood factors associat-ed with potential health outcomes,population vul-nerability, and adaptation.In fact,the relationshipbetween weather and specific health outcomes isunderstood for a relatively small number of diseases,with few quantitative models available for analysis.The costs,benefits and availability of resources toaddress adaptation measures also require evaluation.Research aimed at filling the priority knowledgegaps identified in this assessment would allow formore quantitative assessments in the future.

CHAPTER SUMMARYBecause human health is intricately bound toweather and the many complex natural systems itaffects,it is possible that climate change as project-ed will have a measurable impact,both beneficialand adverse,on health outcomes associated withweather and/or climate.We identified and assessedfive such categories of health outcomes:1) tem-perature-related morbidity and mortality;2) healtheffects of extreme weather events (i.e.,storms,tor-nadoes,hurricanes,and precipitation extremes);3)air pollution-related health effects;4) water- andfood-borne diseases;and 5) insect-,tick-,and rodent-borne diseases.

Temperature-related Morbidity and Mortality The more frequent heat waves projected to accom-pany climate change would pose a risk,particularlyagainst the backdrop of an aging US population,asthe elderly are most susceptible to dying fromextreme heat.Beyond individual behavioral changes,adaptation measures include development of com-munity-wide heat emergency plans,improved heatwarning systems,and better heat-related illness man-agement plans.Death rates are higher in winter thanin summer and it is possible that milder winterscould reduce deaths in winter months.However, therelationship between winter weather and mortalityhas been difficult to interpret.The net effect on win-ter mortality from climatic changes is uncertain andthe overall balance between changes in summer andwinter weather-related deaths is unknown.

438


deficiencies in watershed protection and stormdrainage systems will probably increase the risk ofcontamination events.

Insect-, tick-, and rodent-borne diseases The ecology and transmission dynamics of insect-and rodent-borne infections are complex andunique for each disease.Many of these diseasesexhibit a distinct seasonal pattern,suggestive ofweather sensitivity. But demographic,sociologicaland ecological factors also play a critical role in

determining disease transmission.The moderatingeffect of these other factors makes it unlikely thatincreasing temperatures alone will have a majorimpact on tropical diseases spreading into the US.There is greater uncertainty regarding more indige-nous diseases that cycle through animals and canalso infect humans.Further studies of transmissiondynamics,and of pathogen, vector, and animal reser-voir host ecology are required to determinewhether these diseases will increase or decrease with climate change.

439

Chapter 15 / Human Health in the United States

Key Findings

· Multiple levels of uncertainty preclude any defin-itive statement on the direction of potentialfuture change for each of the health outcomesassessed.

· Although our report mainly addresses adversehealth outcomes,some positive health outcomeswere identified,notably reduced cold-weathermortality, which has not been extensively exam-ined.

· At present, much of the US population is protect-ed against adverse health outcomes associatedwith weather and/or climate,although certaindemographic and geographic populations are atincreased risk.

· Vigilance in the maintenance and improvementof public health systems and their responsive-ness to changing climate conditions and to iden-tified vulnerable subpopulations should help toprotect the US population from adverse healthoutcomes of projected climate change.

440


project that some degree of projected climatechange over the next several decades cannot be pre-vented,as a result of already elevated concentrationsof greenhouse gases in the atmosphere, even if miti-gation steps are taken.Thus,it is important to under-stand what adaptation measures might be desirable,or are feasible, regardless of mitigation, given thecurrent climate projections.Future climate changeassessments might choose to link adaptation andmitigation research and impacts.

CONTEXT Because human health is intricately bound to weath-er and the many complex natural systems it affects,it is possible that climate change as projected willhave a measurable impact,both beneficial andadverse,on health outcomes associated with climateand/or weather (see Figure 1).These outcomesinclude temperature-related illnesses and deaths,injuries or fatalities from extreme weather events(i.e.,storms,tornadoes,hurricanes,and precipitationextremes),air pollution- related health effects,anddiseases carried by water (water- and food-bornediseases) and by organisms such as mosquitoes,ticks,mites,and rodents (vector- and rodent-bornediseases).

To establish a baseline for projections of the poten-tial impacts of climate on health, we reviewed thecurrent status and context of health in the US,asreflected in indicators such as life expectancy andthe leading causes of death.We also identified possi-ble strains on public health and health care systems,such as cost and population growth.Urbanization,funding for public health infrastructure (e.g.,sanita-tion systems and medical research) and scientificdevelopments contributed to advances in health sta-tus in the past and are expected to do so in thefuture.Environmental conditions,such as air andwater quality, are important determinants of health.

Chronic diseases—heart disease,cancer, stroke,andchronic obstructive pulmonary disease are the lead-ing four— accounted for almost 75% of all US

POTENTIAL CONSEQUENCES OF CLIMATEVARIABILITY AND CHANGE FOR HUMAN HEALTH IN THE UNITED STATES

INTRODUCTION

This chapter is based on a literature review and con-sultation with experts,interested researchers,andmembers of the public health community. Using,asan underlying set of assumptions, climate changeprojections developed for the national assessment,we analyzed the potential relationships between cli-mate variability, climate change,and human healthwithin a given framework of questions:

• What is the current status of the nation’s healthand what are current stresses on our health?

• How might climate variability and change affectthe country’s health and existing or predictedstresses on health?

• What is the country’s capacity to adapt to cli-mate change, for example,through modificationsto the health infrastructure or by adopting spe-cific adaptive measures?

• What essential knowledge gaps must be filled tofully understand the possible impacts of climatevariability and change on human health?

These questions were developed to tailor the man-date of the National Assessment to the health sector.That mandate was to identify, for each particularsector or region 1) the current status;2) the expect-ed impacts of climate variability and change;3) theadaptive capacity;and 4) the research gaps.Responding to these questions enabled assessmentparticipants to evaluate a baseline and then to iden-tify adaptation measures and research needs.Consistent with the National Assessment as a whole,the health sector did not address the question of thespecific role of anthropogenic (human-caused) con-tributions to changes to climate or identify measuresto reduce emissions of greenhouse gases or thepresence of greenhouse gases in the atmosphere(e.g.,through carbon dioxide sequestration).Theseissues are critically important,and are the focus ofother past and ongoing research and programs inthe United States and elsewhere.However, theextent and success of current and future mitigationmeasures is uncertain.In addition, climate scientists

441


Similarly, although the proportion of childrenyounger than 5 years of age is not expected to growas significantly as the proportion of the elderly, theirnumber will increase even if immigration levels arekept constant.The variables that may affect chil-dren’s special vulnerability to the possible impactsof climate change include:poverty [currentlyapproximately 20% of US children are poor (NCHS,1998)];access to medical care;and children’s sus-ceptibility to environmental hazards because oftheir size,behavior, and the fact that they are grow-ing and developing (Landrigan et al.,1999).

Finally, it is anticipated that the proportion ofimmunocompromised people in the US mayincrease with the aging of the population and thesuccess of medical treatments (e.g.,cancer therapyand HIV medications),but data are difficult toobtain. For example,survival has improved foracquired immunodeficiency syndrome (AIDS)patients, resulting in a 12% increase from 1996-1997in the number of people living with AIDS (CDC,1998).AIDS patients and other immunocompro-mised individuals may be more susceptible to water-borne and vector-borne pathogens,to the adverseimpacts of exposure to elevated levels of certain airpollutants,and to debilitation due to physical stress-

deaths in 1996 for the 25- to 64- year old age group(NCHS,1998).Injuries and infectious diseases remainsignificant causes of morbidity and mortality in theUS;infectious diseases caused one third of the deathsin the US in 1992,primarily because of respiratorytract infections,human immunodeficiency virus(HIV),and septicemia (Pinner et al.,1996). Patterns ofillness and death vary substantially by socioeconomicstatus, geographic region, race, age,and gender(NCHS,1998).

Certain populations within the US — the poor, theelderly, children,and immunocompromised individu-als — may be more vulnerable to many of the healthrisks that might be initially exacerbated by climatechange.Poverty, for example,is a risk factor for heat-related illnesses and deaths,because the poor aremore likely to live in urban areas and are less likely tobe able to afford air-conditioning systems.Thus,mak-ing air-conditioned environments readily available tothe poor is an adaptive response strategy to reduceillnesses and deaths in heat waves. Understandingwhat groups may be the most affected by climatechange is critical to ef fective targeting of preventionor adaptation strategies. For example,air pollutionand heat advisory warnings should specifically targetchildren and the elderly, respectively.

It is also important to recognize that there are racialdifferences in health outcomes,including those asso-ciated with weather and/or climate,such as heatwaves;these differences may be associated withpoverty status,which is disproportionately highamong African-Americans.For example,data on the1995 heat wave in Chicago indicate that mortalityamong African-Americans was 50% higher thanamong whites (Whitman et al.,1997).The disparitylikely reflects residence in inner- city neighborhoods,poverty, housing conditions,and medical conditions(Applegate et al.,1981; Jones et al.,1982;Kilbourneet al.,1982).

It is important to recognize that the proportion ofelderly (65 years of age and older) and very elderly(85 years of age and older) residents is expected torise in the coming decades.The proportion of thesenior population in the very elderly category isgrowing fast:their numbers rose 274% between 1960and 1994,while the entire US population grew only45% (Hobbs and Damon,1998).Aging can be expect-ed to be accompanied by multiple, chronic illnessesthat may result in increased vulnerability to infectiousdisease or external/environmental stresses such asextreme heat (Hobbs and Damon,1998). Poverty,which increases with age in the elderly, may add tothis vulnerability (Day, 1996).

Potential Health Effects of ClimateVariability and Change

Figure 1: Schematic diagram of the potential health effects of cli-mate variability and change. (Source, Patz et al., 2000)* Moderating influences include non-climate factors that affect cli-

mate-related health outcomes, such as: population growth anddemographic change; standards of living; access to health care;improvements in health care; and public health infrastructure. ** Adaptation measures include actions to reduce risks of adverse

health outcomes, such as: vaccination programs; disease surveil-lance; monitoring; use of protective technologies (e.g., air condi-tioning, pesticides, water filtration/treatment); use of climate fore-casts; and development of weather warning systems; emergencymanagement and disaster preparedness programs; and public edu-cation. See Color Figure Appendix


that over the relevant time period the US climatewill be characterized by increased temperatures,analtered hydrologic cycle,and increased variability.These projections are based in part on historicaldata;however, a detailed systematic record of weath-er parameters is only available for some places forapproximately the last hundred years,although indi-rect measurements from ice cores,tree rings,otherpaleo-data and written history extend further(Houghton,1997).In the past 100 years,the globalsurface temperature has warmed between 0.7 and1.4° F (Easterling et al.,1997; Jones et al.,1999;

es,such as those experienced during heat waves orin adverse emergency weather conditions,unlessthey can be adequately protected from those stress-es with access to air conditioning,sanitation,safewater, and sufficient food.

CLIMATE AND HUMANHEALTH The National Assessment climate models project

Table 1. Summary of the Health Sector Assessment

Potential Weather Direction of Examples of Priority research areashealth factors of possible some specificimpacts interest * change in adaptation

health impact strategies

Heat-related Extreme heat Air conditioning Improved prediction, warning and responseillnesses and and stagnant Urban design and energy systemsdeaths air masses Early warning Exposure assessment

Weather relationship to influenza and other Winter deaths Extreme cold causes of winter mortality

SnowIce

Extreme Precipitation Early warning Improved prediction, warning and responseweather variability Engineering Improved surveillanceevents-related (heavy rainfall Zoning and Investigation of past impacts and health effects events †) building codes effectiveness of warnings

Storms

Air pollution- Temperature Early warning Relationships between weather and air related health Stagnant air Mass transit pollution concentrationseffects masses Urban planning Combined effects of temperature/humidity

Pollution control on air pollutionEffect of weather on vegetative emissions

and allergens (e.g.,pollen)

Water- and Precipitation Surveillance Improved monitoring effects of food-borne Estuary water Improved weather/environment on marine-related diseases temperatures water systems disease

engineering Land use impacts on water quality (watershed protection)

Enhanced monitoring/mapping of fate andtransport of contaminants

Vector- and Temperature Surveillance Rapid diagnostic testsrodent-borne Precipitation Vector control Improved surveillancediseases variability programs Climate-related disease transmission

Relative dynamic studieshumidity

* Based on projections pr ovided by the National Assessment Synthesis Team.Other scenarios might yield different changes.† Projected change in frequency of hurricanes and tornadoes is unknown.

443


NRC,2000).In the contiguous US,temperatureshave increased by approximately 1°F (Karl et al.,1996),and precipitation has been increasing in theUS,with much of this change due to increases inheavy-precipitation events (> 2 inches [5 cm] perday) and decreases in light-precipitation events (Karlet al.,1995;Karl et al.,1996;Karl and Knight,1998).These historical data are consistent with climatechange theory that suggests an altered hydrologicalcycle accompanying warming of the Earth’s surface(Fowler and Hennessey, 1995;Mearns et al.,1995;Trenberth,1999).

We examined the impact of this projected climatechange on five health outcomes:1) temperature-related morbidity and mortality;2) health effects ofextreme weather events (i.e.,storms,tornadoes,hur-ricanes,and precipitation extremes);3) air pollution-related health effects;4) water- and food-borne dis-eases;and 5) vector- and rodent-borne diseases.Some of these outcomes are relatively direct (e.g.,effects of exposure to extreme heat or extremeevents);others involve intermediate and multiplepathways,making assessments more challenging(see Figure 1).

Projections of the extent and direction of somepotential health impacts of climate variability andchange can be made,but there are many layers ofuncertainty (see Table 1). First,methods to projectchanges in climate over time continue to improve,but climate models are unable to accurately projectregional-scale impacts.Second,basic scientific infor-mation on the sensitivity of human health to aspectsof weather and climate is limited.In addition,thevulnerability of a population to any health riskvaries considerably depending on moderating fac-tors such as population density, level of economicand technological development,local environmentalconditions,preexisting health status,the quality andavailability of health care,and the public health infrastructure.

It is also dif ficult to anticipate what adaptive meas-ures might be taken in the future to mitigate risks ofadverse health outcomes,such as vaccines,diseasesurveillance,protective technologies (e.g.,air condi-tioning or water filtration/treatment),use of weatherforecasts and warning systems,emergency manage-ment and disaster preparedness programs,and pub-lic education (see Figure 1).As they do currently, theneed for and the success of adaptation measurescan be expected to vary in different parts of thecountry—for example,Chicago must plan for heatwaves,and communities along the southeast coastmust be prepared for hurricanes. For the most part,

government organizations fund public health sys-tems within the US.Continued investments inadvancing the public health infrastructure are cru-cial for adapting to the potential impacts of climatevariability and change.

KEY ISSUES • Temperature-related illnesses and deaths • Health effects related to extreme weather events • Air pollution-related health effects • Water- and food-borne diseases • Insect-,tick-,and rodent-borne diseases

1. Temperature-related Illnessesand Deaths

Heat and heat waves are projected to increase inseverity and frequency with increasing global mean

Figure 2: Both models project substantial increases in the July heatindex (which combines heat and humidity) over the 21st century.These maps show the projected increase in average daily July heatindex relative to the present. The largest increases are in the south-eastern states, where the Canadian model projects increases ofmore than 25˚F. For example, a July day in Atlanta that now reach-es a heat index of 105˚F would reach a heat index of 115˚F in theHadley model, and 130˚F in the Canadian model.(Map by BenjaminFelzer, UCAR, based on data from Canadian and Hadley modelingcenters.) See Color Figure Appendix

July Heat Index Change - 21 st Century

+25ºF

+20ºF

+15ºF

+10ºF

+5ºF

0ºF

Canadian Model

Hadley Model

444


temperatures (Meehl et al.,2000);see Figure 2.Studies of heat waves in urban areas have shown anassociation between increases in mortality andincreases in heat,measured by maximum or mini-mum temperature,and heat index (a measure of tem-perature and humidity).Some of these studies adjustfor other weather conditions (Semenza et al.,1996;Kalkstein and Greene,1997). For example,after afive-day heat wave in 1995 in which maximum tem-peratures in Chicago,Illinois ranged from 93°F to104°F, the number of deaths increased 85% over thenumber recorded during the same period of the pre-ceding year (CDC,1995);see Figure 3.At least 700excess deaths (i.e.,deaths beyond those expected forthat period in that population) were recorded,mostof which were directly attributed to heat (CDC,1995;Semenza et al.,1996;Semenza et al.,1999).

Exposure to extreme and prolonged heat is associat-ed with heat cramps,heat syncope (fainting),heatexhaustion,and heat stroke.These health effectsappear to be related to environmental temperaturesabove those to which the population is accustomed.Models of weather-mortality relationships indicatethat populations in northeastern and midwestern UScities are likely to experience the greatest number ofheat-related illnesses and deaths in response tochanges in summer temperature, and that the mostsensitive regions are those where extremely hightemperatures occur infrequently or irregularly(Kalkstein and Smoyer, 1993);see Figure 4. For exam-ple,Philadelphia,Chicago,and Cincinnati have eachexperienced a heat wave that resulted in a largenumber of heat-related deaths.Physiologic and

behavioral adaptations among vulnerable popula-tions may reduce morbidity and mortality due toheat.Although long-term physiologic adaptation toheat events has not been documented,adaptationappears to occur as the summer season progresses;heat waves early in the summer often result in moredeaths than subsequent heat waves or than thoseoccurring later in the summer (Kalkstein andSmoyer, 1993).Heat waves are episodic,andalthough populations may adapt to gradual tempera-ture increases,physiologic adaptation for extremeheat events is unlikely.

Within heat-sensitive regions,populations in urbanareas are the most vulnerable to adverse heat- relat-ed health outcomes.Heat indices and heat-relatedmortality rates are higher in the urban core than insurrounding areas (Landsberg,1981).Urban areasretain heat throughout the nighttime more efficient-ly than do outlying suburban and rural areas(Buechley et al.,1972;Clarke,1972). The absence ofnighttime relief from heat for urban inhabitants is afactor in excessive heat-related deaths.

The size of US cities and the proportion of US resi-dents living in them are projected to increase overthe next century, so it is possible that the popula-tion at risk for heat-related illnesses and deaths willincrease.High-risk sub-populations include peoplewho live in the top floors of apartment buildings incities and who lack access to air-conditioned envi-ronments (either at home or elsewhere).The elderly(Ramlow and Kuller, 1990;CDC,1993;Whitman etal.,1997;Semenza,1999), young children (CDC,1993),the poor (Schuman,1972;Applegate et al.,1981),and people who are bedridden or on medica-tions that affect the body’s thermoregulatory ability(Kilbourne et al.,1982;Di Maio and Di Maio,1993;Marzuk et al.,1998) are particularly vulnerable.

There is evidence that heat-related illnesses anddeaths are largely preventable through behavioraladaptations,including the use of air conditioningand increased fluid intake (Kilbourne et al.,1982),although the magnitude of mortality reduction can-not be predicted.The proportion of housing unitswith central and/or room unit air conditioningranges from below 30% in the Northeast to almost90% in the South (Bureau of the Census,1997a).Theuse of air-conditioning systems in homes, work-places,and vehicles has increased steadily over thepast 30 years and is projected to become nearly uni-versally available in the US by the year 2050 (Bureauof the Census,1997a;Bureau of the Census,1997b).

Figure 3: This graph tracks the maximum temperature (Tmax), heatindex (HI), and heat-related deaths in Chicago each day from July11 to 23, 1995. The gray line shows maximum daily temperature,the blue line shows the heat index, and the bars indicate the num-ber of deaths each day. (Source: NOAA/NCDC) See Color FigureAppendix

Heat Related Deaths in Chicago in July 1995

445


extreme weather events such as floods and stormshinge on the vulnerabilities and recovery capacitiesof the natural environment and the local population.A community’s level of preparedness greatly affectsthe severity of the health impacts of an extremeevent.

From 1945 to 1989,145 natural disasters caused14,536 deaths in the United States,an average of323 deaths per year (Glickman and Silverman,1992).According to the National Weather Service,severe storms caused 600 deaths and 3,799 report-ed injuries in 1997 (NWS,1999).Floods are themost frequent natural disaster and the leading causeof death from natural disasters in the US;the averageannual loss of life is estimated to be as high as 146

Overall death rates are higher in the winter than inthe summer, and it is possible that milder winterscould reduce deaths in winter months (Kalksteinand Greene,1997).However, the relationshipbetween winter weather and mortality is difficult tointerpret. For example,many winter deaths are dueto infectious diseases such as influenza and pneu-monia,and it is unclear how influenza transmissionwould be affected by higher winter temperatures.Inaddition,studies indicate an association betweensnowfall and fatal heart attacks (from winter precipi-tation rather than cold temperatures) (Spitalnic etal.,1996;Gorjanc et al.,1999).The net effect on win-ter mortality from climate change is thereforeextremely uncertain,and the overall balancebetween changes in summer and winter weather-related deaths is unknown.

Beyond individual behavioral changes,adaptationmeasures include the development of community-wide heat emergency plans,improved heat warningsystems,and better heat-related illness managementplans.Research can refine each of these measures,including which weather parameters are mostimportant in the weather-health relationship,theassociations between heat and nonfatal illnesses,theevaluation of implemented heat response plans,andthe effectiveness of urban design in reducing heatretention.

2. Health Effects Related toExtreme Weather Events

Climate change may alter the frequency, timing,intensity, and duration of extreme weather events(Fowler and Hennessey, 1995;Karl et al.,1995;Mearns et al.,1995),i.e.,meteorological events thathave a significant impact on local communities.Increases in heavy precipitation have occurred overthe past century (Karl et al.,1995;Karl and Knight,1998).Future climate scenarios show likely increasesin the frequency of extreme precipitation events,including precipitation during hurricanes (Knutsonand Tuleya,1999).This poses an increased risk offloods (Meehl et al.,2000). Frequencies of tornadoesand hurricanes cannot reliably be projected.Whetherthese changes in climate risk result in increasedhealth impacts cannot currently be assessed.

Injury and death are the direct health impacts mostoften associated with natural disasters. Secondaryhealth effects have also been observed. Theseimpacts are mediated by changes in ecological sys-tems (such as bacterial and fungal proliferation) andpublic health infrastructures (such as the availabilityof safe drinking water).The health impacts of

Figure 4: Deaths due to summer heat are projected to increase inUS cities, according to a study using time-dependent results ( forgreenhouse gas increase only) from several climate models(Kalkstein and Greene, 1997). Mortality rates (number of deaths per100,000 population) are shown from the Max Planck Institute model,the results from which lie roughly in the middle of the models exam-ined (the other climate scenarios used were from Geophysical FluidDynamics Laboratory (GFDL) and the Hadley Centre). Because heat-related illness and death appear to be related to temperatures muchhotter than those to which the population is accustomed, cities thatexperience extreme heat only infrequently appear to be at greatestrisk. For example, Philadelphia, New York, Chicago, and St. Louishave experienced heat waves that resulted in a large number ofheat-related deaths, while heat related deaths in Atlanta and LosAngeles are much lower. In this study, statistical relationshipsbetween heat waves and increased death rates are constructed foreach city based on historical experience. Deaths under a city’sfuture climate are then projected by applying that city’s projectedincidence of extreme heat waves to the statistical relationship thatwas estimated for the city whose present climate is most similar tothe projected climate for the city in question. This approachattempts to represent how people will acclimate to the new averageclimate that they experience. See Color Figure Appendix

Average Summer Mortality Rates Attributed to Hot Weather Episodes

deaths per year (NWS,1992).Hurricanes also posean ongoing threat;an average of two each year makelandfall on the US coastline (NWS,1993).Theimpacts of hurricanes include injuries and deathsresulting from strong winds and heavy rains.

Depending on the severity and nature of the weath-er event,people may experience disabling fear oraversion (Drabek,1996).There is controversy aboutthe incidence and continuation of significant mentalproblems,such as post traumatic stress disorder(PTSD),after disasters (Quarantelli,1985).However,an increase in the number of mental disorders hasbeen observed after several natural disasters in theUS.Increased psychological problems were reportedduring a 5-year period after Hurricane Agnes causedwidespread flooding in Pennsylvania in 1972 (Logueet al.,1979).More recently, a longitudinal study oflocal residents who lived through Hurricane Andrewshowed that 20-30% of the adults in the area metthe criteria for PTSD at 6 months and 2 years afterthe event (Norris et al.,1999).

A population’s ability to minimize the potentialhealth effects associated with extreme weatherevents is based on a number of diverse and interre-lated factors including:building code regulations,warning systems,and disaster policies; evacuationplans;adequate relief efforts;and recovery (Noji,1997).There are many federal,state,and local gov-ernment agencies and nongovernmental organiza-tions involved in planning for and responding tonatural disasters in the US. For example,the FederalEmergency Management Agency (FEMA) recentlylaunched its National Mitigation Strategy (FEMA,1996),which is designed to increase public aware-ness of natural hazard risk and to reduce the risk ofdeath,injury, community disruption,and economicloss.This strategy represents a comprehensive effortto address severe events with a series of initiativesand public-private partnerships.

Future research on extreme weather events andassociated health effects should focus on improvingclimate models to project trends,if any, in regionalextreme events.This type of improved predictioncapability will assist in public health mitigation andpreparedness.In addition,epidemiologic studies ofhealth effects beyond the direct impacts of disasterwill provide a more accurate measure of the fullhealth impacts and will assist in planning andresource allocation.

446


3. Air Pollution-related HealthEffects

Air pollutants have many sources—natural (e.g., veg-etation and volcanoes), agricultural (e.g.,methaneand pesticides),commercial (e.g.,dry cleaning oper-ations and auto body shops),industrial (e.g.electricpower plants and manufacturing facilities),trans-portation (e.g.truck and automobile emissions),andresidential (e.g.home gas and oil burners and woodstoves).Ambient levels of regulated air pollutants(which include particulate matter, ozone,carbonmonoxide,and sulfur and nitrogen oxides) have gen-erally dropped since the mid-1970s,but air qualityin many parts of the country falls short of health-based air quality standards.In 1997, about 107 mil-lion people in the US lived in counties that did notmeet the air quality standards for at least one regu-lated pollutant (USEPA,1998a).

Air pollution is related to weather both directly andindirectly. Climate change may affect exposures toair pollutants by:1) affecting weather and therebylocal and regional pollution concentrations (Penneret al.,1989;Robinson,1989);2) affecting human-caused emissions,including adaptive responsesinvolving increased fuel combustion for power gen-eration;3) affecting natural sources of air pollutantemissions (USEPA,1997a;USEPA,1998a);and 4)changing the distribution and types of airborneallergens (Ahlholm et al.,1998).Local weather pat-terns,including temperature,precipitation, clouds,atmospheric water vapor, wind speed,and winddirection influence atmospheric chemical reactions.They can also affect atmospheric transport process-es and the rate of pollutant export from urban andregional environments to the global scale environ-ments (Penner et al.,1989;Robinson,1989).In addi-tion,the chemical composition of the atmospheremay in turn have a feedback effect on the local cli -mate.

If the climate becomes warmer and more variable,air quality is likely to be affected. For example,ifwarmer temperatures lead to more air-conditioninguse,power plant emissions could increase withoutadditional air pollution controls.Analyses show thathigher surface temperatures are conducive to theformation of ground-level ozone,particularly inurban areas (Morris et al.,1989;NRC,1991;Sillmanand Samson,1995;USEPA,1996;USEPA,1998a);seeFigure 5.

Changing weather patterns contribute to yearly dif-ferences in ozone concentrations (USEPA,1998a);for example,the hot,dry, stagnant meteorological

conditions in 1995 in the central and eastern USwere highly conducive to ozone formation.However,the specific type of change (i.e.,local, regional,orglobal),the direction of change in a particular loca-tion (i.e.,positive or negative),and the magnitude ofchange in air quality that may be attributable to cli-mate change are not known.

Because the effect of climate change on all of the airpollutants of concern,especially particulate matter, isunknown,it is difficult to determine the overalleffect of climate variability and change on respirato-ry health.Health effects associated with climateimpacts on air pollution will depend on future airpollution levels.Since 1970,emissions and ambientair pollutants have declined overall (USEPA,1996).However, the majority of regulated air pollutants arefrom fossil fuel combustion (USEPA,1997a;USEPA,1998a) and,as a result,increased energy and fuel-usewould increase emissions of air pollutants withoutadditional air pollutant controls.Integrated air quali-ty modeling studies will be necessary to assess morequantitatively the potential health impacts of airquality changes associated with global climatechange.These models would need to incorporatevariables such as:1) future human-caused emissions(driven by economic growth,air pollution controls,vehicle usage,and possible changes in use of fuel forheating and cooling);2) future natural emissions(factoring in possible responses to changing cli-mate);and 3) changes in local meteorology due toglobal climate change.

Current exposures to air pollutants have seriouspublic health consequences.Ground-level ozone canexacerbate respiratory diseases by damaging lung tis-sue, reducing lung function,and sensitizing the lungsto other irritants (Romieu,1999).Short-term drops inlung function caused by ozone are often accompa-nied by chest pain,coughing,and pulmonary conges-tion (American Thoracic Society, 1996).Epidemio-logic studies have found that exposure to particulatematter can aggravate existing respiratory and cardio-vascular diseases,alter the body’s defense systemsagainst foreign materials,damage lung tissue,lead topremature death,and possibly contribute to cancer(American Thoracic Society, 1996;Lambert et al.,1998).Health effects of exposures to carbon monox-ide,sulfur dioxide,and nitrogen dioxide can includereduced work capacity, aggravation of existing car-diovascular diseases,effects on respiratory function,respiratory illnesses,lung irritation,and alterations inthe lung’s defense systems (American ThoracicSociety, 1996;Lambert et al.,1998).

In addition to affecting exposure to air pollutants(whether man-made or naturally emitted),there issome chance that climate change will play a role inhuman exposure to airborne allergens.Plant speciesare sensitive to weather, and climate change willpossibly alter pollen production in some plants orthe geographic distribution of plant species(Ahlholm et al.,1998).Consequently, there is somechance that climate change will affect the timing orduration of seasonal allergies,such as hay fever.Theimpact of pollen and of pollen changes on theoccurrence and severity of asthma,the most com-mon chronic disease of childhood,is currently veryuncertain.

447


Figure 5: These graphs illustrate the observed association betweenground-level ozone concentrations and temperature in Atlanta andNew York City (May to October 1988-1990). The projected highertemperature across the US in the 21st century will likely increasethe occurrence of high ozone concentrations, especially becauseextremely hot days frequently have stagnant air circulation pat-terns, although this will also depend on emissions of ozone precur-sors and meteorological factors. Ground-level ozone can exacer-bate respiratory diseases and cause short-term reductions in lungfunction. (Maximum Daily Ozone Chart provided by USEPA.) – SeeColor Figure Appendix

Maximum Daily Ozone Concentrations versus Maximum Daily Temperature in Atlanta and New York

448


There is some chance that climate change will affectthe amount of time individuals spend indoors (e.g.,individuals may spend more time in air conditionedenvironments to avoid extreme heat,or may spendmore time outdoors if winter temperatures aremilder), resulting in changed exposure to indoor airpollutants and allergens.In some cases,these indoorenvironments may be more dangerous than theambient conditions.

Adaptation measures include ensuring the respon-siveness of federal and state air quality protectionprograms to changing pollution levels.These stan-dards are designed to protect the public health bylimiting emissions of key air pollutants and thusreducing ambient concentrations.The PollutantsStandards Index (Davies and Mazurek,1998),an EPA-coordinated health advisory system that provideswarnings for both the general population and sus-ceptible individuals,could be further strengthenedfor specific pollutants.

Future research in the area of health effects associat-ed with air pollution should include:basic atmos-pheric science elucidating the association betweenweather, ozone,particulate matter, and other air pol-lutants and aeroallergens;improving existing models(e.g., expanding the spatial domain and lengtheningthe duration of modeled events) and their linkagewith climate change scenarios;and closing the gapsin our understanding of common pollutants,such asparticulate matter and ozone,and of individualexposures to these pollutants.

4. Water- and Food-borne Diseases

More than 200 million people in the US have directaccess to treated public water supply systems, yet asmany as 9 million annual cases of water-borne dis-ease are estimated (Bennett et al.,1987);high uncer-tainty accompanies this estimate,and reporting isvariable by state (Frost et al.,1996).Although mostof these cases of water-borne disease involve mildgastrointestinal illnesses,other severe outcomessuch as myocarditis (infection of the heart) are nowrecognized.These infections and illnesses can bechronic and even fatal in infants,the elderly, preg-nant women,and people with weakened immunesystems (Gerba et al.,1996;ASM,1998).

In the US, food-borne diseases are estimated tocause 76 million cases of illness,with 325,000 hos-pitalizations and 5,000 deaths per year (Mead et al.,1999).Microbiologic agents in water (e.g.,viruses,bacteria,and protozoa) can contaminate food (e.g.shellfish and fish).In addition,there have been

instances of contamination of fresh fruits and veg-etables by water-borne pathogens (Tauxe,1997).

The routes of exposure to water- and food-bornediseases include ingestion,inhalation,and dermalabsorption of microbial organisms or algal toxins.For example,people can ingest water-borne micro-biologic agents by drinking contaminated water, byeating seafood from contaminated waters,or by eat-ing fresh produce irrigated or processed with con-taminated water (Tauxe,1997).They also can beexposed by contact with contaminated waterthrough commerce (e.g., fishing) or recreation (e.g.,swimming) (Coye and Goldoft,1989).The water-borne pathogens of current concern include virus-es,bacteria,and protozoa.Examples include Vibriovulnificus, a naturally occurring estuarine bacteriumresponsible for a high percentage of the deaths asso-ciated with shellfish consumption (Johnston et al.,1985;Shapiro et al.,1998); Cryptosporidiumparvum and Giardia lamblia, protozoa associatedwith gastrointestinal illnesses (Craun,1998);and bio-logic toxins associated with harmful algal blooms(Baden et al.,1996).Many of these were discoveredonly recently and are the subject of ongoingresearch.

Between 1980 and 1996,401 disease outbreaks asso-ciated with drinking water were reported,withmore than 750,000 associated cases of disease(Craun,1998).More than 400,000 of those cases(including 54 deaths,primarily of individuals whoseimmune systems were compromised by HIV infec-tion or other illness) occurred in a 1993 outbreakof Cryptosporidiosis that resulted from the contami-nation of the Milwaukee,Wisconsin, water supply(Hoxie et al.,1997).A contributing factor in the con-tamination,in addition to treatment system malfunc-tions, was heavy rainfall and runoff that resulted in adecline in the quality of raw surface water arrivingat the Milwaukee drinking water plants (MacKenzieet al.,1994).Studies from other locations in the USfound positive correlations between rainfall andCryptosporidium oocyst and Giardia cyst concen-trations in river water (Atherholt et al.,1998) andhuman disease outbreaks (Weniger et al.,1983;Curriero et al.,2001). Many water treatment facili-ties still have difficulty removing these pathogens.

Changes in precipitation,temperature,humidity,salinity, and wind have a measurable effect on thequality of water used for drinking, recreational andcommercial use,and as a source of fish and shellfish(see Figure 6).Direct weather associations havebeen documented for water-borne disease agentssuch as Vibrio bacteria (Motes et al.,1998),viruses

(Lipp et al.,1999),and harmful algal blooms(Harvell et al.,1999).In Florida during the strongEl Niño of 1997-1998,high precipitation andrunoff greatly elevated the counts of fecal bacteriaand infectious viruses in local coastal waters(Harvell et al.,1999).In Gulf Coast waters, Vibriovulnificus bacteria are especially sensitive towater temperature,which dictates their seasonali-ty and geographic distribution (Lipp and Rose,1997;Motes et al.,1998).In addition,toxic redtides proliferate as seawater temperatures increase(Valiela,1984).Over the past twenty-five yearsalong the East Coast, reports of marine-related ill-nesses increased in correlation with El Niñoevents (Harvell et al.,1999).

For many water-borne diseases,the managementand disposal of sewage,biosolids and other animalwastes,and the protection of watersheds andfresh water flows are critical variables that impactwater quality and the risk of water-borne disease(ASM,1998).In September, 1999,the largestreported water-borne associated outbreak ofEscherichia coli 0157:H7 occurred at a fairgroundin the state of New York and was linked to con-taminated well water (CDC,1999a).Heavy rainsfollowing a period of drought coincided with thismajor outbreak event (New York Department ofHealth,2000).The likelihood of this type of prob-lem occurring could increase under conditions ofhigh soil saturation that enhances the rapid trans-port of microbiologic organisms (Yates and Yates,1988). Finally, many communities in the US contin-ue to use combined sewer and storm waterdrainage systems (Figure 7). These systems maypose a health risk should the frequency or intensi-ty of storms increase,because raw sewage bypass-es treatment and is discharged into receiving sur-face waters during storms (Rose and Simonds,1998).

Climate changes projected to occur in the nextseveral decades,in particular the likely increase inextreme precipitation events,will probably raisethe risk of contamination events.However,whether these increases materialize depends onpolicy responses and the level of maintenance orimprovement of infrastructure.Current adapta-tions for assessing and preventing water-borne dis-eases include legal and administrative measuressuch as water safety criteria,monitoring require-ments,and health outcome surveillance,as man-dated under the Safe Drinking Water Act,withamendments in 1996 (USEPA,1997b).Recent leg-islative and regulatory attention has focused onimproved treatment of surface water to address

449


Figure 6: Monthly distribution of oyster-associated Vibrio vulnifi-cus illness (or shellfish poisoning) and deaths occurring in Floridafrom 1981-1994. Over the 14-year period, higher numbers of casesoccur during summer. Monitoring in Florida shows a statisticallysignificant association between concentrations of this pathogen inestuaries and temperature and salinity, the latter being affected byrainfall and runoff. (Adapted from: Lipp and Rose, 1997.) – SeeColor Figure Appendix

Figure 7: Wastewater systems that combine storm water drainageand sewage and industrial discharges are still in use in about 950communities in the US, mostly in the Northeast and Great Lakesregions. These combined sewer systems deliver both stormdrainage and wastewater to sewage treatment facilities. However,during rain or snowmelt, the volume of incoming water can exceedthe capacity of the treatment system. Under those conditions, com-bined sewer systems are designed to overflow and dischargeuntreated wastewater into surface water bodies, and are termed asa combined sewer overflow (CSO) event. EPA, in 1994, developed aCSO Control Policy that sets forth a national framework for preven-tion of combined sewer overflows through the federal Clean WaterAct’s water discharge permit program. It has been suggested thatif they continue to discharge untreated wastewater during stormevents, combined sewer systems may pose a greater health riskshould the frequency or intensity of storms increase. (Source:USEPA, http://www.epa.gov/owmitnet/cso.htm) See Color FigureAppendix

Seasonality of Shellfish Poisoning in Florida1981-1994

Locations of Combined Wastewater Systems

broad generalizations on the effect of climate onvector-borne diseases (Reiter, 1996;Reiter, 2000).Many of these diseases are no longer present in theUS,mainly because of changes in land use, agricul-tural methods, residential patterns,human behavior,and vector control.However, diseases that may betransmitted to humans from wild animals(zoonoses) continue to circulate in nature in manyparts of the country. Humans can become infectedwith the pathogens that cause these diseasesthrough transmission by insects or ticks. For exam-ple, Lyme disease,which is tick- borne,circulatesamong white-footed mice in woodland areas of theMid-Atlantic,Northeast,upper Midwest and WestCoast of the US,and humans acquire the pathogenwhen they are bitten by infected ticks (Gubler,1998).Flea-borne plague incidence increased in con-junction with increasing rodent populations afterunseasonal winter-spring precipitation in NewMexico (Parmenter et al.,1999).

Humans may also become infected withpathogens that cause zoonotic diseases by directcontact with the host animals or their body flu-ids,as occurs with Hantavirus PulmonarySyndrome (HPS).Hantaviruses are carried bynumerous rodent species and are transmitted tohumans through contact with rodent urine,drop-pings,and saliva,or by inhaling aerosols of theseproducts.In 1993,a previously undocumentedhantavirus, Sin Nombre, emerged in the FourCorners region of the rural southwestern US,caus-ing HPS (Schmaljohn and Hjelle 1997).As of June2000,274 cases had been confirmed in the US,30 inCanada,and 475 in Central and South America.Inthe US,the mortality rate is currently 39% in other-wise healthy individuals (personal communication,James Mills,CDC).

The impact of weather on rodent populations mayaffect disease transmission.The Four Corners out-break was attributed to an explosion in the mousepopulation caused by an increase in their food sup-ply resulting from unusually prolonged rainfall asso-ciated with the 1991-1992 El Niño event(Engelthaler et al.,1999;Glass et al.,2000).

Flooding has also been associated with rodent-borneleptospirosis,as occurred in the 1995 epidemic inNicaragua.A case-control study showed a 15-foldrisk of disease associated with walking throughflood waters (Trevejo et al.,1998).In Salvador,Brazil,a large epidemic of leptospirosis peaked twoweeks after severe flooding in 1996 (Ko et al.,1999).Although leptospirosis cases are rare in theUS,the disease is under-diagnosed (Demers et al.,

microbial contaminants and on ground water andwatershed protection (ASM,1998;USEPA,1998b).

With respect to marine-related human disease out-breaks,protection is provided by measures such asadequate sewage/sanitation systems and safe foodstorage infrastructures,and beach and recreationalwater monitoring (USEPA,1999).However, thesemeasures are inadequate for microbial contami-nants.With increasing trends in food importation,improved surveillance and preventive measures arerequired (Tauxe,1997),as well as a better under-standing of how climate and weather might affectfood and water safety outside the US.

Important knowledge gaps must be addressed toimprove the assessment of the association of climatewith water-borne disease issues.Determinants oftransport and fate of microbial pollutants associatedwith rainfall and snowmelt are not well quantified.Further studies should address the influence of vary-ing land use on the water quality in watersheds. Forurban watersheds, much of the current annual loadof contaminants is transported into fresh andmarine water bodies during storm events. For thesereasons, regional and even localized projections ofchanges in the intensity and frequency of stormsand changes in land use are required for improvingclimate variability/health assessments.

Advances in monitoring are necessary to improveour knowledge base and enhance early warning andprevention capabilities.Application of existing tech-nologies could be expanded,such as molecular fin-gerprinting to track contaminant sources (CDC,1999b),improvement of monitoring systems (CDC,1999c),and the use of satellite remote sensing usedto detect coastal algal blooms (Gower, 1995).Coordination and integration of monitoring acrossthe varying agencies responsible for water-borne,food-borne,and coastal surveillance systems couldgreatly enhance our knowledge and adaptive poten-tial.

5. Insect-, Tick-, and Rodent-borneDiseases

Diseases transmitted between humans by blood-feeding arthropods (insects,ticks,and mites),suchas plague,typhus,malaria, yellow fever, and denguefever were once common in the US and Europe(Philip and Rozeboom,1973;Beneson,1995;Reiter,1996).The ecology and transmission dynamics ofthese “vector-borne”infections are complex,and thefactors that influence transmission are unique toeach disease.It is not possible,therefore,to make450


1983),and the bacteria has been found in samplesfrom both rats and children from surveys conductedin urban areas (Demers et al.,1983;Childs et al.,1992).

Changes in ecosystems and sociologic factors play acritical role in the occurrence of these diseases. Forinstance,the increasing numbers of cases and spreadof Lyme disease in the US and Europe stemmed fromthe reversion of large tracts of agricultural land towoodland and the subsequent increase in mouse,deer, and tick populations,combined with thespread of residential areas into undeveloped areasand farmland (IOM,1992).

Most vector-borne diseases exhibit a distinct season-al pattern,which clearly suggests that they areweather sensitive.Rainfall,temperature,and otherweather variables affect in many ways both vectorsand the pathogens they transmit.Rainfall mayincrease the abundance of some mosquitoes byincreasing the number of their breeding sites(Reisen et al.,1995),but excessive rainfall can flushthese habitats and thus destroy the mosquitoes intheir aquatic larval stages.Increased humidity canextend vector survival times (Reisen et al.,1995).Dry conditions may eliminate the smaller breedingsites,such as ponds and puddles,but create produc-tive new habitats as river flow is diminished.Thus,epidemics of malaria are associated with rainy peri-ods in some parts of the world but with drought inothers.High temperatures can increase the rate atwhich mosquitoes develop into adults,the rate ofdevelopment of the pathogens in the mosquitoes(Watts et al.,1987),and feeding and egg-laying fre-quency.The key factor in transmission is the survivalrate of the vector (Gilles,1993).Higher temperaturesmay increase or reduce survival rate,depending onthe vector, its behavior, ecology, and many other fac-tors.Thus,the probability of transmission may ormay not be increased by higher temperatures.

In some cases,specific weather patterns over severalseasons appear to be associated with increasedtransmission rates. For example,in the midwesternUS,outbreaks of St.Louis encephalitis (SLE,a viralinfection of birds that can also infect and cause dis-ease in humans) appear to be associated with thesequence of warm, wet winters,cold springs,andhot dry summers (Monath,1980).The factors under-lying this association are complex and require moreinvestigation (Reeves and Hammon,1962;Reiter,1988).

In the western US,one study (Reeves et al.,1994)predicted that a 5.5 to 9°F (3-5°C ) increase in aver-

age temperature may cause a northern shift in thedistribution of both Western Equine Encephalitis(WEE) and SLE outbreaks,and a decreased range ofWEE in southern California based on temperaturesensitivity of both virus and mosquito carrier.

Many other factors are important in transmissiondynamics. For example,dengue fever — a viral dis-ease mainly transmitted by Aedes aegypti, a mosqui-to that is closely associated with human habitation— is greatly influenced by house structure,humanbehavior, and general socioeconomic conditions.There is a marked difference in the incidence of thedisease above and below the US- Mexico border:inthe period 1980-1996,43 cases were recorded inTexas,as compared to 50,333 reported cases in thethree contiguous border states in Mexico; Figure 8shows data updated through 1999 (Reiter, 2001).

The tremendous growth in international travelincreases the risk of importation of vector-borne dis-eases,some of which can be transmitted locallyunder suitable circumstances at the right time of theyear (Gubler, 1998). Key preventive measures mustbe directed both at protecting the increasing num-ber of US travelers going to disease-endemic areas,as well as preventing importation of disease by USand non-US citizens.The recent importation of WestNile virus encephalitis into New York illustrates thecontinued need for vigilant surveillance for zoonotic

451


Figure 8: Dengue along the US-Mexico border. Dengue, a mosqui-to-borne viral disease, was once common in Texas (where therewere an estimated 500,000 cases in 1922), and the mosquito thattransmits it remains abundant. The striking contrast in the inci-dence of dengue in Texas versus three Mexican states that borderTexas (64 cases vs. 62,514) in the period from 1980-1999 provides agraphic illustration of the importance of factors other than tempera-ture, such as use of air conditioning and window screens, in thetransmission of vector-borne diseases. (National Institute of Health,Mexico; Texas Department of Health; US Public Health Service.Unpublished data.) – See Color Figure Appendix

Reported Cases of Dengue 1980-1999

on health of economic losses or gains due to cli-mate variability or attempt to assign a monetaryvalue to the health outcomes of climate change.Wedid not address the potential impact that changes inthe hydrologic cycle might have on crop productionand food storage in the US (McMichael et al.,1996).Finally, we did not address stratospheric ozonedepletion (McKenzie et al.,1999),although climatechange may contribute to the delayed recovery ofthe stratospheric ozone hole (Shindell et al.,1998;Kirk-Davidoff et al.,1999),and possibly lead toadverse health impacts from increased UV expo-sure.

ADAPTATION STRATEGIES If climate change occurs as anticipated,it may havesignificant impacts on virtually all systems on whichhuman life depends—biologic, hydrologic,and eco-logic.The extent of the impacts that climate changemay have on human health is very uncertainbecause it is dependent on multiple interrelatedvariables as well as on the condition of our publichealth infrastructure.Climate variability and changewill likely have both positive and negative conse-quences for the health of the US population (seeTable 2).

The future vulnerability of the US population to thehealth impacts of climate change depends on ourcapacity to adapt to potential adverse changesthrough legislative,administrative,institutional,tech-nological,educational,and research-related meas-ures.Examples include building codes and zoning toprevent storm or flood damage,severe weatherwarning systems,improved disease surveillance andprevention programs,improved sanitation systems,education of health professionals and the public,and research addressing key knowledge gaps in cli -mate/health relationships.

diseases potentially brought in by imported animalsor international travelers (Lanciotti et al.,1999).Anactive survey in Florida recently documented under-reporting for some diseases,such as dengue fever(Gill et al.,2000),further demonstrating the need forimproved surveillance to better estimate risk.

Preventive measures from these types of risksinclude vaccinations and drug prophylaxis for trav-elers,information for travelers,and the use of repel-lants and other protective measures.In the US,med-ical personnel should be made aware of thisincreased risk to travelers and of the need toimprove surveillance of imported vector-borne dis-eases.

A high standard of living and well-developed publichealth infrastructure are central to the currentcapacity to adapt to changing risks of vector- androdent-borne diseases in the US.Maintaining andimproving this infrastructure—including surveil-lance,early warning,prevention,and control—remain a priority. Integration of climate,environmen-tal,health,and socioeconomic data may facilitateimplementing public health prevention measures.For example, climate forecasts can assist in diseaseprevention by predicting seasonal or interannualevents such as El Niño,and early warning fromimproved vector and disease surveillance can helpprevent local transmission of imported vector-bornediseases (Colwell and Patz,1998).

ADDITIONAL ISSUES Other health outcomes identified in the literatureand by researchers as potentially affected by climatevariability and change may warrant future study butare beyond the scope of this current assessment. Forexample, we did not address the potential impacts

452


Table 2: Human Health: Key Summary Messages

Multiple levels of uncertainty preclude any definitive statement on the direction of potential future change foreach of the health outcomes assessed.

Although our report mainly addresses adverse health outcomes,some positive health outcomes were identified,notably reduced cold-weather mortality, which has not been extensively examined.

At present, much of the US population is protected against adverse health outcomes associated with weather and/or climate,although certain demographic and geographic populations are at increased risk.

Vigilance in the maintenance and improvement of public health systems and their responsiveness to changing climate conditions and to identified vulnerable subpopulations should help to protect the US popula-tion from adverse health outcomes of projected climate change.

453


Table 3. Summary of Research Needs and Knowledge Gaps

Temperature- Improvement of the early prediction of these events by determining the key weather parametersrelated associated with health morbidity Improvement of urban design to facilitate trees,shade,wind,and other heat-reducing conditions to limit and the “urban heat island effect”mortality Better personal exposure assessment

Heat morbidity modelingUnderstanding of weather relationship to causes of winter mortality

Extreme Improvement of warning systems to provide early, easily understood messages to the populations most weather likely to be affectedevents– Research on the ef fectiveness of educational materials and early warning systemsrelated Long-term health ef fects from severe events,such as nutritional deficiency and mental health effectshealth Standardization of information collection after disasters to better measure morbidity and mortalityeffects Effects of altered land use on vulnerability to extreme weather

Air Association between weather and pollutantspollution- Health impacts of chronic exposure to high levels of ozonerelated Health effects of exposure to ozone in people with asthma and other lung diseaseshealth Interaction of ozone with other air pollutantseffects Mechanisms responsible for the adverse ef fects of air pollutants in the general population and within

susceptible subgroupsMeasures that can modulate the impact of air pollution on health,such as nutrition and other life-style

characteristicsUrban weather modeling for inversions,etc.

Water- and Links between land use and water quality, through better assessment at the watershed level of the food-borne transport and fate of microbial pollutants associated with rain and snowmeltdiseases Methods to improve surveillance and prevention of water-borne disease outbreaks

Epidemiologic studiesMolecular tracing of water-borne pathogensLinks between drinking water, recreational exposure,and food-borne disease monitoringLinks between marine ecology and toxic algaeVulnerability assessment to improve water and waste water treatment systems

Vector- and Improvement of rapid diagnostic tests for pathogensrodent-borne Vaccinesdiseases Improvement of active laboratory-based disease surveillance and prevention systems at the state

and local levelTransmission dynamics (including reservoir host and vector ecology) studiesImprovement of surveillance systems for the arthropod vector and vertebrate hosts involved in the

pathogen maintenance/transmission cycles to allow for more accurate predictive capability for epidemic/epizootic transmission

More effective and rapid electronic exchange of surveillance data

Many of these adaptive responses are desirable froma public health perspective ir respective of climatechange.For example, reducing air pollution obvious-ly has both short- and long-term health benefits.Improving warning systems for extreme weatherevents and eliminating existing combined sewerand storm water drainage systems are other meas-ures that can ameliorate some of the potentialadverse impacts of current climate extremes and ofthe possible impacts of climate change.Improved

disease surveillance and prevention systems at thestate and local levels are already needed.Adaptationis a complex undertaking,as demonstrated by thevarying degrees of effectiveness of current efforts tocope with climate variability. Considerable work stillneeds to be done to assess the feasibility (e.g., abili-ty of a community to incur the costs) and the effec-tiveness of alternative adaptive responses,and todevelop improved mechanisms for coping with cli -mate variability and change.

454


CRUCIAL UNKNOWNS/RESEARCH NEEDS

We are still learning about the linkages betweenweather and human health in the present, even aswe try to anticipate the health effects of climatevariability and change in the future.(Specific knowl-edge gaps are discussed in the Health Sector report,in press,and are listed in Table 3.)

LITERATURE CITEDAhlholm J.U.,M.L.Helander, J. Savolainen.Genetic andenvironmental factors affecting the allergenicity ofbirch (Betula pubescens ssp.czerepanovii [Orl.] Hamet-ahti) pollen. Clinical and Experimental Allergy28:1384-8,1998.

American Society for Microbiology (ASM),Microbialpollutants in our Nation’s water:Environmental andpublic health issues,American Society for Microbiology,Office of Public Affairs,Washington, DC,1998.

American Thoracic Society, Health effects of outdoorair pollution,part 2, American Journal of Respiratoryand Critical Care Medicine, 153, 477-98,1996.

Applegate,W. B., J.W. Rynyan,L.Brasfield,M.L.Williams,C. Konigsberg,and C. Fouche,Analysis of the 1980 heatwave in Memphis, Journal of the American GeriatricSociety, 29, 337-42,1981.

Atherholt,T.B.,M.W. LeChevallier,W. D. Norton,and J. S.Rosen,Effect of rainfall on giardia and crypto, Journalof the American Water Works Association, 90(9),66-80,1998.

Baden, D. G.,L.E.Glemming,and J.A.Bean,Marine tox-ins,in Handbook of Clinical Neurology, Intoxicationsof the Nervous System: Part II, edited by F. de Wolf,Elsevier Science,Amsterdam,141-75,1996.

Beneson,A.S., Control of Communicable DiseasesManual, 16th edition,American Public HealthAssociation,Washington DC,577,1995.

Bennett, J.V.,S. D. Homberg,M. F. Rogers,and S.L.Soloman,Infectious and parasitic diseases, AmericanJournal of Preventive Medicine, 55, 102-14,1987.

Buechley, R.W., J.V. Bruggen,and L.E.Truppi,Heatisland = death island?, Environmental Research, 5, 85-92,1972.

Bureau of the Census,American housing survey for theUnited States in 1995,US Bureau of the Census,Washington,DC,1997a.

Bureau of the Census, Statistical Abstracts of theUnited States:1997 (11th edition),US Bureau of theCensus,Washington,DC,1997b.

Centers for Disease Control and Prevention (CDC),Heat-related deaths — United States,1993, Morbidityand Mortality Weekly Report 42, 558-560,1993.

Centers for Disease Control and Prevention (CDC),Heat-related mortality—Chicago, July 1995, Morbidityand Mortality Weekly Report 44, 577-579,1995.

Centers for Disease Control and Prevention (CDC),HIV/AIDS surveillance report,December 1998,CDC,Atlanta,Georgia,10,December 1998.

Centers for Disease Control and Prevention (CDC),FoodNet,1999c. http://www.cdc.gov/ncidod/dbmd/foodnet/98surv.htm

Centers for Disease Control and Prevention (CDC),Outbreak of Escherichia coli O157:H7 andCampylobacter among attendees of the WashingtonCounty Fair – New York,1999, Morbidity andMortality Weekly Report, 48(36),803,1999a.

Centers for Disease Control and Prevention (CDC),PulseNet,1999b.http://www.cdc.gov/ncidod/dbmd/pulsenet/pulsenet.htm

Childs, J. E.,B.S.Schwartz,T. G. Ksiazek,R.R.Graham, J.W. LeDuc,and G. E.Glass,Risk factors associated withantibodies to leptospires in inner-city residents ofBaltimore:A protective role for cats, American Journalof Public Health, 82(4),597-9,1992.

Clarke, J. F.,Some effects of the urban structure on heatmortality, Environmental Research, 5, 93-104, 1972.

Colwell,R.R.,and J.A. Patz, Climate, Infectious Diseaseand Health:An Interdisciplinary PerspectiveAmerican Academy of Microbiology, Washington,DC,pp.24,1998.

Coye,M. J.,and M.Goldoft,Microbiological contamina-tion of the ocean,and human health, New JerseyMedicine, 86(7),533-8,1989.

Craun, G. F., Waterborne Disease in the United States,CRC Press,Boca Raton,Florida,pp.295,,1998.

Curriero, F., J. Patz, J. Rose,and S.Lele,Analysis of theassociation between extreme precipitation and water-borne disease outbreaks in the United States:1948-1994, American Journal of Public Health, in press,2001.

Davies,C. J.,and J. Mazurek, Pollution Control in theUnited States: Evaluating the system,Resources for theFuture,Washington,DC,pp.366,1998.

455


Day, J. C., Population projections of the United States byage,sex, race and Hispanic origin:1995-2050,Currentpopulation reports provided by the US Bureau of theCensus,US Government Printing Office,Washington,DC,1996.

Demers,R.Y.,A.Thiermann, P. Demers,and R. Frank,Exposure to Leptospira icterohaemorrhagiae in inner-city and suburban children:A serologic comparison,Journal of Family Practice, 17(6),1007-11,1983.

Di Maio, D. J.,and V. J. M.Di Maio, The Effects of Heatand Cold:Hyperthermia and hypothermia, ForensicPathology, CRC Press,Boca Raton,Florida,pp.503,1993.

Drabek,T. E., The Social Dimensions of Disaster,Federal Emergency Management Agency (EmergencyManagement Institute),Emmitsburg,Maryland,619 pp.,1996.

Easterling, D. R.,et al.,Maximum and minimum temper-ature trends for the globe, Science, 277, 364-367,1997.

Engelthaler, D. M.,et al.,Climatic and environmentalpatterns associated with Hantavirus pulmonary syn-drome, Four Corners region,United States, EmergingInfectious Diseases, 5(1),87-94,1999.

Federal Emergency Management Agency (FEMA),National mitigation strategy: Partnerships for buildingsafer communities,1996,http://www.fema.gov/mit/ntmstrat.htm

Fowler,A.M.,and K. J. Hennessey, Potential impacts ofglobal warming on the frequency and magnitude ofheavy precipitation, Natural Hazards, 11, 283-303,1995.

Frost, F. J., G. F. Craun,and R.L.Calderon,Waterbornedisease surveillance, Journal of the American WaterWorks Association, 88(9),66-75,1996.

Gerba,C. P., J. B.Rose,and C.N Haas,Sensitive popula-tions:Who is at the greatest risk? InternationalJournal of Food Microbiology, 30(1-2),113-23,1996.

Gill, J.,L.M.Stark,and G. C.Clark,Dengue surveillancein Florida,1997-1998, Emerging Infectious Diseases,6(1),in press,2000.

Gilles,H.M.,Epidemiology of Malaria,in Bruce-Chwatt’s Essential Malariology edited by H.M.Gillesand D.A.Warrell,Edward Arnold Division of Hodder &Stoughton,London,England,124-63,1993.

Glass G.E.,Cheek J.E., Patz J.A.,Shields T.M.,Doyle T.J.,Thoroughman D.A.,Hunt D.K.,Enscore R.E.,Gage K.L.,Irland C., Peters C.J.,Bryan R. Using remotely senseddata to identify areas at risk for hantavirus pulmonarysyndrome, Emerging Infectious Diseases, 6(3):238-47.

Glickman,T. S.,and E. D. Silverman, Acts of God andActs of Man. Resources for the Future,,CRM 92-02,Washington,DC,1992.

Gorjanc,M.L.,W. D. Flanders, J.VanDerslice, J. Hersh,and J. Malilay, Effects of temperature and snowfall onmortality in Pennsylvania, American Journal ofEpidemiology, 149, 1152-60,1999.

Gower, J.F.R..Detection and mapping of bright plank-ton blooms and river plumes using AVHRR imagery, inThird Thematic Conference, Remote Sensing forMarine and Coastal Environments, Seattle,Washington,pp.151-62,1995.

Gubler, D. J.,Resurgent vector-borne diseases as a glob-al health problem, Emerging Infectious Diseases, 4(3),442-50,1998.

Harvell,C. D.,et al.,Emerging marine diseases—climatelinks and anthropogenic factors, Science, 285(5433),1505-1510,1999.

Hobbs, F. B.,and B.L.Damon,65+ in the United States,Bureau of the Census,US Government Printing Office,Washington,DC,1998.

Houghton, J., Global Warming:The Complete Briefing,Cambridge University Press,Cambridge, UnitedKingdom,1997.

Hoxie, N. J., J. P. Davis, J. M.Vergeront,R. D. Nashold,andK.A.Blair, Cryptosporidiosis - associated mortality fol-lowing a massive waterborne outbreak in Milwaukee,Wisconsin, American Journal of Public Health,87(12),2032-35,1997.

Institute of Medicine (IOM),Emerging infections:Microbial threats to health in the United States,Institute of Medicine, Washington,DC,1992. Johnston,J. M.,S. F. Becker, and L.M.McFarland,Vibrio vulnificus.Man and the sea, JAMA, 253(19),2850-2853,1985.

Jones, P. D., D. M.New, D. E. Parker, S.Martin,and I. G.Rigor, Surface air temperature and its changes over thepast 150 years, Review of Geophysics, 37, 173-200,1999.

Jones,T. S.,A. P. Liang,E.M.Kilbourne,M.R.Griffin, P.A.Patriarca,S. G. Fite-Wassilak,R. J. Mullan,R. F. Herrick,H.D. Donnell,C. Keewhan,et al.et al.,Morbidity and mor-tality associated with the July 1980 heat wave in St.Louis and Kansas City, Missouri, JAMA, 247, 3327-3331,1982.

Kalkstein,L.S.,and J. S.Greene,An evaluation of cli-mate/mortality relationships in large US cities and thepossible impacts of climate change, EnvironmentalHealth Perspectives, 105(1),84-93,1997.

456


Kalkstein,L.S.,and K.E.Smoyer,The impact of climatechange on human health:Some international implica-tions, Experientia, 49(11),969-79,1993.

Karl,T. R.,and R.W. Knight,Secular trends of precipita-tion amount,frequency, and intensity in the USA,Bulletin of the American Meteorological Society, 79,231-241,1998.

Karl,T. R.,R.W. Knight, D. R.Easterling,and R. G. Quayle,Indices of climate change for the United States,Bulletin of the American Meteorological Society, 77,279-303,1996.

Karl,T. R.,R.W. Knight,and N. Plummer,Trends in high-frequency climate variability in the twentieth century,Nature, 377(6546),217-20,1995.

Kilbourne,E.M.,K.Choi,T. S. Jones,S.B.Thacker, and F.I.Team,Risk factors for heat stroke:A case-controlstudy, JAMA, 247, 3332-3336,1982.

Kirk-Davidoff, D.B.,E. J. Hintsa, J. G.Anderson,and D.W.Keith,The ef fect of climate change on ozone depletionthrough changes in stratospheric water vapour, Nature,402, 399-401,1999.

Knutson,T. R.,and R.E.Tuleya,Increased hurricaneintensities with CO 2 -induced global warming as simu-lated using the GFDL hur ricane prediction system,Climate Dynamics, 15, 503-19,1999.

Ko,A.I.,M.Galvao Reis,C.M.Ribeiro Dourado,WD, Jr.,Johnson,and L .W. Riley, Urban epidemic of severe lep-tospirosis in Brazil,Salvador Leptospirosis Study Group,Lancet, 354(9181), 820-825,1999.

Lambert,W. E., J. M.Samet,and D.W. Dockery,Community air pollution,in Environmental andOccupational Medicine, edited by W. N. Rome,pp.1501-1522,Lippincott-Raven,Philadelpia, Pennsylvania,1998.

Lanciotti,R.S.,et al.,Origin of the West Nile virusresponsible for an outbreak of encephalitis in thenortheastern United States, Science, 286(5448),2333-2337,1999.

Landrigan, P. J.,W. Suk,and R.W.Amler, Chemical wastes,children’s health,and the Superfund basic research pro-gram, Environmental Health Perspectives, 107(6),423-7,1999.

Landsberg,H.E., The Urban Climate, Academic Press,New York,1981.

Lipp,E.K.,and J. B.Rose,The role of seafood in food-borne diseases in the United States of America, RevueScientifique et Technique, 16: 620-40,1997.

Lipp,E.K., J. B.Rose,R.Vincent,R. Kurz,and C.Rodriguez-Palacios,Assessment of the microbiologicalwater quality in Charlotte Harbor, Florida,Technicalreport to the Southwest Florida water management dis-trict,surface water improvement,and managementplan,Tampa:Southwest Florida Water ManagementDistrict,1999.

Logue J. N.,H.Hansen,and E.Struening,Emotional andphysical distress following Hurricane Agnes inWyoming Valley of Pennsylvania, Public Health Report94, 495-502,1979.

McKenzie,R.,B.Connor, and G. Bodeker, Increased sum-mertime UV radiation in New Zealand in response toozone loss, Science, 285, 1709-1711,1999 .

MacKenzie,W. R.,et al.,A massive outbreak inMilwaukee of cryptosporidium infection transmittedthrough the public water supply, New EnglandJournal of Medicine, 331(3),161-167,1994.

Marzuk, P. M.,K.Tardiff,A.C.Leon,C.S.Hirsch,L.Portera,M.I.Iqbal,M.K.Nock,and N. Hartwell,Ambienttemperature and mortality from unintentional cocaineoverdose, JAMA, 279(22),1795-800,1998.

McMichael,A. J.,A.Haines,R.Sloof, and S. Kovats,Climate Change and Human Health, World HealthOrganization,Geneva,297,1996.

Mead, P. S.,L.Slutsker,V. Dietz,L. F. McCaig, J. S.Bresee,C.Shapiro, P. M.Griffen ,and R.V.Tauxe, Food-related ill-ness and death in the United States, EmergingInfectious Diseases 5(5),607-25,1999.

Mearns,L. O., F. Giorgi,L.McDaniel,and C.Shields,Analysis of daily variability of precipitation in a nestedregional climate model:Comparison with observationsand doubled CO2 results, Global and PlanetaryChange, 10, 55-78,1995.

Meehl, G.A., F. Zwiers, J. Evans,T. Knutson,L. O. Mearns,and P.Whetton,Trends in extreme weather and climateevents:Issues related to modeling extremes in projec-tions of future climate change, Bulletin of theAmerican Meteorological Society, 81, 427-436,2000.

Monath,T. P.,Epidemiology, in St.Louis encephalitis,edited by T. P. Monath,American Public HealthAssociation,Washington,DC,1980.

Morris,R.E.,M.S.Gery, M.K.Liu, G. E.Moore,C.Daly,and S.M.Greenfield,Sensitivity of a regional oxidantmodel to variations in climate parameters,in ThePotential Effects of Global Climate Change on theUnited States, edited by J. B.Smith and D.A.Tirpak,USEPA,Of fice of Policy, Planning and Evaluation,Washington,DC,1989.

Motes,M.L.,A.DePaola, D.W. Cook, J. E.Veazey, J. C.Hunsucker,W. E.Garthright,R. J. Blodgett,and S. J.Chirtel,Influence of water temperature and salinity onVibrio vulnificus in Northern Gulf and Atlantic Coastoysters (Crassostrea virginica), AppliedEnvironmental Microbiology, 64(4),1459-65,1998.

National Center for Health Statistics (NCHS),Health,United States,1998 with socioeconomic status andhealth chartbook,US Department of Health and HumanServices,Hyattsville,Maryland,pp.465,1998.National Research Council (NRC), Rethinking theOzone Problem in Urban and Regional Air Pollution,National Academy Press,Washington,DC,1991.

National Research Council (NRC),ReconcilingObservations of Global Temperature Change, NationalAcademy Press,Washington,DC,pp.86,2000.

National Weather Service (NWS),Summary of naturalhazard deaths for 1991 in the US,National WeatherService,Rockville,Maryland,1992.21

National Weather Service (NWS),Hurricane Andrew:South Florida and Louisiana, August 23-26,1992,Naturaldisaster survey repor, US Dept.of Commerce,NationalOceanic and Atmospheric Administration,NationalWeather Service,Silver Spring,Maryland,1993.

National Weather Service (NWS),Summary of naturalhazard statistics,1999.http://www.nws.noaa.gov/om/hazstats.htm

New York Department of Health,Health Commissionerreleases E.coli outbreak report,March 31 2000.http://www.health.state.ny.us/nysdoh/commish/2000/ecoli.htm

Noji,E.K.,The nature of disaster:General characteris-tics and public health ef fects,in The Public HealthConsequences of Disasters, edited by E.K.Noji,pp.3-20,Oxford University Press,New York,1997.31

Norris, F. H., J. L. Perilla, J. K.Riad,K.Kaniasty, and E.Lavizzo,Stability and change in stress, resources,andpsychological distress following natural disaster:Findings from a longtitudinal study of HurricaneAndrew, Anxiety, Stress and Coping, 12,363-96,1999.

Parmenter R.R.,E.P.Yadav, C.A. Parmenter, P. Ettestad,K.L.Gage.Incidence of plague associated withincreased winter-spring precipitation in New Mexico.American Journal of Tropical Medicine and Hygiene61:814-21,1999.

Patz, J.A.,M.A.McGeehin,S.M.Bernard,K.L.Ebi, P.R.Epstein,A.Grambsch, D.J. Gubler, P. Reiter, I Romieu,J.B.Rose, J.M.Samet, J.Trtanj,The Potential HealthImpacts of Climate Variability and Change in theUnited States:Executive Summary of the Report of theHealth Sector of the U.S. National Assessment,Environmental Health Perspectives, 108(4),367-376,2000.

Penner, J. E., P. S.Connell, D. J.Wuebbles,and C.C.Covey, Climate change and its interactions with airchemistry: Perspective and research needs,in Thepotential effects of global climate change on theUnited States, edited by J. B.Smith and D.A.Tirpak,USEPA,Office of Policy, Planning and Evaluation,Washington,DC,1989.

Philip,C.B.,and L.E.Rozeboom,Medico-veterinaryentomology:A generation of progress,in History ofEntomology, edited by R. F. Smith,T. E.Mittler, and C. N.Smith,Annual Reviews,Inc., Palo Alto,California,1973.

Pinner, R.W.,S.M.Teutsch,L.Simonsen,L.A.Klug, J. M.Graber, M. J. Clarke,and R.L.Berkelman,Trends in infec-tious diseases mortality in the United States, JAMA,275(3),189-93,1996.

Quarantelli,E.L.,An assessment of conflicting views onmental health:The consequences of traumatic events,in Trauma and Its Wake:The Study and Treatment ofPost-Traumatic Stress Disorder, vol 4 (Figley CR,ed).Brunner/Mazel,New York,173-215,1985.

Ramlow, J. M.,and L.H. Kuller, Effects of the summerheat wave of 1988 on daily mortality in AlleghenyCounty, Pennsylvania, Public Health Report 105, 283-9,1990.

Reeves,W. C.,and W. M.Hammon,Epidemiology of thearthropod-borne viral encephalitides in Kern County,California,1943-1952,University of California atBerkeley Public Health Report 4,1-257,1962.

Reeves,W. C., J. L.Hardy,W. K.Reisen,and M.M.Wilby,Potential effect of global warming on mosquito-bornearboviruses, Journal of Medical Entomology, 31(3),323-32,1994.

Reisen,W. K.,H. D. Lothrop,and J. L.Hardy, Bionomicsof Culex tarsalis (Diptera:Culicidae) in relation toarbovirus transmission in southeastern California,Journal of Medical Entomology 32(3), 316-27,1995.

Reiter, P.,Weather, vector biology, and arboviral recrude-scence,in The Arboviruses:Epidemiology and Ecology,edited by T. P. Monath,pp.245-255,CRC Press,BocaRaton,Florida,1988.

Reiter, P.,Global warming and mosquito-borne diseasein USA, Lancet, 348(9027),622,1996.

Reiter, P.,Global climate change and mosquito-bornedisease, Environmental Health Perspectives, 109(sup-plement 1),in press, Feb.2001.

Reiter, P., From Shakespeare to Defoe:Malaria inEngland in the Little Ice Age, Emerging InfectiousDiseases, 6, 1-11,2000.

457


Trenberth,K.E.,Conceptual framework for changes ofextremes of the hydrologic cycle with climate change,Climatic Change, 42, 327-39,1999.

Trevejo,R.T.,et al.,Epidemic leptospirosis associatedwith pulmonary hemorrha ge - Nicaragua,1995,Journal of Infectious Diseases, 178(5),1457-63,1998.

United States Environmental Protection Agency(USEPA),Air quality criteria for ozone and related pho-tochemical oxidant:Volume I of III,USEPA,Office ofResearch and Development,Washington,DC,1996.

United States Environmental Protection Agency(USEPA),National air pollutant emission trends update:1970-1996,USEPA, Washington,DC,1997a.

United States Environmental Protection Agency(USEPA),Water on tap:A consumer’s guide to thenation’s drinking water, USEPA,Office of Water,Washington,DC,1997b.

United States Environmental Protection Agency(USEPA),National air quality and emissions trendsreport,1997,USEPA,Office of Air Quality Planning andStandards,Washington,DC,1998a.

United States Environmental Protection Agency(USEPA), National Primary Drinking WaterRegulations:Interim Enhanced Surface WaterTreatment; Final Rule, USEPA,Washington,DC,1998b.

United States Environmental Protection Agency(USEPA),Action plan for beaches and recreationalwaters,USEPA,Office of Research and Development,Office of Water,Washington,DC,1999.

Valiela,I., Marine Ecological Processes, Springer-Verlag,New York,1984.

Watts, D. M., D. S.Burke,B.A.Harrison,R.E.Whitmire,and A.Nisalak,Effect of temperature on the vector effi-ciency of Aedes aegypti for Dengue 2 virus, AmericanJournal of Tropical Medicine and Hygiene, 36(1),143-52,1987.

Weniger, B. G.,M J. Blaser, J. Gedrose,E.C.Lippy, and D.D. Juranek,An outbreak of waterborne giardiasis associ-ated with heavy water runoff due to warm weatherand volcanic ashfall, American Journal of PublicHealth, 73(8),868-72,1983.

Whitman,S., G. Good,E.R.Donoghue, N. Benbow,W.Shou,and S.Mou,Mortality in Chicago attributed tothe July 1995 heat wave, American Journal of PublicHealth, 87, 9,1515-8,1997.

Yates,M.L.,and S.R.Yates,Modeling microbial fate inthe subsurface environment, Critical Reviews inEnvironmental Control, 17(4),307-44,1988.

Robinson, P.,The ef fects of climate change,in GlobalClimate Change Linkages:Acid Rain,Air Quality,and Stratospheric Ozone, edited by J. C.White,Elsevier, New York,1989.

Romieu,I.,Epidemiological studies of the health effectsarising from motor vehicle air pollution,in UrbanTraffic Pollution, edited by D. Schwela and O. Zali,pp.10-69,World Health Organization,New York,1999.

Rose, J. B.,and J. Simonds,King County water qualityassessment:Assessment of public health impacts associ-ated with pathogens and combined sewer overflows,Seattle,Washington,Report for Water and LandResources Division.Department of Natural Resources,1998.

Schmaljohn,C.,and B.Hjelle,Hantaviruses:A global dis-ease problem, Emerging Infectious Diseases 3(2),95-104,1997.

Schuman,S.H., Patterns of urban heat-wave deaths andimplications for prevention:Data from New York andSt.Louis during 1966, Environmental Research, 5, 59-75,1972.

Semenza, J. C.,Are electronic emergency departmentdata predictive of heat-related mortality? Journal ofMedical Systems, 23(5),419-424,1999.

Semenza, J. C., J. McCullough, D.W. Flanders,M.A.McGeehin,and J. R.Lumpkin,Excess hospital admis-sions during the 1995 heat wave in Chicago, AmericanJournal of Preventive Medicine, 16, 269-77,1999.

Semenza, J. C.,C.H.Rubin,and K.H. Falter, Heat-relateddeaths during the July 1995 heat wave in Chicago, NewEngland Journal of Medicine, 335(2),84-90,1996.

Shapiro,R.L.,S.Altekruse,and P. M.Griffin,The role ofGulf Coast oysters harvested in warmer months inVibrio vulnificus infections in the United States,1988-1996, Journal of Infectious Diseases, 178(3),752,1998.

Shindell, D.T., D. Rind,and P. Lonergan,Increased polarstratospheric ozone losses and delayed eventual recov-ery owing to increasing greenhouse-gas concentra-tions, Nature, 392, 589-592,1998.

Sillman,S.,and P. J. Samson,Impact of temperature onoxidant photochemistry in urban,polluted, rural,andremote environments, Journal of GeophysicalResearch, 100, 11497-508,1995.

Spitalnic,S. J.,L. Jagminas,and J. Cox,An associationbetween snowfall and ED presentation of cardiacarrest, American Journal of Emergency Medicine,14(6),572-3,1996.

Tauxe,R.V.,Emerging foodborne diseases:An evolvingpublic health challenge, Emerging Infectious Diseases,3(4),425-34,1997.458


461

CHAPTER 16

POTENTIAL CONSEQUENCES OF CLIMATEVARIABILITY AND CHANGE ON COASTALAREAS AND MARINE RESOURCESJohn C. Field1, Donald F. Boesch2, Donald Scavia1, Robert Buddemeier3,Virginia R.Burkett4,5, Daniel Cayan6, Michael Fogarty2, Mark Harwell7, Robert Howarth8, CurtMason1, Leonard J. Pietrafesa10, Denise Reed11,Thomas Royer12,Asbury Sallenger4,Michael Spranger13, James G.Titus14


Chapter Summary

Introduction

Context

Climate and Coastal Environments

Key Issues

Shoreline Systems,Erosion and Developed Coastal Areas

Threats to Estuarine Health

Coastal Wetland Survival

Coral Reef Die-Offs

Stresses on Ocean Margins and Marine Fisheries



Literature Cited

Acknowledgments

1National Oceanic and Atmospheric Administration,2University of Maryland, 3University of Kansas, 4U.S.Geological Survey, 5Coordinating author for the National Assessment Synthesis Team, 6Scripps Institution ofOceanography, 7University of Miami, 8Eco-Modeling, 10North Carolina State University, 11University of NewOrleans, 12Old Dominion University, 13University of Washington, 14 U.S.Environmental Protection Agency


• Sea level has risen by 4 to 8 inches (10-20 cm) inthe past century.

• Ocean temperatures have risen over the last 55years,and there has been a recent increase in thefrequency of extreme ocean warming events inmany tropical seas.

• Sea ice over large areas of the Arctic basin hasthinned by 3 to 6 feet (1 to 2 meters),losing 40%of its total thickness since the 1960s;it continuesto thin by about 4 inches (10 cm) per year.

• Marine populations and ecosystems have beenhighly responsive to climate variability.


• Sea level is projected to rise an additional 19inches (48 cm) by 2100 (with a possible range of5 to 37 inches [13 cm to 95 cm]) along most ofthe US coastline.

• Ocean temperatures will continue to rise,but therate of increase is likely to lag behind tempera-ture changes observed on land.

• The extent and thickness of Arctic sea ice isexpected to continue to decline. All climatemodels project large continued losses of sea ice,with year-round ice disappearing completely inthe Canadian model by 2100.

• Increased temperature or decreased salinitycould trigger abrupt changes in thermohalineocean circulation.

CHAPTER SUMMARY

Context

The US has over 95,000 miles of coastline andapproximately 3.4 million square miles of oceanwithin its territorial sea,all of which provide a widerange of essential goods and services to human sys-tems. Coastal and marine ecosystems supportdiverse and important fisheries throughout thenation’s waters,hold vast storehouses of biologicaldiversity, and provide unparalleled recreationalopportunities. Some 53% of the total US populationlives on the 17% of land in the coastal zone,andthese areas become more crowded every year.Demands on coastal and marine resources are rapid-ly increasing,and as coastal areas become moredeveloped,the vulnerability of human settlementsto hurricanes,storm surges,and flooding events alsoincreases. Coastal and marine environments areintrinsically linked to climate in many ways. Theocean is an important distributor of the planet’sheat,and this distribution could be strongly influ-enced by changes in global climate. Sea-level rise isprojected to accelerate in the 21st century, with dra-matic impacts in those regions where subsidenceand erosion problems already exist.


462

Chapter 16 / Coastal Areas and Marine Resources

463

Key Findings

• Climate change will increase the stresses alreadyoccurring to coastal and marine resources as aresult of increasing coastal populations,develop-ment pressure and habitat loss, overfishing,excess nutrient enrichment,pollution,and inva-sive species.

• Marine biodiversity will be further threatened bythe myriad of impacts to all marine ecosystems,from tropical coral reefs to polar ecosystems.

• Coral reefs are already under severe stress fromhuman activities,and have experienced unprece-dented increases in the extent of coral bleaching,emergent coral diseases,and widespread die-offs.

• The direct impact of increasing atmospheric car-bon dioxide on ocean chemistry will possiblyseverely inhibit the ability of coral reefs to growand persist in the future.

• Globally averaged sea level will continue to rise,and the developed nature of many coastlinesmakes both human settlements and ecosystemsmore vulnerable to flooding and inundation.

• Barrier islands are especially vulnerable to thecombined effects of sea-level rise and uncon-trolled development that hinders or prevents nat-ural migration.

• Ultimately, choices will have to be made betweenthe protection of human settlements and theprotection of coastal ecosystems such as beach-es,barrier islands,and coastal wetlands.

• Human development and habitat alteration willlimit the ability of coastal wetlands to migrateinland as sea levels rise,however, if sedimentsupplies are adequate,many wetlands may sur-vive through vertical adjustments.

• Increases in precipitation and runoff are likely tointensify stresses on estuaries in some regions byintensifying the transport of nutrients and con-taminants to coastal ecosystems.

• As rivers and streams also deliver sediments,which provide material for soil in wetlands andsand in beaches and shorelines,dramatic declinesin streamflows could have negative effects onthese systems.

• Changes in ocean temperatures,currents,andproductivity will affect the distribution, abun-dance,and productivity of marine populations,with unpredictable consequences to marineecosystems and fisheries.

• Increasing carbon dioxide levels could triggerabrupt changes in thermohaline ocean circula-tion,with massive and severe consequences forthe oceans and for global climate.

• Extreme and ongoing declines in the thicknessand extent of Arctic sea ice will have enormousconsequences for Arctic populations,ecosystems,and coastal evolution.


INTRODUCTIONHuman caused alterations and impacts to coastaland marine environments must be considered in thecontext of evaluating the health and viability ofthese resources. In many instances,it is difficult toassess the effects of climate variability and changesbecause human activities are often responsible forthe greatest impacts on coastal and marine environ-ments. Such disturbances often reduce the capacityof systems to adapt to climate variations and climatestress,and often mask the physical and biologicalresponses of many systems to climate forces. Forexample,naturally functioning estuarine and coastalwetland environments would typically be expectedto migrate inland in response to relative sea-levelrise,as they have responded to sea-level variationsthroughout time. When this natural migration isblocked by coastal development,such habitat isgradually lost by “coastal squeeze”as rising sea levelspush the remaining habitat against developed orotherwise altered landscapes. Similarly, estuariesalready degraded by excess nutrients could recovermore slowly from droughts or floods.

CONTEXTThe US has over 95,000 miles of coastline andapproximately 3.4 million square miles of oceanwithin its territorial sea,all of which provide a widerange of essential goods and services to society.These include overlapping and often competinguses,including but not limited to tourism,coastaldevelopment,commercial and recreational fisheries,aquaculture,biodiversity, marine biotechnology, navi-gation,and mineral resources.

The coastal population of the US is currently grow-ing faster than the nation’s population as a whole, atrend that is projected to continue. Currently some53% of the total population of the US live in 17% ofthe land area considered coastal (Culliton,1998).Over the next 25 years,population gains of approxi-mately 18 million people are projected to occur inthe coastal states of Florida,California,Texas,andWashington alone (Figure 1) (NPA,1999). With this

growth,as well as increased wealth andaffluence,there are rapidly increas-ing demands on coastal and marineresources for both aesthetic enjoy-ment and economic benefits.

This large and growing populationpressure in coastal areas is responsi -ble for many of the current stresses tocoastal resources. For example,theEPA (1996) estimated that nearly 40%of the nation’s surveyed estuarieswere impaired by some form of pollu-tion or habitat degradation. Some 30to 40% of shellfish-growing waters inthe nation’s estuaries are harvest pro-hibited or restricted each year, primari-

ly due to bacterial contaminationfrom urban and agricultural

runoff and septic systems(Alexander, 1998). Additionally,

over 3,500 beach advisories andbeach closings occurred in the

United States in 1995,primarily due to storm-waterrunoff and sewage overflows (NOAA,1998).

POTENTIAL CONSEQUENCES OF CLIMATEVARIABILITY AND CHANGE ON COASTAL AREAS AND MARINE RESOURCES

Population Distribution across the US

464

Figure 1: Over the next 25 years, population gains of some 18 mil-lion people are projected to occur in the coastal states of Florida,California, Texas, and Washington (NPA, 1999). See color figureappendix.


465

in the US, representing the second largest employerin the nation (after health care) and employingsome 6 million people (NOAA,1998). It has beenestimated that the US receives over 45% of thedeveloped world’s travel and tourism revenues,andoceans,bays,and beaches are among the most popu-lar tourist destinations in the nation (Houston,1996). As many as 180 million people visit the coasteach year for recreational purposes in all regions ofthe country, and many regions depend upon tourismas a key economic activity. For example one studyestimated that in the San Francisco Bay area alonetourism has been estimated to generate over $4 bil-lion a year (EPA,1997). Clean water, healthy ecosys-tems,and access to coastal areas are critical to main-taining tourism industries;ironically, however, theseindustries themselves often pose additional impactsto coastal environments and local communities(Miller and Auyong,1991).

As coastal populations increase,the vulnerability ofdeveloped coastal areas to natural hazards is simulta-neously expanding. Disaster losses are currentlyestimated at about $50 billion annually in the US,compared to just under $4.5 billion in 1970. Asmuch as 80% of these disasters were meteorological-ly related storms,hurricanes,and tornadoes (asopposed to geologically related disasters such asearthquakes and volcanoes) and many of these hadtheir greatest impacts on coastal communities. Thepotential increase in such events related to climatechange will pose an even greater threat to coastalpopulation centers and development in the future.However, the ongoing trends of population growthand development in coastal areas alone will ensurethat losses due to hurricanes,storms,and other dis-asters in coastal areas will continue to increase.

In addition to direct economic benefits,coastal andmarine ecosystems,like all ecosystems,have charac-teristic properties or processes which directly orindirectly benefit human populations. Costanza etal.(1997) have attempted to estimate the economicvalue of sixteen biomes,or ecosystem types,andseventeen of their key goods and services,includingnutrient cycling,disturbance regulation, waste treat-ment, food production, raw materials, refugia forcommercially and recreationally important species,genetic resources,and opportunities for recreationaland cultural activities. For example,the societalvalue per hectare of estuaries,tidal marshes,coralreefs,and coastal oceans were estimated at $22,832,$9,990,$6,075,and $4,052 per hectare, respectively.On a global basis,the authors suggested that theseenvironments were of a disproportionately higher

Population pressures from further inland can alsohave detrimental impacts on coastal resources.Effluent discharges as well as agricultural runoffhave caused significant nutrient over-enrichment inmany coastal areas. Sewage and siltation are signif i-cant contributors to coral reef degradation inHawaii,Florida,and US-affiliated islands of thePacific and Caribbean. Dams,irrigation projects,andother water management activities have furtherimpacted coastal ecosystems and shorelines bydiverting or otherwise altering the timing and flowof water, sediments,and nutrients.

As population and development in coastal areasincrease,many of these stresses can be expected toincrease as well,further decreasing the resilience ofcoastal systems. This,in turn,increases the vulnera-bility of coastal communities and economies whichdepend upon healthy, functioning ecosystems. Theinteraction of these ongoing stresses with theexpected impacts imposed by future climate changeis likely to greatly accentuate the detrimentalimpacts to coastal ecosystems and communities(Reid and Trexler, 1992,Mathews-Amos andBernston,1999).

Despite these ongoing stresses to coastal environ-ments,the oceans and coastal margins provideunparalleled economic opportunities and revenues.One estimate suggests that as many as one out ofevery six jobs in the US is marine-related,and nearlyone-third of the gross domestic product (GDP) isproduced in coastal areas (NOAA,1998;NRC,1997).In 1996,approximately $590 billion worth of goodspassed through US ports, over 40% of the total valueof US trade and a much larger percentage by vol-ume. The US is also the world’s fifth largest fishingnation and the third largest seafood exporter;totallandings of marine stocks have averaged about 4.5million metric tons over the last decade. Ex-vesselvalue (the amount that fishermen are paid for theircatch) of commercial fisheries alone was estimatedat approximately $3.5 billion in 1997,and the total(direct and indirect) economic contribution ofrecreational and commercial fishing has been esti-mated at over $40 billion per year (NRC,1999). Thegrowing field of marine biotechnology has also gen-erated substantial opportunities; for example recentresearch has yielded five drugs originating frommarine organisms with a cumulative total potentialmarket value of over $2 billion annually (NOAA,1998).

Coastal tourism also generates enormous revenues.Travel and tourism are multi-billion dollar industries


466

value,covering only some 6.3% of the world’s sur-face area but responsible for some 43% of the esti-mated value of the world’s ecosystem services.These results suggest that the oceans and coastalareas contribute the equivalent of some $21 trillionper year to human activities globally (Costanza,1999). The approach of Costanza et al.(1997) to val-uation is not universally accepted,and the authorsthemselves agree that ecosystem valuation is diffi-cult and fraught with uncertainties. However, themagnitude of their estimates,and the degree towhich coastal and marine ecosystems rank asamongst the most valuable to society, serve to placethe importance of the services and functions ofthese ecosystems in an economic context.

CLIMATE AND COASTALENVIRONMENTSSea-level Change

Global sea levels have been rising since the conclu-sion of the last ice age approximately 15,000 yearsago. During the last 100 years,globally averaged sealevel has risen approximately 4 to 8 inches (10-20cm,or about 1 to 2 millimeters per year). This rep-resents “eustatic”sea-level change,the change in ele-vation of the Earth’s oceans that has been deter-mined from tidal stations around the globe. Most of

Figure 2 : Projected rise in global average sea level based on theHadley and Canadian General Circulation Model (GCM) scenarios.See color figure appendix.

Global Average Sea Level Rise

Spatial Distribution Around North America in Sea Level Rise

Figure 3 : Projections of the regional pattern of global sea level rise by the year 2100 based on the Canadian(left) and Hadley (right) scenarios. These estimates do not include contributions to sea-level change due to verti-cal movement of coastal lands. See color figure appendix.


467

the observed global sea-level change is accountedfor by two major variables:the thermal expansion ofseawater in the oceans with rising ocean tempera-tures and changes in the amount of the Earth’swater that is locked up in glaciers and ice sheets.The vast majority of landlocked water, enough toraise global sea levels by some 80 meters,is found inthe Greenland and Antarctic Ice sheets. Of these,there has been considerable concern regarding thestability of the West Antarctic Ice Sheet (WAIS).Oppenheimer (1998) suggested that the probabilityof mass wasting (melting) of this ice sheet over thenext 100 years is relatively low, but the probabilityof wasting after 2100 will be considerably greater.Over the next 100 years,most studies predict thatthe majority of observed sea-level change is expect-ed to come as a result of thermal expansion of theoceans (Gornitz,1995).

In addition to global changes in mean sea level,there are large regional variations in sea level asmeasured at the shoreline due to changes in coastalland masses such as subsidence (sinking),isostatic(glacial) rebound and tectonic uplift. The combinedeffects of eustatic and land mass factors contributeto relative sea-level change in each locality. Forexample,within the US,portions of the Gulf Coastare experiencing a relative sea-level rise of nearlyhalf an inch (approximately 10 mm) per year due tosubsidence. Concurrently, some portions of thesoutheastern Alaska coastline are experiencing upliftassociated with tectonic activity, causing a local rela-tive sea-level fall of slightly less than one half inch(approximately 8 mm) per year.

Finally, there are also regional changes in mean sealevel that result from dynamic changes to the oceangeoid,or the “topography”of the sea surface. Theseresult from changes in ocean circulation,wind andpressure patterns,and ocean-water density (IPCC,1996). The result is that sea-level rise will affect var-ious coastal regions differently, independent of theland motion that contributes to sea-level change.Figures 2 and 3 show the results of Hadley andCanadian General Circulation Model (GCM) projec-tions for regional changes in sea level,independentof land movement. In general,the Hadley modelpredicts a greater sea-level rise for the Pacific coastthan for the Atlantic and Gulf coasts as a result ofcurrent and wind patterns. By contrast,theCanadian model predicts a more complex pattern ofsea-level rise but with increases relatively similaralong all US coasts.

The Hadley climate model projects that sea levelwill rise between 8 and 12 inches (20 to 30 cen-

timeters) by 2100 for the Atlantic and Gulf Coasts,and 13 to 16 inches (33 cm to 41 cm) for the PacificCoast (Figure 3). The Canadian model projects a sig-nificantly greater estimate of 20 to 24 inches (51 cmto 61 cm) along parts of the US coast (Figure 3).These results are very similar to those found by theIntergovernmental Panel on Climate Change (IPCC,1996),which estimated that sea levels would mostlikely increase by approximately 19 inches (48 cm)by 2100,with a range of 5 to 37 inches (13 to 95cm). Most additional studies conducted since thenyield similar estimates, generally projecting increasesof about one foot above current trends over the 21st

century, for a total sea-level rise of approximately 18to 20 inches (45 to 52 cm) above their current levelby the year 2100 (Titus and Narayanan,1996;Wigley,1999). In addition,sea-level rise will continue tooccur, even accelerate,beyond 2100 as a result ofthe long time frame necessary for oceans and icesheets to approach equilibrium under the long-termperturbations anticipated with climate change.

Hurricanes and Non-TropicalStorms

Storm flooding, wave forces,and coastal erosion arenatural processes that pose hazards only when theyaffect people,homes,and infrastructure that areconcentrated in coastal areas. During the 20th cen-tury, loss of life and threats to human health duringhurricanes has decreased substantially because ofimproved tracking and early warning systems;how-ever, property losses have increased greatly (Herbertet al.,1996). Yet even if storm intensity and frequen-cy remain the same in the future,continued acceler-ation in property losses is likely as the increasedconcentration of people and infrastructure alongthe coasts continues. This acceleration in propertylosses will be amplified should global climatechange increase storm activity, although the currentrelationship between climate change and hurricanefrequency and intensity is not clear.

Historical records of hurricanes suggest strongannual to decadal variability. For example,the num-ber of hurricanes occurring per year can vary by afactor of three or more for consecutive years,andduring El Niño years,hurricanes are less prevalent inthe Atlantic Basin (Pielke and Landsea,1999).Furthermore,during the 25-year period from 1941to 1965,there were seventeen Category 3 hurri-canes landfalling on the US East Coast or peninsularFlorida, yet between 1966 and 1990 there were onlytwo (Figure 4). The mid-1990s have seen a recur -rence of large numbers of hur ricanes in the NorthAtlantic,perhaps suggesting a return to an active


regime,although the long term implications are notyet clear (Landsea et al.,1996). Globally, the histori-cal record shows “no discernible trends in tropicalcyclone number, intensity, or location”(Henderson-Sellers et al.,1998). However, regional variability,such as observed in the North Atlantic,can be large.

While interdecadal and intera n nual va ri ability ofh u rricane frequency and strength is like ly to domi-nate ch a n ges in hurricanes for at least the fi rs thalf of the next century, i n c reases in hurri c a n ewind strength could result from future eleva t e dsea surface tempera t u res over the next 50 to 100ye a rs . A recent model investigation shows thati n c reases in hurricane wind strength of 5 to 10%a re possible with a sea surface wa rming of 4˚F(2.2˚C) (Knutson et al., 1 9 9 8 ) . Other re s e a rchs u p p o rts the possibility that tropical cycl o n e scould become more intense (Ke rr, 1 9 9 9 ) . For a

m o d e rate hurri c a n e ,s u ch an increase in winds t rength would translate into approx i m a t e ly a 25%i n c rease in the destru c t i ve power of the winds;thus the resulting increase in wave action ands t o rm surge would be greater than the perc e n t agei n c rease in wind speed. S i m i l a r ly, wave height ands t o rm surge would have a greater perc e n t agei n c rease than wind speed, p o t e n t i a l ly yieldingi n c reased impacts to coasts.

H oweve r, it should be noted that recent global cl i-mate model investigations have shown that ElN i ñ o / S o u t h e rn Oscillation (ENSO) ex t remes couldbecome more frequent with increasing gre e n h o u s egas concentra t i o n s . For ex a m p l e , wo rk byTi m m e rmann et al. ( 1 9 9 9 ) , using a GCM with suffi-cient re s o l u t i o n ,s u g gests that the tropical Pa c i fic isl i ke ly to ch a n ge to a state similar to pre s e n t - d ay ElNiño conditions. Since fewer hurricanes occur in

the Atlantic during El Niño ye a rs ,t h e i rresults suggest that Atlantic hurricanes willl i ke ly decrease in frequency in the future .I n t e re s t i n g ly, d u ring seve re El Niño eve n t ss u ch as those in 1982-83 and 1997-98, the jets t ream over the North Pa c i fic brought win-ter storms fa rther south causing ex t e n s i vecoastal erosion and flooding in Califo rn i a .H e n c e , a pro l o n ged El Niño state shouldd e c rease the occurrence of hurricanes inthe Atlantic but lead to an increase in coastalimpacts by winter storms on the West Coast.On the other hand, the results ofTi m m e rmann et al. (1999) also sugge s ts t ro n ger intera n nual va ri ab i l i t y, with re l a t i ve-ly strong cold (La Niña) events becomingm o re fre q u e n t . Although these La Niñaevents would be superimposed upon a high-er mean tempera t u re , this could sugge s tm o re intera n nual va ri ability in Atlantic hurri-canes with more intense activity during thes t ro n ger cold eve n t s .

Even if storm magnitudes and frequenciesof occurrence remain the same,an impor-tant impact of future storms,whether tropi-cal or extratropical,will be their superposi-tion on a rising sea level. This has recentlybeen demonstrated by examining historicalstorm surge magnitudes calculated from sea-level records (Zhang et al.,1997). Theserecords included surges induced by bothwinter storms and hurricanes but weredominated by the more frequent winterstorms. No significant long term trendswere found in storm frequency or severity.When considering that the storms were

468

Figure 4: Loss of life and property from hurricanes making landfallin the continental U.S. over the past 20th century Source: NationalHurricane Center: NOAA. See color figure appendix.

Hurricanes and their Impacts in the 20th Century(1900-1995)


469

strong effects on biotic distributions,life histories,and geochemistry. Coastal runoff also affects circu-lation in estuaries and continental shelf areas,andincreases in runoff have the potential to increasethe vertical stratification and decrease the rate ofthermohaline circulation by adding more freshwaterto the system.

In the event that increased river flows result fromclimate change,more suspended sediments could betransported into the coastal regions,increasing theupper layer turbidity and potentially reducing avail-able light to both plankton and submerged aquaticvegetation. Changes in sediment transport couldalso alter the amount of sediment available for soilaggradation (accumulation) in wetlands and sandsfor littoral systems. Increased sediment transportwill provide needed material for accretion to coastalwetlands threatened by sea-level rise,whereas adecrease in sediment transport will concurrentlydiminish the ability of some wetlands to respond tosea-level rise.

Increased river flows could also increase the flux ofnutrients and contaminants into coastal systems,which influence eutrophication and the accumula-tion of toxins in marine sediments and livingresources. Both increased temperatures anddecreased densities in the upper layers might alsoreduce the vertical convection enough to preventoxygenation of the bottom waters,further contribut-ing to anoxic conditions in the near-bottom waters(Justic´ et al.,1996). Decreased freshwater inflowsinto coastal ecosystems would be likely have thereverse effect, reducing flushing in estuaries,increasing the salinity of brackish waters,and possi-bly increasing the susceptibility of shellfish to dis-eases and predators.

Ocean Temperatures

The oceans represent enormous reservoirs of heat;their heat capacity is such that the upper 16 to 33feet (5 to 10 meters) of the water column generallycontain as much heat as the entire column of airabove it. The oceans work to distribute heat global-ly, and atmospheric and oceanic processes worktogether to control both pelagic and coastal oceantemperatures.

Strong evidence for ocean warming was recentlypublished by Levitus et al.(2000),who evaluatedsome five million profiles of ocean temperaturetaken over the last 55 years. Their results indicatethat the mean temperature of the oceans between 0and 300 meters (0 and 984 feet) has increased by

superimposed on a rising sea level (0.15 inch/yearor 3.9 mm/year at Atlantic City),there is an inferredincrease in storm impact over an 82-year record.For example,the number of hours of extreme waterlevels per year increased from less than 200 in theearly 1900s to abnormally high values of up to 1200hours (averaging typically 600 hours) in the 1990s.Thus sea-level rise will increase impacts to the coastby a storm of a given magnitude.

Freshwater Runoff

Hydrological cycles are fundamental components ofclimate,and climate change is likely to affect bothwater quality and water availability. In general,theconsensus is that the hydrologic cycle will becomemore intense,with average precipitation increasing,especially at high latitudes. Extreme rainfall condi-tions,already demonstrated to have increased overthe 20th century, are likely to become more com-mon,as could droughts and floods (Karl et al.,1995a;Karl et al.,1995b). Changes in freshwaterrunoff will result both from climate-related factorsand changes in population and land use patternsrelating to supply and demand. While the reader isreferred to the Water Sector chapter for further dis-cussion of both climatic and human-inducedchanges in hydrological cycles,the close relation-ship between precipitation,streamflow, and coastalecosystems is a topic of special interest to coastalresearchers,managers,and planners.

The Canadian and Hadley climate models have beenused in concert with hydrology models to estimatethe changes in freshwater runoff for three portionsof coastline:the Atlantic,the Gulf of Mexico,and thePacific. In contrast to GCM scenarios of tempera-ture changes under the influence of increased CO2,the estimates of runoff vary widely. Wolock andMcCabe (1999) determined that for some regionsthe Hadley and Canadian climate models produceopposite results. For the Atlantic coast,the Hadleymodel projects an increase of more than 60% by the2090s,whereas the Canadian model projects arunoff decrease of about 80%. The large differencesare attributed to the projected increases in precipi-tation in the Hadley model, versus small changes tosignificant decreases in precipitation in theCanadian model during the 21st century.

Freshwater runoff affects coastal ecosystems andcommunities in many ways. The delivery of sedi -ment, nutrients,and contaminants is closely linkedto both the strength and timing of freshwaterrunoff. Salinity gradients are driven by freshwaterinputs into estuaries and coastal systems,and have

0.31˚C (0.56˚F) over that same period,which corre-sponds to an increase in heat content of approxi-mately 1 x 1023 Joules of energy. Furthermore,thewarming signal was obser vable to depths of some3000 meters (9843 feet),and the total heat contentbetween 300 and 3000 meters increased by an addi-tional 1 x 1023 Joules of energy (see Figure 5).Although Levitus et al.(2000) could not concludethat the signal was primarily one of climate change,as opposed to climate variability, they note that theirresults are in strong agreement with those projectedby many general circulation models. Earlier workdone by Cane et al.(1997) also suggest thatobserved patterns of change in ocean temperatureare consistent with warming scenarios.

Understanding how future ocean temperatures willbe affected by climate variability and change willdepend on improved understanding of the couplingbetween atmospheric and oceanic processes,andpredicting future variations in the forcing functions.The coupling between these processes is regionally


470

or locally controlled,whereas the climate changemodels are usually more global. As these physicalstructures and processes,along with their naturalmodes of variability, have important implications foroverall levels of biological productivity in the ocean,an accurate assessment of potential changes inocean dynamics will necessarily be speculative untilGCMs with finer resolution of ocean processes aredeveloped.

The ecosystem responses to increased ocean tem-p e ra t u res can ch a n ge productivity both dire c t ly andi n d i re c t ly. Ocean tempera t u re ch a n ges will affe c tnot only the metabolic rate of organisms but alsosea leve l ,c u rre n t s ,m ovement of larva e ,e ro s i o nra t e s ,s u b s t rate stru c t u re s ,t u r b i d i t y, water columns t ra t i fi c a t i o n , nu t rient cycl i n g , and subsequentlyp roduction (McGowa n , et al., 1 9 9 8 ;R i c e ,1 9 9 5 ) .With population ch a n ge s ,c o m munity dynamics willch a n ge ,a l t e ring such things as pre d a t o r - p rey re l a-t i o n s h i p s . Ocean tempera t u re increases will affe c tthe distribution of marine species, l i ke ly with ap o l ewa rd migration of tropical and lower latitudeo rg a n i s m s . The result could be major ch a n ges inthe composition, f u n c t i o n , and productivity ofm a rine ecosystems, i n cluding potential impacts tom a rine biodive rs i t y. For ex a m p l e ,o b s e rvabl ech a n ges in the distribution and abundance of inter-tidal species along the Califo rnia coast have beendocumented by Sag a rin et al. (1999) which are con-sistent with wa rming tre n d s . S i m i l a r ly, l a rge inter-a n nual and interdecadal tempera t u re ch a n ges in theN o rth Pa c i fic have alre a dy been associated withch a n ges in the mixed layer depth over the Nort hPa c i fic and diminished biological pro d u c t i v i t y( M c G owan et al., 1 9 9 8 ).

Increases in temperature will also result in furthermelting of sea ice in polar and subpolar regions.Recent observations in the Arctic have alreadyshown significant declines in ice extent,which hasbeen shrinking by as much as 7% per decade overthe last 20 years (Johannessen et al.,1999) as well asice thickness,with Arctic sea ice thinning (and sub-sequently decreasing in volume) by as much as 15%per decade (Rothrock et al.,1999). Additionally,record low levels of ice extent occurred in theBering and Chukchi seas during 1998,the warmestyear on record (Maslanik et al.,1999). While a possi-ble mechanism could be related to the ArcticOscillation,a decadal scale mode of atmosphericvariability in the Arctic,some comparisons withGCM outputs strongly suggest that these observeddeclines in sea ice are related to anthropogenicallyinduced global warming (Vinnikov et al.,1999). Thereduction and potential loss of sea ice has enormous

Figure 5: A comprehensive analysis of over 5 million temperatureprofiles by Levitus, et al. (2000) reveals a pattern of warming inboth the surface and the deep ocean over the last 40 years. Thelargest warming has occurred in the upper 300 meters (984 feet),which have warmed by an average of 0.31°C (0.56°F), with addition-al warming as deep as 3000 meters (9843 feet). See Color FigureAppendix

Ocean Heat Content in the 0-3000 m Layer

feedback implications for the climate sys-tem as well;ice and snow are highlyreflective surfaces that reflect the vastmajority of the Sun’s incoming radiativeheat back to outer space. By contrast,open oceans reflect only 10 to 20% of theSun’s energy. Thus,the conversion of theArctic ice cap to open ocean could greatlyincrease solar energy absorption,and actas a positive feedback to global warming.

Ocean Currents

Major ocean current systems play a signifi-cant role in the dispersal of marine organ-isms and in the production characteristicsof marine systems. These current systemsare likely to be affected in critical ways bychanges in global and local temperatures,precipitation and runoff, and wind fields.Similarly, oceanic features such as frontsand upwelling and downwelling zoneswill be strongly influenced by variations in tempera-ture,salinity, and winds. These changes will be man-ifest on scales ranging from the relatively small spa-tial and temporal scales characteristic of turbulentmixing processes to very large scales characteristicof the deep water “conveyor belt”circulation,withpotentially dramatic feedback influences on climatepatterns. Changes occurring on this spectrum ofspatial and temporal scales have important implica-tions for overall levels of biological productivity inthe ocean,and are critical to understanding theimplications of global climate change on livingmarine resources.

Research suggests that the processes of formationand circulation of deep-water through the so-calledconveyor-belt circulation (see Figure 6) could bestrongly influenced by changes in temperature andsalinity, with significant implications for the NorthAtlantic region and global ocean circulation(Broecker et al.,1999;Taylor 1999). Generallywarm,saline surface waters are transported north-wards by the Gulf Stream,ultimately feeding theNorwegian,East Greenland,and West Greenland cur-rents. Here winter air-sea interactions cool thealready highly saline water masses, resulting in arapid increase in density. This rapid increase in den-sity drives convection,in which the heavier, denserwater sinks and flows away from the polar regionsto fill the deep ocean basins,driving deep sea ther-mohaline circulation.

Increased temperature or decreased salinity (result-ing from changes in precipitation patterns or melt-

ing ice sheets) at high latitudes could result in areduction in the deep-water formation by decreas-ing the density of surface waters,with importantconsequences for this convective system(Schmittner and Stocker, 1999;Broecker, 1997).Hadley Centre models have suggested a decline inthe strength of deep-water circulation of approxi-mately 25% under some scenarios (Wood et al.,1999). Such a decline could lead to a general cool-ing throughout the North Atlantic region, resultingfrom a reduction in transport of warmer watersfrom lower latitudes (Driscoll and Haug,1998).Additionally, a complete cessation of the conveyorbelt circulation is possible,which could cause win-ter temperatures throughout the North Atlantic tofall abruptly by as much as 9˚F (5˚C). Sedimentaryrecords in the northern Atlantic suggest that pastcirculation shutdowns have occurred in extremelyshort time intervals,associated with abrupt climateshifts and dramatic cooling in Europe. Ironically, theresult would likely be further acceleration of globalwarming trends as a result of a decrease in theoceanic uptake of carbon dioxide that would followweakening or cessation of thermohaline circulation(Sarmiento and Le Quere,1996).


471

Figure 6: The ocean plays a major role in the distribution of theplanet's heat through deep sea circulation. This simplified illustra-tion shows this "conveyor belt" circulation which is driven by dif-ferences in heat and salinity. Records of past climate suggest thatthere is some chance that this circulation could be altered by thechanges projected in many climate models, with impacts to climatethroughout lands bordering the North Atlantic (Modified fromBroecker, 1991).See color figure appendix.

The Global Ocean Conveyor Belt


472

KEY ISSUES1. Shoreline Systems,Erosion,and Developed

Coastal Areas2. Threats to Estuarine Health3. Coastal Wetland Survival4. Coral Reef Die-offs5. Stresses on Ocean Margins and Marine Fisheries

Coastal and marine resources are uniquely influ-enced by the long-term climate dynamics describedabove. For the purposes of this Assessment,thepotential impacts on five principal ecosystem typeswere evaluated:shorelines,estuaries,coastal wet-lands,coral reefs,and ocean margins/marine fish-eries. While there is significant overlap betweensome of these ecosystem types,and also further divi-sions that could be made within these ecosystems,this breakdown provides a methodology for under -standing what may be the most significant genericimpacts on ecological systems. These are conse-quences which are either ongoing or could be rea-sonably expected on the basis of past climate vari-ability, forcing scenarios,and change thresholds. Inaddition,individual case studies were conducted inorder to provide specific examples of the complexset of interactions between the effects of climateand human activities; results from several of thesecase studies have been incorporated here.

1. Shoreline Systems, Erosion, andDeveloped Coastal Areas

Storms,hurricanes,typhoons,and similarly extremeatmospheric phenomena along coasts produce highwinds that in turn generate large waves and cur-rents. Storms and hurricanes produce storm surgesthat temporarily raise water levels by as much as 23feet (7 meters) above normal. Although theseevents are sporadic,they are a primary cause ofbeach erosion and shoreline impacts throughout theUS. Hurricanes account for far more insured lossesof property than do other hazards such as earth-quakes and wildfires. Sea-level rise increases thevulnerability of shorelines and floodplains to stormdamage by increasing the baseline water level forextreme storms and coastal flooding events. In addi-tion to storm and flooding damages,specificimpacts associated with sea-level rise could includeincreased salinity of estuaries and freshwateraquifers,altered tidal ranges in rivers and bays,changes in sediment and nutrient transport whichdrive beach processes,and the inundation of wastedisposal sites and landfills which could reintroducetoxic materials into coastal ecosystems.

Individual hurricanes have resulted in enormouseconomic impacts,particularly to the Southeast.Hurricane Hugo in 1989 caused an estimated $9 bil-lion in damages;Hurricane Andrew in 1992 causedan estimated $27 billion in damages;and HurricaneGeorges in 1998 caused an estimated $5.9 billion indamages (Source:National Climatic Data Center).However, hurricanes are not the only storm threatto coastal inhabitants and property owners.Extratropical winter storms have significant impactsas well,such as the Halloween “nor’easter”of 1991which caused damages amounting to over $1.5 bil-lion along the Atlantic Coast. Along the PacificCoast, extratropical storms battered California caus-ing an estimated $500 million in damage during the1997-98 El Niño (Griggs and Brown,1999). Ascoastal population and property constructionincreases,the economic vulnerability of humandevelopments in coastal areas to hurricane andstorm activity will continue to rise,and will beaggravated by any climate-induced changes in thefrequencies or intensities of these events.

Coastal erosion is already a widespread problemthroughout much of the country (Figure 7),and hassignificant impacts to both undeveloped shorelinesand coastal development and infrastructure. In addi-tion to storms and extreme events,interannualmodes of variability such as ENSO have been shownto impact many shorelines. For example,along thePacific Coast,cycles of beach and cliff erosion havebeen linked to El Niño events,which elevate aver-age sea levels over the short term and alter the fre-quency, intensity, and direction of storms impactingthe coastline. During the 1982-83 and the 1997-98El Niños,impacts were especially severe alongshorelines throughout California,Oregon,andWashington (Komar, 1999;Kaminsky et al.,1999).Good (1994) has shown that along the centralOregon coast,a rapid buildup of seawalls and revet-ments routinely followed major El Niño events ofthe last two decades,as coastal property ownersattempted to protect their shorelines from increas-ing erosion.

Atlantic and Gulf coastlines are especially vulnerabl eto sea-level ri s e , as well as to ch a n ges in the fre q u e n-cy and seve rity of storms and hurri c a n e s . Most of theEast and Gulf coasts are rimmed by a series of barri e risland and bay systems which separate the ge n t lysloping mainland coastal plains from the continentals h e l f. These islands bear the brunt of fo rces fro mwinter storms and hurri c a n e s ,p rotecting the main-land from wave action. H oweve r, these islands andcoastlines are not stable but are instead highlydynamic systems that are highly sensitive and re s p o n-

s i ve to the rate of re l a t i ve sea-level rise as well as thef requency and seve rity of storms and hurri c a n e s .

In response to rising sea levels,these islands typical-ly “roll over”towards the mainland,through aprocess of beach erosion on their seaward flank andoverwash of sediment across the island. Humanactivities block this natural landward migrationthrough the construction of buildings, roads,andseawalls;as a result,shorelines erode,increasing thethreat to coastal development and infrastructure.Subsequent impacts include the destruction of prop-erty, loss of transportation infrastructure,increasedcoastal flooding,negative effects on tourism,saltwa-ter intrusion into freshwater aquifers,and impactsto fisheries and biodiversity. Such impacts willunquestionably further modify the functioning ofbarrier islands and inland waters that have alreadybeen impacted by pollution,physical modification,sediment starvation by dams,and material inputsrelated to human activities.

B a rrier islands are cert a i n ly not the only coastlinesat ri s k . P ri vate pro p e rty on mainland and estuari n es h o re l i n e s , harbor installations, other wa t e r - d e p e n d-ent activities and infra s t ru c t u re ,t ra n s p o rt a t i o ni n f ra s t ru c t u re , re c reational are a s , and agri c u l t u ra la reas are all part i c u l a r ly vulnerable to inu n d a t i o nand increased erosion as a result of climate ch a n ge ,e s p e c i a l ly in low - lying are a s . Assessing the total


473

economic impacts from sea-level rise on coastala reas and on a national scale is still somewhat spec-u l a t i ve . N eve rt h e l e s s , a study by Yohe et al. ( 1 9 9 6 )q u a n t i fied the economic costs (protection plusabandonment) to coastal stru c t u res with a onemeter sea-level ri s e . This analysis estimated costs ofas mu ch as $6 billion (in 1990 dollars) betwe e n1996 and 2100. H oweve r, this number re p re s e n t sm a rke t - valued estimates only, w h i ch are deri ve df rom pro p e rt y - value appre c i a t i o n ,m a rket adapta-t i o n , and protection costs. As such , this is a mini-mum cost estimate, as it does not include the lostecosystem services value of non-market re s o u rc e s ,s u ch estuaries and tidal we t l a n d s , or the costs toc o m munities resulting from reductions in coastaleconomic activities such as fi s h i n g ,t o u ri s m ,a n dre c re a t i o n . A compre h e n s i ve rev i ew of the poten-tial cost of sea-level rise to developed coastlineswas re c e n t ly completed by Neumann et al. ( 2 0 0 0 ) ;and should be re fe rred to for a wider ra n ge of esti-mates of the potential economic consequences ofs e a - l evel rise to developed are a s .

Many coastal structures were designed with the100- year flood as their basis. This flooding leveldetermines the elevations to which the federal proj-ects are built,such as the Army Corps of Engineers

Figure 7: A general classification scheme of shoreline erosionrates throughout the US. (modified from Dolan et al., 1985). SeeColor Figure Appendix

climate va ri ability and ch a n ge . Wa rmer spri n gand fall tempera t u res will alter the timing ofseasonal tempera t u re tra n s i t i o n s , w h i ch affe c ta number of ecologi c a l ly important pro c e s s e sand will thus impact fi s h e ries and marine pop-u l a t i o n s . As re fe renced ab ove in the sectionon fre s h water fo rc i n g , the amount and timingof fre s h water f l ow into estuaries gre a t ly inf l u-ence salinity, s t ra t i fi c a t i o n , c i rc u l a t i o n , s e d i-m e n t , and nu t rient inputs. T h u s , i n c reases inw i n t e r - s p ring disch a rges in particular couldd e l i ver more excess nu t rients as well as con-taminants to estuari e s , while simu l t a n e o u s lyi n c reasing the density stra t i fi c a t i o n . Both ofthese would increase the potential for algalblooms (including potentially harmful species)and the development of hy p oxic (low ox y ge n )c o n d i t i o n s , w h i ch could in turn incre a s es t resses on sea grasses and affect commerc i a lfishing and shellfish harve s t i n g .

Currently, nutrient over-enrichment is one of thegreatest threats to estuaries in the US,with over halfthe nation’s estuaries having at least some of thesymptoms of moderate to high states of eutrophica-tion (Bricker et al.,1999). Eutrophication has multi-ple impacts to estuaries,as well as other coastal

levees that protect New Orleans. It is also the levelto which coastal structures must be built to qualifyfor flood insurance through the Federal EmergencyManagement Agency’s (FEMA) Flood InsuranceProgram. If sea level rises,the statistics used todesign these structures change. For example,whatwas once considered a 50-year flood could becomeas severe as a 100-year flood following an increasein sea level (Pugh and Maul,1999). Coastal insur-ance rates would have to be adjusted to reflect thechange in risk. Furthermore,FEMA has estimatedthat the number of households in the coastal flood-plain could increase from 2.7 million currently tosome 6.6 million by the year 2100,as a result of thecombination of sea-level rise and rapidly growingcoastal populations and development (FEMA,1991).Again,it is clear that the vulnerability of coastaldeveloped areas to natural disasters can be expect-ed to rise throughout the 21st century, even inde-pendent of climate-induced changes in risk.

2. Threats to Estuarine Health

E s t u a ries are among the most pro d u c t i vecoastal ecosystems, c ritical to the health of agreat many commercial and re c reational fi s h-e ri e s , and are affected in nu m e rous ways by


474

South Florida Case Study

The natural South Florida ecosystem was largely defined by a high degree of variability in rainfall both withinand across years. The region originally supported a mosaic of multiple interactions among many differentecosystem types,including freshwater, wetland,mangrove,estuarine,and coral habitats all intimately coupledby the regional hydrology. All of these ecosystems now severely impacted by a large and very rapidly growinghuman population perched along a very narrow strip of coastal land. Tourism in South Florida is a multibil-lion-dollar-a-year industry, and the existence of the Everglades,Biscayne National Park and the Florida KeysNational Marine Sanctuary illustrate the high value society places on this unique environment.

The South Florida region is vulnerable to multiple climate change stresses,as sea-level rise, changes in the fre-quency of freezing events,hurricanes,droughts and associated fires,sea surface temperatures and many oth-ers all affect the full diversity of ecosystems. The responses of all of these ecosystems will be significantlymodified or constrained by human development. In particular, South Florida has developed one of theworld's largest water management systems with the primary objectives of dampening hydrological variability,providing flood protection,and supplying water to urban and agricultural systems (Harwell,1998;US COE1994). This alteration of hydrological functions has caused serious degradation to the ecosystems,whichwere formerly adapted to natural modes of variability but have been made more vulnerable as a result ofhuman disturbance and development.

Any significant change to the precipitation regime could have major consequences to coastal ecosystems;already the Everglades and associated systems are not considered sustainable due to regional hydrologic alter-ations. While a massive and costly restructuring of the water management system is currently underway, areduced precipitation regime could further increase multi-year droughts,increase fire frequency and intensity,reduce water supply to urban users,and increase estuarine hypersalinity. Thus, global climate change can beexpected to exacerbate the effects of natural and anthropogenic stresses on South Florida,making the currentchallenge of restoring and sustaining its ecosystems even greater.

ecosystems,including more frequent and longer last-ing harmful algal blooms,degradation of seagrassbeds and coral reefs,alteration of ecological struc-ture,decreased biological diversity, and the loss offishery resources resulting from low oxygen events.In this instance, climate change is only one aspect ofhuman-accelerated global environmental change,and particularly for estuaries,the broader effects ofglobal change have to be considered. Specifically,human activities have altered the biological availabil-ity of nitrogen through the production of inorganicfertilizer, the combustion of fossil fuel,and the man-agement of nitrogen-fixing agricultural crops. As aresult,since the 1960s the world has changed fromone in which natural nitrogen fixation was the dom-inant process to a situation in which human-con-trolled processes are at least as important in makingnitrogen available on land.

This increased availability has led to an increase inthe delivery of nitrogen to coastal marine systems,particularly estuaries (Vitousek et al.,1997). Thedelivery of nutrients has been closely linked to vari-ability in streamflow; for example,in the Gulf ofMexico,increases in freshwater delivery from theMississippi River are closely linked with increases innutrient delivery and the subsequent developmentof hypoxia. The spring floods of 1993 resulted inthe greatest nitrogen delivery ever recorded in theGulf of Mexico (Rabalais et al.,1999),and the arealextent of the resulting hypoxic zone was twice aslarge as the average for the preceding eight years.However, the response of individual estuaries tochanges in freshwater and nutrient delivery will dif-fer significantly as related to variation in water resi-dence times,stratification,and the timing and mag-nitude of inputs. Yet because the solubility of oxy-gen in warmer waters is lower than in coolerwaters,any additional warming of estuaries duringsummer months would be likely to aggravate theproblems of hypoxia and anoxia that already plaguemany estuaries.

With few exceptions,the potential consequences ofclimate change to estuarine ecosystems are not yetbeing considered in long-term estuarine and coastalzone management. In the Mid-Atlantic,estuaries arealready in a degraded environmental condition,butare currently the subject of substantial societal com-mitments for their restoration through pollutionreduction,habitat rehabilitation,and more sustain-able use of living resources. An increase in winter-spring precipitation that delivers additional nutri-ents to estuaries such as the Chesapeake Bay wouldmake achieving the management goals of reducingnutrient inputs and improving dissolved oxygen lev-

els more difficult. More efficient nutrient manage-ment practices and more extensive restoration ofriparian zones and wetlands would be required tomeet current nutrient goals in seasonally wetterwatersheds. However, it should be recognized thatclimate change is expected to bring about alter-ations to many estuaries for which mitigatingresponses may not exist.

3. Coastal Wetland Survival

Coastal wetlands are some of the most va l u abl eecosystems in the nation as well as some of the mostt h re a t e n e d . By providing hab i t a t , re f u ge , and fo rageo p p o rtunities for fishes and inve rt e b ra t e s ,m a rs h e sand mangroves around the coastal US are the basisof many commu n i t i e s ’ economic live l i h o o d s(Costanza et al., 1 9 9 7 ;M i t s ch and Gosslink, 1 9 9 3 ) .The role of wetlands in nu t rient uptake ,i m p rov i n gwater quality, and reducing nu t rient loads to thecoastal ocean is widely re c o g n i z e d , as is their va l u ein providing re c reational opportunities and pro t e c t-ing local communities from flooding — either bydampening storm surges from the ocean or prov i d-ing storage for ri ver fl o o dwa t e rs . Climate va ri ab i l i t yand ch a n ge compound existing stresses from humanactivities (Mark h a m ,1 9 9 6 ) ,s u ch as dre d ging and/orfilling for deve l o p m e n t ,n avigation or mineral ex t ra c-t i o n ,a l t e red salinity and water quality, and the dire c tp re s s u res of increasing nu m b e rs of people living andre c reating in close prox i m i t y.

These ecosystems are sensitive to a number of cli-mate related variables. Changes in atmospherictemperatures and carbon dioxide concentrationsgenerally result in increased plant production,although the response varies considerably amongspecies. Increases in freshwater discharge wouldgenerally benefit many coastal wetlands,anddecreases might result in salinity stress for somecommunities,particularly in the western Gulf ofMexico where already limited freshwater inputs areexpected to decrease dramatically. For coastal wet-lands facing current or future relative sea-level rise,increased sediment delivery will be necessary forvertical accumulation of the substrate.

Coastal wetlands can cope with changes in sea levelwhen they are capable of remaining at the same ele-vation relative to the tidal range,which can occur ifsediment buildup equals the rate of relative sea-levelrise or if the wetland is able to migrate. Migrationoccurs when the wetland moves upslope alongwith the tidal range,with the seaward edge of thewetland drowning or eroding and the landwardedge invading adjacent upslope communities


475

lands become extremely difficult to restore. Theremaining 3.5 million acres of wetlands in SouthLouisiana provide critical nursery areas for finfishesand crustaceans,particularly shrimp and crabs,which make up the bulk of the state’s multimilliondollar seafood industry. Additionally, these coastalwetlands serve as important buffers against stormsurges,protecting inland residential and commercialinfrastructure from severe flooding. Thus,if climatechange exacerbates the current rate of sea-level rise,those regions of coastal Louisiana where marshesare on the margin of survival will suffer, and evenmore extensive land loss will result.

4. Coral Reef Die-Offs

Although coral reefs might seem exotic to many re s i-dents of the US, these ecosystems play a major ro l ein the env i ronment and economies of two states( F l o rida and Hawaii) as well as most US terri t o ries inboth the Cari bbean and the Pa c i fi c . C o ral reefs areva l u abl e , if often ove r - u t i l i z e d , economic re s o u rc e sfor many tropical coastal re gi o n s ,p roviding nu m e r-ous fi s h e ries opport u n i t i e s , re c re a t i o n ,t o u ri s m ,a n dcoastal protection (Wilkinson and Buddemeier,1 9 9 4 ) . It is widely recognized that reefs and re l a t e dc o m munities are also some of the largest store h o u s-es of marine biodive rs i t y, with untapped re s o u rces ofgenetic and biochemical materi a l s , and of scientifi ck n ow l e d ge (Ve ro n ,1 9 9 5 ) . F u rt h e r, the living re e fc o m munities are not isolated dots on a map, but partof an interacting mosaic of oceanic and terre s t ri a l

h abitats and commu n i t i e s , and many of the org a n-isms and processes found on reefs are impor-

tant in a mu ch wider sphere of related env i-ronments (Buddemeier and Smith,

1 9 9 9 ) . In many important way s ,the condition and re s p o n s e s

of coral reef commu n i-ties can be seen as diag-nostic of the conditionof the wo r l d ’s low - l a t i-tude coastal oceans.

Degradation of reefcommunities has been

increasing worldwide over thepast few decades,and the last few years

have seen a dramatic increase in the extent ofcoral bleaching,new emergent coral diseases andwidespread reef die-offs. Coral bleaching,whichoccurs when symbiotic algae that live in the coralsthemselves are expelled from their hosts due tohigh sea surface temperatures,has been occurringwith increasing frequency in the last severaldecades,notably during El Niño events. The 1998 El

(Figure 8). However, if wetlands are unable to keeppace with relative sea-level change,or if their migra-tion is blocked by bluffs,coastal development,orshoreline protection structures,then the wetlandwill become immersed and eventually lost as risingseas submerge the remaining habitat.

Currently, many US coastlines are areas of coastalsubsidence where both natural and human-inducedprocesses strongly influence wetlands. CoastalLouisiana has experienced the greatest wetland lossin the nation,where a combination of anthro-pogenic (human-caused) and natural factors havedriven losses between 24 and 40 square miles peryear during the last 40 years. As approximately 25%of the nation’s brackish and freshwater coastal wet-lands are found in Louisiana,this constitutes asmuch as 80% of the total loss of these wetlands inthe US (Boesch et al.,1994). Changes have occurredso rapidly in the many bald cypress forests nearNew Orleans that they have been converted directlyto open water rather than being gradually overtakenby salt marsh. Once lost to open water, these wet-


476

Figure 8: The rate of sea-level rise is projected to accelerate 2 to 5fold over the next 100 years. The delivery of sediments to coastalwetlands is extremely important in determining the potential ofthese systems to maintain themselves in the face of current andfuture sea-level changes (based on Reed, 1995). See Color FigureAppendix

Processes Affecting Wetland Migration

Niño in particular was associated with unprecedent-ed high sea surface temperatures and thus the mostwidespread coral bleaching ever observed (Stronget al.,2000;Wilkinson et.al.,1999;Hoegh-Guldberg,1999). In many past bleaching events,both thealgae and the corals are capable of recovery;howev-er, warming events in 1998 resulted in unusuallyhigh levels of mortality from which many reefs haveyet to recover. In addition to bleaching ef fects,there has been an upsurge in the variety, incidence,and virulence of coral diseases in recent years,withmajor die-offs reported,particularly in Florida andthe Caribbean region. The causes of these epi-zootics are not known,but they suggest that coralecosystems and organisms are weakened and vul-nerable.

One of the most significant unanticipated directconsequences of increased atmospheric carbondioxide concentration is the reduction of carbonateion concentrations in seawater. Carbon dioxide actsas an acid when dissolved in seawater, causing sea-water to be less alkaline. A drop in alkalinity subse-quently decreases the amount of calcium carbonate(aragonite) which can be dissolved in seawater,which in turn decreases the calcification rates ofreef-building corals and coraline algae (Figure 9)(Kleypas et al.,1999a;Gattuso,1999). The impactson coral reefs are weaker skeletons, reduced growthrates,and an increased vulnerability to erosion,witha wide range of impacts on coral reef health andcommunity function. Additionally, the effects ofincreasing atmospheric carbon dioxide on the con-centration of carbonate ions is greatest at the mar-gins of coral distributions,due to the fact that car-bon dioxide is more soluble in cooler waters. Thus,these effects will be most severe at high latitudes,socoral reefs at the margins of their distribution arenot expected to expand their ranges as might other-wise be predicted by some ocean warming scenar-ios (Figure 10).

Human-induced disturbances have already taken asignificant toll on coral reef systems,from activitiessuch as over-fishing,the use of destructive fishingtechniques, recreational activities,ship groundings,anchor damage,sedimentation,pollution,andeutrophication. Thus,the future for many coral reefsappears bleak. The synergy of stresses and theresponse time scales involved to restore reefs makeintegrated management essential but dif ficult. Giventrends already in motion,some of the more marginalreef systems will almost certainly continue to deteri-orate,making allocation of resources an importantpolicy issue. The demise or continued deteriorationof reef communities will have profound social and

economic implications for the US,as well as seriouspolitical and economic implications for the world.

5. Stresses on Ocean Margins andMarine Fisheries

Projected changes in the marine environment haveimportant implications for marine populations andfisheries resources. In addition to their economicvalue,marine fisheries hold a special social and cul-tural significance in many coastal communities.However, over-fishing and habitat loss have alreadytaken serious tolls on the nation’s fisheries and thecommunities that depend upon them. Currentlysome 33% of the stocks for which trends are knownare either over-fished or depleted (NMFS,1998),andfor the vast majority of stocks,an accurate assessmentof their status is unknown. The possible effects of cli-mate change range from shifts in distribution to


477

Figures 9: Increasing levels of atmospheric CO2 are projected todecrease carbonate ion concentrations in seawater, which willdecrease the calcification rates of many coral building species andimpact coral reef health (Gattuso et al., 1999).

Bleached coral.

Effects of CO2 on Coral


478

changes in survival and growth rates,with directimplications for fishery yields.

Changes in climate forcing will have importanteffects in ocean margin ecosystems through expect-ed changes in the distribution and abundance ofmarine organisms and in fundamental changes inthe production characteristics of these systems.Changes in temperature,precipitation,wind fields,and sea level can all be expected to affect oceano-graphic conditions in the ocean margins with directramifications for marine life in these areas.Physiological effects of temperature and salinitychanges can also be expected with potentiallyimportant consequences for growth and mortalityof marine species (Jobling,1996). Impacts tomarine ecosystems and fisheries associated with ElNiño events illustrate the extent to which climateand fisheries can interact. For example,the high sea

surface temperatures and anomalous conditionsassociated with the 1997-98 El Niño had a tremen-dous impact on marine resources off of Californiaand the Pacific Northwest. Landings of marketsquid,California’s largest fishery by volume and sec-ond largest in value, fell from over 110,000 metrictons in the 1996-97 season to less than 1000 metrictons during the 1997-98 El Niño season (Kronman,1999). Amongst the many other events associatedwith this El Niño were high sea lion pup mortalitiesin California (Wong,1997),poor reproductive suc-cess in seabirds off of Oregon and Washington,andunexpected catches of warm-water marlin off of theWashington coast (Holt,1997). Even further northwere rare coccolithophore blooms,massive seabirddie-offs along the Aleutian Islands,and poor salmonreturns in Alaska’s Bristol Bay sockeye salmon fish-ery (Macklin,1999).

Calcium Carbonate Saturation in Ocean Surface Waters

>4.0 Optimal

3.5 - 4 Adequate

3 - 3.5 Marginal

<3.0 Extremely Low

Preindustrial (~1880) Current (2000)

Projected (~2050)Figure 10: Map of current and projected changes in calciumcarbonate saturation in ocean surface waters. Corals requirethe right combination of temperature, light, and calcium car-bonate saturation. At higher latitudes, there is less light andlower temperatures than nearer the equator. The saturationlevel of calcium carbonate is also lower at higher latitudes,in part because more CO2, an acid, can be dissolved in cold-er waters. As the CO2 level rises, this effect dominates,making it more difficult for corals to form at the polewardedges of their distribution. These maps show model resultsof the saturation level of calcium carbonate for pre-industri-al, present, and future CO2 concentrations. The dots indi-cate present coral reefs. Note that under model projectionsof the future, it is very unlikely that calcium carbonate satu-ration levels will provide fully adequate support for coralreefs in any US waters. The possibility of this future sce-nario occurring demands continued research on effects ofincreasing CO 2 on entire coral reef systems. Classificationintervals for saturation effects on reef systems are derivedfrom Kleypas et al. (1999b). See Color Figure Appendix


479

Climate change is likely to cause even greaterimpacts to marine populations. Poleward shifts indistribution of marine populations can be expectedwith increasing water temperatures (Murowski,1993). Species at the southern extent of their rangealong the East Coast can be expected to shift theirdistribution to the north with important overall con-sequences for ecosystem structure. Cod,Americanplaice,haddock,Atlantic halibut, redfish,and yellow-tail flounder all would be expected to experiencesome poleward displacement from their southerlylimits in the Gulf of Maine and off New Englandunder increasing water temperatures (Frank et al.,

1990). Thus,the loss of populations or sub-popula-tions due to shifts in temperature under these con-straints is likely, with regional impacts to communi-ties that depend on local populations. An expansionof species commonly occurring in the middleAtlantic region,such as butterfish and menhaden,into the Gulf of Maine could also be expected.

In the Pacific,Welch et al.(1998) suggest that pro-jected changes in water temperatures in the NorthPacific could possibly result in a reduction of suit-able thermal habitat for sockeye salmon. The mostsurprising prediction is that under scenarios of

Understanding Global Climate and Marine Productivity

The US Global Ocean Ecosystems Dynamics (US GLOBEC) research program is operated in collaborationbetween the NOAA/National Ocean Service (NOS) Coastal Ocean Program and the NSF BiologicalOceanography Program. GLOBEC is designed to address the questions of how global climate change mayaffect the abundance and production of marine animals. Two large marine regions have been the subjects ofintensive research,the Northwest Atlantic and the Northeast Pacific. In the Atlantic program,the overall goalis to improve the predictability and management of the living marine resources of the region. This is beingdone through improved understanding of ecosystem interactions and the coupling between climate change,the ocean's physical environment and the ecosystem components (Figure 11). Particularly crucial physicaldrivers are the North Atlantic Oscillation and the salinity variations derived from flows from the Labrador Sea.A major objective of this program is to apply the understanding of the physical processes that affect the dis-tribution, abundance and production of target species. This information will be used in the identification ofcritical variables that will support ecosystem-based forecasts and indicators as a prelude to the implementa-tion of a long-term ecosystem monitoring strategy.

Remarkable changes have been observed in recentdecades in the Northeast Pacific. Concurrentchanges in atmospheric pressure and ocean temper-atures indicate that in 1976 and 1977 the NorthPacific shifted from one climate state or regime toanother that persisted through the 1980s. Analysisof records of North Pacific sea surface temperatureand atmospheric conditions show a pattern of suchregime shifts lasting several years to decades.Although the important linkages are poorly under-stood,there is growing evidence that biological pro-ductivity in the North Pacific responds quite strong-ly to these decadal-scale shifts in atmospheric andoceanic conditions by alternating between periodsof high and low productivity. Some of the keyissues being addressed in this region concern theevaluation of life history patterns,distributions,growth rates,and population dynamics of highertrophic level species and their direct and indirectresponses to climate variability. Additionally, theprogram seeks to better understand complexecosystem interactions,such as how the NorthPacific ecosystem is structured and whether highertrophic levels respond to climate variability solely asa consequence of bottom-up forcing.

Figure 11: Biological processes in the ocean are related to climatein many ways, both directly and indirectly. Improvements in ourunderstanding of the direct and indirect effects of climate on bio-logical processes in the oceans are essential to predicting howmarine populations might respond to future change Source: USGLOBEC.

Climatic Pathways Affecting the AbioticEnvironment and Biological Processes


480

For example, changes in the abundance of sardineand anchovy populations appear to be stronglylinked to global climate shifts (Lluch-Belda et al.,1992). The California sardine population wasrecently declared “recovered,” decades after its neardisappearance from Pacific waters. The recoveryhas in part been related to an increase in sea surfacetemperatures off the Pacific Coast observed over thelast 20 years (California Department of Fish andGame,1999). Sardines are once again found in large

future climate,none of the Pacific Ocean is project-ed to lie within the thermal limits that have definedthe distribution of sockeye salmon over the last 40years. Under such scenarios,the distribution of allspecies of salmonids could be restricted to marginalseas in the North Pacific region (see also the PacificNorthwest chapter and Mantua et al.1997 for moreinformation on interactions between salmon and cli-mate variability). Conversely, higher sea surfacetemperatures will be beneficial for other species.

Predicting Coastal Evolution at Societally-Relevant Time and Space Scales

One of the most important applied problems in coastal geology today is determining the response of thecoastline to sea-level rise. Prediction of shoreline retreat and land loss rates is critical to the planning offuture coastal zone management strategies,and assessing biological impacts due to habitat changes or destruc-tion. Presently, long-term (~50 years) coastal planning and decision-making has been done piecemeal,if at all,for the nation's shoreline. Consequently, facilities are being located and entire communities are being devel-oped without adequate consideration of the potential costs of protecting them from or relocating them dueto sea-level rise-related erosion, flooding and storm damage.

The prediction of future coastal evolution is not stra i g h t fo r wa rd . T h e re is no standard methodology, and eve nthe kinds of data re q u i red to make such predictions are the subject of mu ch scientific debate. A number of pre-d i c t i ve appro a ches could be used, i n cl u d i n g : 1) ex t rapolation of historical data, ( e .g . , coastal erosion ra t e s ) ,2 )i nundation modeling, 3) application of a simple ge o m e t ric model (e. g . , the Bruun Rule), 4) application of a sedi-

ment dy n a m i c s / b u d get model, or 5) MonteCarlo (pro b abilistic) simu l a t i o n . E a ch of thesea p p ro a ch e s ,h oweve r, has its shortcomings orcan be shown to be invalid for certain applica-t i o n s . S i m i l a r ly, the types of input datare q u i red va ry widely, and for a gi ven appro a ch( e .g . , sediment budge t ) , existing data may bei n d e t e rminate or simply not ex i s t .

The relative susceptibility of different coastalenvironments to sea-level rise,however, maybe quantified at a regional to national scale(e.g.,Gornitz et al.1994) using basic data oncoastal geomorphology, rate of sea-level rise,and past shoreline evolution (Figure 12). Apilot project is underway at the USGeological Survey, Coastal and MarineGeology Program to assess the susceptibilityof the nation's coasts to sea-level rise from ageologic perspective. The long-term goal ofthis project is to predict future coastalchanges with a degree of certainty useful tocoastal management, following an approach

similar to that used to map national seismic and vol-canic hazards. This information has immediateapplication to many of the decisions our society willbe making regarding coastal development in boththe short- and long-term.

Figure 12: These preliminary results illustrate the relative vulnera-bility to sea-level rise along the New York and New Jersey coastlineas assessed by ongoing USGS research. Note that the vulnerabilitymapped here is likely to change as methodologies in this pilot pro-gram are critically evaluated and improved (Source: USGS). SeeColor Figure Appendix

numbers as far north as British Columbia(Hargreaves et al.,1994),and their recovery has ledto a concurrent resurgence of the fishery long agoimmortalized in John Steinbeck’s Cannery Row.

ADAPTATION STRATEGIESAssessing the effects of climate variability andchange on coastal and marine resources is especiallydifficult given that human activities are generallyresponsible for the greatest impacts on coastal andmarine environments. The nature of climate effects— both detrimental and beneficial to resources inquestion — are likely to vary greatly in the diversecoastal regions of the US. Anthropogenic distur-bance often results in a reduction in the adaptivecapacity of systems to cope with change and stress,making the real or potential impacts of climate diffi-cult to observe. It is in this context that climatechange acts as an increased stress on coastal andmarine environments,adding to the cumulativeimpact of both natural and anthropogenic stress onecological systems and resources.

This is most abundantly clear in coral reef systems.Coral reefs,both in US waters and worldwide,areclearly stressed,and many are degraded to the pointof destruction. The prospects for the future are thatin many, if not most,instances,the levels of both cli -mate-related stress and local or regional stress result-ing from anthropogenic impacts will increase. Theimplication for coral reefs is that local and regionalreef protection and management efforts must beeven more effective in controlling local stresses,toprovide some compensation for large scale impacts.If local and regional anthropogenic stresses contin-ue or increase,many of the reefs that are heavilyused or affected by humans will be poor candidatesfor survival. This would result in substantial impactsto the communities and regional economies thatdepend upon healthy reefs for fisheries,subsistence,recreation,and tourism.

For responding or adapting to sea-level rise,threegroups of strategies have been discussed,thesebeing:(planned) retreat,accommodation,and pro-tection (IPCC,1996). A retreat strategy would pre-vent or discourage major developments in vulnera-ble coastal areas and could include rolling ease-ments,which allow development but explicitly pre-vent property owners from preventing the uplandmigration of wetlands and beaches. An accommoda-tion strategy might elevate land surfaces or humanstructures,modify drainage systems,or otherwisechange land use practices,thus allowing many

coastal ecosystems to be maintained. A protectionstrategy could utilize beach nourishment and dunestabilization,as well as dikes,seawalls,bulkheads,and revetments,to form a barrier between waterand land. This might generally lead to a loss of natu-ral functions for beaches, wetlands,and other inter-tidal zones,but would be capable of roughly main-taining the coastline in place.

In areas where beaches or wetlands must migrateinland in response to sea-level change,it has beenshown that planning and implementing protectionor retreat strategies for coastal developments cansubstantially reduce the long-term economicimpacts of inundation and shoreline migration.While some regulatory programs continue to permitstructures that block the migration of wetlands andbeaches,others have tried to decrease economicmotivations for developing in vulnerable areas,orhave experimented with retreat strategies. Forexample,coastal management programs in Maine,Rhode Island,South Carolina,and Massachusettshave implemented various forms of rolling easementpolicies in an attempt to ensure that wetlands andbeaches can migrate inland as sea-level rises. Severalother states require coastal counties to consider sea-level rise in their coastal management plans.However, the difficulties and obstacles facing coastalmanagers in effectively implementing setback androlling easement policies are substantial. Many ofthe programs initiated to date are not mandatory,and the effectiveness of state coastal zone manage-ment programs often varies significantly.

While the potential impacts of sea-level rise havebegun to initiate concern and some discussion ofpotential response strategies,the broader ramifica-tions of changes in temperature,freshwater dis-charges,and the frequency and intensity of stormevents,have scarcely been assessed. For example,inestuaries such as Chesapeake Bay, a particular focusof restoration efforts is the reduction of nutrientover-enrichment,or eutrophication,from both pointdischarges and diffuse sources throughout thewatersheds draining into the estuaries. However,efforts to reduce eutrophication will have to con-tend with multiple climate change-related problemssuch as sea-level rise,increased winter-spring dis-charges but reduced summer runoff, warmer seasurface temperatures,and greater shoreline erosion.

For marine fisheries,adaptations to changes in theproduction characteristics of exploited populationswill include adjustments in the recommended har-vest levels and/or exploitation rates and in the sizeor age at which fish and invertebrate populations


481

attempts to manage or mitigate the effects of cli-mate must be tightly coupled with management ofhuman behavior at all spatial and temporal scales.Most importantly, those who are or will likely beaffected by climate impacts must be made aware ofthe risks and potential consequences that futurechange will pose to their communities and theirlivelihoods.

CRUCIAL UNKNOWNS ANDRESEARCH NEEDSDirect human impacts dominate the changes occur-ring in most coastal areas,making it difficult to sepa-rate climatic effects from these direct stresses. Inmost areas,study of climate-related coastal impactsis only beginning. Consequently, for nearly allcoastal and marine ecosystems considered in thisAssessment,knowledge is inadequate to assesspotential impacts fully and to determine effectiveadaptations. The Assessment has identified siximportant areas for research related to climateimpacts on coastal and marine systems.

Coastal Hazards and the Physical Transformations ofCoastlines and Wetlands. The significant erosion ofbeach fronts,barrier islands,and coastal marshes,coupled with accelerated sea-level rise,increases thevulnerability of coastal life and property to stormsurge. Regardless of projected changes in the fre-quency and severity of coastal storms (hurricanesand nor’easters),storms will be riding on a highersea level in the future. Research and assessmentsare needed to fully evaluate the vulnerability ofhuman and natural coastal systems to the combinedeffects of sea-level rise,land subsidence,and stormsurge. This information is required for rationalresponses in coastal protection,setbacks,and mitiga-tion approaches to sustaining coastal wetlands.

Changes in Freshwater Loads to Coastal Ecosystems.Because of the importance of changes in land-usepatterns and freshwater inflow to coastal ecosys-tems,particularly estuaries and wetlands,and to keyspecies like Pacific salmon,considerable effort isneeded to improve assessments of the impact ofchanges in the extent and timing of freshwaterrunoff. While contemporary GCM estimates ofpotential runoff vary widely, it is clear that changesare likely to occur and that the impacts could besubstantial. Thus,new research is needed to assessthe consequences of changes in runoff and theattendant changes in nutrient,contaminant,and sed-iment supply, circulation,and biological processes.

are first harvested. The limiting level of exploitation(which is the rate at which the risk of populationcollapse is high) for a population is directly relatedto the rate of recruitment at low abundance levels.Thus,environmental changes that result in a reduc-tion in recruitment rates must be countered byreductions in exploitation rates. Conversely, somehigher levels of exploitation should be sustainablefor some stocks under favorable environmental con-ditions. Adaptation to changes in the compositionof fish stocks will also be necessary; regional mar-kets will undoubtedly have to adjust to shifts inspecies composition due to changes in the availabili-ty of different species. Additionally, changes in dis-tribution are likely to lead to more complex con-flicts regarding the management of transboundarystocks and species.

Although much remains to be learned in order toconfidently project impacts of climate change oncoastal areas and marine resources,the trends andrelationships already apparent suggest that the man-agers,decision makers,and the public must take cli-mate change impacts into account. To be success-ful,this will have to be done in the context ofcoastal and resource management challenges alreadybeing addressed, for example:

• Strategic adaptation of coastal communities (e.g.,barrier islands and other low-lying areas) to sea-level rise and increased storm surge.

• Adaptive management of coastal wetlands toimprove their prospects of soil building to keepup with sea-level rise and allow their migrationover adjacent lowlands.

• Comprehensive and forward-looking water useand management policies that factor in require-ments for coastal ecosystems,such as reducednutrient and pollutant delivery.

• Control procedures to reduce the risk of inva-sions by non-indigenous species.

• Fishery management regimes that incorporateknowledge of fluctuations in productivity andpopulations resulting from varying modes of cli-matic variability.

• Controls on impacts such as runoff from landand unsustainable fishing pressures that reducethe resilience of coral reef ecosystems.

In general,many of the strategies which might beappropriate for coping with future climate changehave been or are currently being discussed or imple-mented in response to current stressors on coastaland marine environments. The future impacts of cli-mate will be deeply integrated with the ongoingimpacts resulting from human activities;thus,


482

Decline of Coral Ecosystems. The decline of coralecosystems is significant and global. Contributionsto this decline include changes in ocean tempera-tures,levels of atmospheric CO2, and a series ofmore direct anthropogenic stress (e.g., over-fishing,eutrophication,and sedimentation). Increased effortis needed to adequately understand and predict thecumulative effects of these multiple stresses oncoral ecosystems. It is important in this work to rec-ognize the significance of the full ecosystem(including e.g.,sand beds,sea grasses,and the watercolumn) associated with corals,and not only thecoral reefs alone.

Alterations and Geographic Shifts in MarineEcosystems. Changes in ocean temperature and cir -culation (e.g.,ENSO, PDO, and NAO),coupled withchanges in nutrient supplies (driven by changes infreshwater fluxes and arctic ice dynamics),are likelyto modify patterns of primary productivity, the dis-tribution and recruitment success of marine fish,thereproductive success of protected species,and theeconomic viability of marine fisheries. Whileresearch on environmental variability and marineecosystems is advancing (e.g.,in GLOBEC),the cur-rent effort is limited to relatively few importantregional ecosystems. More research is needed tounderstand and predict potential changes andregional shifts for all important coastal and USmarine ecosystems,including the socioeconomicimpacts to fishing communities.

Loss of Arctic Sea Ice. Loss of sea ice in Arcticre gions will have widespread re gional impacts oncoastal env i ronments and marine ecosystems.Recent dramatic reductions in the extent of sea icein the A rctic Ocean and Bering Sea have led tom o re seve re storm surges because the larger openwater areas are capable of ge n e rating mu ch large rwave s . This has led to unprecedented ero s i o np ro blems both for Native villages and for oil andgas ex t raction infra s t ru c t u re along the Beaufo rt Seac o a s t . Reductions in sea ice also result in a loss ofc ritical habitat for marine mammals such as wa l ru sand polar bears , and significant ch a n ges in the dis-t ribution of nu t rients supporting the base of thefood we b . R e s e a rch to better understand howch a n ging ice re gimes will affect the productivity ofpolar ecosystems, and to assess the long-term con-sequences of these impacts is essential, both tosustain marine ecosystems and to develop copings t ra t e gies for the Native communities that dependon hunting for their food and other aspects oftheir culture .


483

Driscoll, N.W.,and G. H.Haug,A short circuit in thermo-haline circulation: A cause for Northern Hemisphereglaciation?, Science, 282, 436-438,1998.

Environmental Protection Agency (EPA),Naturalresource valuation: A report by the Nation’s EstuaryProgram,1997.

Environmental Protection Agency (EPA), Nationalwater quality inventory: 1996 Report to Congress,1996.URL:http://www.epa.gov/305b/

Federal Emergency Management Agency, Projectedimpact of relative sea-level rise on the National FloodInsurance Program,Report to Congress, FederalEmergency Management Agency,Washington,DC,Federal Insurance Administration,1991.

Frank,K.T.,R.I. Perry, and K. F. Drinkwater, Predictedresponse of Northwest Atlantic invertebrate and fishstocks to CO2-induced climate change, Transactions ofthe American Fisheries Society, 119, 353-365,1990.

Gattuso, J. P., D.Allemand,and M. Frankignoulle,Photosynthesis and calcification at cellular, organismal,and community levels in coral reefs: A review on inter-actions and control by the carbonate chemistry,American Zoologist, 39(1),160-183,1999.

Good, J.W.,Shore protection policy and practices inOregon: An evaluation of implementation success,Coastal Management, 22, 335-352,1994.

Gornitz,V. M.,Sea-level rise: A review of past and near-future trends, Earth Surface Processes and Landforms,20, 7-20,1995.

Gornitz,V. M.,R.C.Daniels,T.W.White,and K.R.Birdwell,The development of a coastal risk assessmentdatabase: Vulnerability to sea-level rise in the USSoutheast,in Coastal Hazards: Perception,Susceptibility, and Mitigation, edited by C. W. Finkl,Jr.,Journal of Coastal Research, Special Issue 12,pp.327-338,1994.

Griggs, G. B.,and K.M.Brown,Erosion and shorelinedamage along the Central California Coast: A compari-son between the 1997-1998 and 1982-1983 ENSO win-ters, Shore and Beach, 66(3),18-23,1999.

Hargreaves, N. B., D. M.Ware,and G.A.McFarlane,Return of the Pacific sardine (Sardinops sagax) to theBritish Columbia coast in 1992, Canadian Journal ofFisheries and Aquatic Sciences, 51, 460-463,1994.

Harwell,M.A.,Science and environmental decisionmaking in South Florida, Ecological Applications, 8(3),580-590,1998.

Henderson-Sellers,A.,H.Zhang, G. Berz,K.Emanuel,W.Gray, C.Landsea, G. Holland, J. Lighthill,S.-L.Shieh, P.Webster, and K.McGuffie.Tropical cyclones and global

LITERATURE CITEDAlexander, C.E.,Classified shellfish growing waters,inNOAA’s State of the Coast Report, National Oceanicand Atmospheric Administration (NOAA),Silver Spring,Maryland,1998. URL:http://state-of-coast.noaa.gov/topics/html/state.html

Boesch,B. F.,M. N. Josselyn,A. J. Mehta, J.T. Morris,W. K.Nuttle,C.A.Simenstad,and D. J. P. Swift,Scientificassessment of coastal wetland loss, restoration,andmanagement in Louisiana, Journal of CoastalResearch, Special Issue 20,1994.

Bricker, S.B.,C. G. Clement, D. E.Pirhall,S. P. Orlando,and D. R. G. Farrlow, National estuarine eutrophicationassessment: A summary of conditions,historical trends,and future outlook,National Oceanic and AtmosphericAdministration (NOAA),Silver Spring,Maryland,1999.

Broecker,W. S.,Thermohaline circulation,the AchillesHeel of our climate system: Will man-made CO2 upsetthe current balance?, Science, 278, 1582-1588,1997.

Broecker,W. S.,The great ocean conveyor,Oceanography, 4(2),79-89,1991.

Broecker,W. S.,S.Sutherland,and T. H. Peng,A possible20th century slowdown of southern ocean deep waterformation, Science, 286, 1132-1135,1999.

Buddemeier, R.W.,and S.V. Smith,Coral adaptation andacclimatization: A most ingenious paradox, AmericanZoologist, (39)1,66-79,1999.

California Department of Fish and Game, Pacific sar-dine resource now officially recovered, press release,January 15,1999.

Cane,M.A.,A.C.Clement,A.Kaplan,Y. Kushnir, D.Pozdnyakov, R.Seager, S.E.Zebiak,and R.Murtugudde,Twentieth-century sea surface temperature trends,Science, 275, 957-960,1997.

Costanza,R.,The ecological,economic,and socialimportance of the oceans, Ecological Economics, 31,199-213,1999.

C o s t a n z a ,R . , et al.,The value of the wo r l d ’s ecosystems e rvices and natural capital, N a t u re , 3 8 7 , 2 5 3 - 2 6 0 ,1 9 9 7 .

Culliton,T. J., Population: Distribution,density andgrowth, NOAA state of the coast report,NationalOceanic and Atmospheric Administration (NOAA),Silver Spring,Maryland,1998. URL:http://state-of-coast.noaa.gov/topics/html/pressure.html

Dolan,R., F.Anders,and S.Kimball,Coastal erosion andaccretion in US, Geological Survey National AtlasDepartment of Interior, US Geological Survey, Reston,Virginia,1985.


484

climate change:A post-IPCC assessment Bulletin of theAmerican Meteorological Society, 79:19-38.1998.

Herbert, P. J., J. D. Jarrell,and M.Mayfield,The deadliest,costliest,and most intense hurricanes of this century(and other frequently requested hurricane facts),NOAATechnical Memorandum NWS TPC-1,NationalHurricane Center 1996.

Hoegh-Guldberg, O.,Climate change,coral bleaching,and the future of the world’s coral reefs, Marine andFreshwater Research, 50(8),839-866,1999.

H o l t , G. , Marlin hooked in state’s ‘ t ro p i c a l ’ wa t e rs , S e a t t l ePost Intellige n c e r, p .C 1 , Fri d ay, September 5, 1 9 9 7 .

Houston, J. R.,International tourism and US beaches,Shore and Beach, 64(3),27-35,1996.

Intergovernmental Panel on Climate Change (IPCC),Climate Change 1995: Impacts, Adaptations, andMitigation of Climate Change: Scientific-TechnicalAnalysis, Cambridge University Press,Cambridge,United Kingdom,1996.

Jo bl i n g ,M . 1 9 9 6 . Te m p e ra t u re and grow t h : M o d u l a t i o nof growth rate via tempera t u re ch a n ge , in G l o b a lWa rm i n g : Implications for Marine and Fre s h wa t e rFi s h , edited by C. M .Wood and D. G. M a c D o n a l d ,p p .2 2 5 - 2 5 3 ,C a m b ri d ge Unive rsity Pre s s .1 9 9 6 . .

Johannessen, O. M.,E.V. Shalina,and M.W. Miles,Satelliteevidence for an Arctic sea ice cover in transformation,Science, 286, 1937-1939,1999.

Justic´, D., N. N. Rabalais,and R.E.Turner, Effects of cli-mate change on hypoxia in coastal waters: A doubledCO2 scenario for the northern Gulf of Mexico,Limnology and Oceanography, 41, 992-1003,1996.

Kaminsky, G. M., P. Ruggeiero,and G. Gelfenbaum,Monitoring coastal change in Southwest Washingtonand Northwest Oregon during the 1997-1998 El Niño,Shore and Beach, 66(3),42-51,1999.

Karl,T. R.,R.W. Knight,and N. Plummer,Trends in high-frequency climate variability in the twentieth century,Nature, 377, 217-220,1995a.

Karl,T. R.,R.W. Knight, D. R.Easterling,and R. G. Quayle,Indices of climate change for the United States,Bulletin of the American Meteorological Society, 77,279-292,1995b.

Kerr, E.A.,Thermodynamic control of hurricane intensi-ty, Nature, 401, 665-669,1999.

Kleypas, J.A.,R.W. Buddemeier, D.Archer, J. P. Gattuso,C.Langdon,and B. N. Opdyke, Geochemical conse-quences of increased atmospheric carbon dioxide oncoral reefs, Science, 284(2),118-120,1999a.

Kleypas, J.A., J.W. McManus,and L.A.B.Menez,Environmental limits to coral reef development: Wheredo we draw the line?, American Zoologist, 39, 146-159,1999b.

Knutson,T. R.,R.E.Tuleya,and Y. Kurihara,Simulatedincrease of hurricane intensities in a CO 2-warmed cli-mate, Science, 279, 1018-1020,1998.

Komar, P. D.,The 1997-1998 El Niño and erosion on theOregon coast, Shore and Beach, 66(3),33-41,1999.

Knutson,R.R.,R.E.Tuleya,and Y. Kurihara,Simulatedincrease of hurricane intensities in a CO 2 warmed cli-mate, Science, 279, 1018-1020,1998.

Kronman,M.,Market report: Year in review, NationalFisherman, p.40,April 1999.

Landsea,C.W., N. Nicholls,W. M.Gray and L.A.Avila,Downward trends in the frequency of intense Atlantichurricanes during the past five decades,GeophysicalResearch Letters, 23, 1697-1700,1996.

Levitus,S., J. I.Antonov,T. P. Boyer, and C.Stephans,Warming of the world ocean, Science, 287, 2225-2229,2000.

Lluch-Belda, D. S.,H. Vasquez, D. B.Lluch-Cota,C.A.Salinas-Zavela,and R.A.Schwartzlose,The recovery ofthe California sardine as related to global change,California Cooperative Oceanic FisheriesInvestigations Reports, 33, 50-59,1992.

Macklin,S.A.,Report on the Fisheries OceanographyCoordinated Investigations International Workshop onRecent Conditions in the Bering Sea,National Oceanicand Atmospheric Administration ERL.Contribution No.2044 from NOAA/Pacific Marine EnvironmentalLaboratory, Contribution B358 from Fisheries-Oceanography Coordinated Investigations. 1999.

Mantua, N. J.,S.R.Hare,Y. Shang, J. M.Wallace,and R.C.Francis,A Pacific interdecadal climate oscillation withimpacts on salmon production, Bulletin of theAmerican Meteorological Society, 78, 1069-1079,1997.

Markham,A., Potential impacts of climate change onecosystems: A review of implications for policymakersand conservation biologists, Climate Research, 6(2),179-191,1996.

Maslanik, J.A.,M.C.Serreze,and T.Agnew, On therecord reduction in 1998 Western Arctic Sea-ice cover,Geophysical Research Letters, 26, 1905-1908,1999.

Mathews-Amos,A.,and E.A.Berntson,Turning up theheat: How global warming threatens life in the sea,Marine Conservation Biology Institute,1999. URL:http://www.mcbi.org/


485

Reed, D.J. The response of coastal marshes to sea-levelrise:Survival of submergence. Earth Surface Processesand Landforms, 20:39-48.1995.

Reid,W.V.,and M.C.Trexler, Responding to potentialimpacts of climate change on US coastal biodiversity,Coastal Management, 20(2),117-142,1992.

Rice, J., Food web theory, marine food webs,and whatclimate change may do to northern marine fish popula-tions,in Climate Change and Northern FishPopulations,edited by R. J. Beamish, Canadian SpecialPublication of Fisheries and Aquatic Sciences, 121,1995.

Rothrock, D.A.,Y.Yu,and G.A.Maykut,Thinning of theArctic Sea-ice cover, Geophysical Research Letters,26(23),3469-3472,1999.

Sagarin,R. D., J. P. Barry, S.E.Gilman,and C.H.Baxter,Climate-related change in an intertidal community overshort and long time scales. Ecological Monographs,69(4),465-490,1999.

Sarmiento, J. L.,and C.Le Quere,Oceanic carbon diox-ide uptake in a model of century-scale global warming,Science, 274, 1346-1350,1996.

Schmittner,A.,and T. F. Stocker,The stability of the ther-mohaline circulation in global warming experiments,Journal of Climate, 12, 1117-1134,1999.

Strong,A.E.,E. J. Kearns and K.K.Gjovig,Sea surfacetemperature signals from satellites: An update,Geophysical Research Letters, 27(11),1667-1670,2000.

Taylor, K.,Rapid climate change, American Scientist,87, 320-327,1999.


Titus, J. G.,and V. K.Narayanan,The risk of sea-level rise,Climatic Change, 33, 151-212,1996.

US Army Corps of Engineers,Central and SouthernFlorida Project (C&SF) comprehensive review study,Draft integrated feasibility report and programmaticenvironmental impact statement, US Army Corps ofEngineers, Jacksonville District, Jacksonville,Florida,1998.

Veron, J. E. N., Corals in Space and Time. CornellUniversity Press,Ithaca1995.

Vinnikov, K.Y.,A.Robock,R. J. Stouffer, J. E.Walsh,C.L.Parkinson, D. J. Cavalieri, J. F. B.Mitchell, D. Garrett,andV. F. Zakharov, Global warming and NorthernHemisphere sea ice extent, Science, 286, 1934-1937,1999.

McGowan, J.A., D. R.Cayan,and L.M.Dorman,Climate,ocean variability, and ecosystem response in theNortheast Pacific, Science, 281, 210-217,1998.

Miller, M.L.,and J.Auyong,Coastal zone tourism: Apotent force affecting environment and society, MarinePolicy, 15(2),75-99,1991.

Mitsch,W. J.,and J. G. Gosslink, Wetlands: Second edi -tion, Van Norstrand Reinhold,New York,1993.

Murowski,S.A.,Climate change and marine fish distri-butions: Forecasting from historical analogy,Transactions of the American Fisheries Society, 122,657-658,1993

National Marine Fisheries Service.1998.Status ofFisheries of the United States:A Report to Congress.National Oceanic and Atmospheric Administration,Silver Spring,MD. 88pp.

National Research Council, Sustaining Marine Fi s h e ri e s ,National A c a d e my Pre s s ,Wa s h i n g t o n ,D C ,1 9 9 9 .

National Research Council, Striking a Balance:Improving Stewardship of Marine Areas, NationalAcademy Press,Washington DC,1997.

Neumann, J. E., G.Yohe,R.Nicholls,and M.Maino, Sea-level rise and global climate change: A Review ofImpacts to US Coasts, Pew Center on Global Climate,2000. URL:http://www.pewclimate.org/projects/env_sealevel.html

National Oceanic and Atmospheric Administration(NOAA),Year of the Ocean discussion papers,1998.

NPA Data Services,Inc.,Analytic documentation ofthree alternate socioeconomic projections,1997-2050,Washington,DC,May 1999.

Oppenheimer, M.,Global warming and the stability ofthe West Antarctic ice sheet,Nature, 393, 325-332,1998.

Pielke,R.A.,Jr., and C.W. Landsea,La Niña,El Niño,andAtlantic hurricane damages in the United States,Bulletin of the American Meteorological Society,80(10),pp.2027-2033,1999.

Pugh, D.T.,and G.A.Maul,Coastal sea level predictionfor climate change,in Coastal and Estuarine Studies56: Coastal Ocean Prediction, edited by C.N.K.Mooers,American Geophysical Union,Washington D.C.pp.377-404,1999.

Rabalais, N. N.,R.E.Turner, D. Justic´, Q. Dortch,and W.J.Wiseman,Jr., Characterization of hypoxia, NOAACoastal Ocean Program Decision Analysis Series No.15, National Oceanic and Atmospheric Administration,Silver Spring,Maryland,1999.


486

Vitousek, P. M., J.Aber, S.E.Bayley, R.W. Howarth, G. E.Likens, P.A.Matson, D.W. Shindler,W. H.Schlesinger, andG. D.Tilman,Human Alteration of the global nitrogencycle: Sources and consequences, EcologicalApplications, 7(3),737-750,1997.

Welch, D.W.,Y. Ishida,and K.Nagasawa,Thermal limitsand ocean migrations of sockeye salmon(Onchorhynchus nerka): Long term consequences ofglobal warming, Canadian Journal of Fisheries andAquatic Sciences, 55, 937-948,1998.

Wigley,T. M.L., The Science of Climate Change:Global and US Perspectives, Pew Center on GlobalClimate Change,Arlington, Virginia,1999.

Wilkinson,C., O. Linden,H.Cesar, G. Hodgson, J.Rubens,and A.E.Strong,Ecological and socioeconomicimpacts of 1998 coral mortality in the Indian Ocean:An ENSO impact and a warning of future change?,Ambio, 28(2),188-196,1999.

Wilkinson,C.R.,and R.W. Buddemeier, Global climatechange and coral reefs: Implications for people andreefs,Report of the UNEP-IOC-ASPEI-IUCN Global TaskTeam on the Implications of Climate Change on CoralReefs, International Union for the Conservation ofNature.Gland,Switzerland,1994.

Wolock, D.M.and G.J. McCabe.Simulated effects of cli-mate change on mean annual runoff in the contermi-nous United States. Journal of the American WaterResources Association, 35: 1341-1350,1999.

Wong, J.,Thousands of seals,sea lions starving, SeattleTimes, pp.A8,Thursday, December 11,1997.

Wood,R.A.,A.B. Keen, J. F.B.Mitchell,and J. M.Gregory,Changing spatial structure of the thermohaline circula-tion in response to atmospheric CO2 forcing in a cli -mate model, Science, 399, 572-575,1999.

Yohe, G., J. Neumann, P. Marshall,and H.Ameden,Theeconomic cost of greenhouse induced sea-level rise fordeveloped property in the United States, ClimaticChange, 32, 387-410,1996.

Zhang,K.,et al.,East Coast storm surges provideunique climate record, Eos, Transactions of theAmerican Geophysical Union, 78(37),389,396-97,1997.

ACKNOWLEDGMENTSSpecial thanks to:

Jim Allen,Paul Smith CollegeDonald Cahoon,US Geological SurveyBenjamin Felzer, National Center for Atmospheric

ResearchMichael MacCracken,US Global Change Research

ProgramLaShaunda Malone,US Global Change Research

ProgramMelissa Taylor, US Global Change Research Program Rob Thieler, US Geological SurveyElizabeth Turner, National Oceanic and Atmospheric

AdministrationJustin Wettstein,US Global Change Research

Program


487

489

CHAPTER 17

POTENTIAL CONSEQUENCES OF CLIMATEVARIABILITY AND CHANGE FOR THEFORESTS OF THE UNITED STATESLinda Joyce1,2, John Aber3, Steve McNulty1, Virginia Dale4, Andrew Hansen5, LloydIrland6, Ron Neilson1, and Kenneth Skog1


Chapter Summary

Background

Climate and Forests

Key Issues

Forest Processes

Forests and Disturbance

Biodiversity

Socioeconomic Impacts

Adaptations: Forest Management Strategies under Climate Change


Literature Cited

Acknowledgments

1USDA Forest Service; 2Coordinating author for the National Assessment Synthesis Team; 3University of New

Hampshire, 4Oak Ridge National Laboratory; 5Montana State University, 6The Irland Group

490


CHAPTER SUMMARY

Context

Forests cover nearly one-third of the US,providingwildlife habitat, clean air and water, cultural and aes-thetic values,carbon storage, recreational opportuni-ties such as hiking,camping, fishing,and autumnleaf tours,and products that can be harvested suchas timber, pulpwood,fuelwood,wild game, ferns,mushrooms,and berries. This wealth depends onforest biodiversity—the variety of plants,animals,and microbe species,and forest functioning—waterflow, nutrient cycling,and productivity. Theseaspects of forests are strongly influenced by climateand human land use.

491

Chapter 17 / Forests

to predict; climate changes, changes in these distur-bances and their effects on forests are possible.

Analyses of the results of ecological models whendriven by several different climate scenarios indi-cate changes in the location and area of potentialhabitats for many tree species and plant communi-ties. For example,alpine and subalpine habitats andthe variety of species dependent upon them arelikely to be greatly reduced in the conterminous US.The ranges of some trees are likely to contract dra-matically in the US and largely shift into Canada.Expansion of potential habitats is possible foroak/hickory and oak/pine in the eastern US andPonderosa pine and arid woodland communities inthe West. How well plant and animal species adaptto or move with changes in their potential habitat isstrongly influenced by their dispersal abilities andthe disturbances to these environments. Introducedand invasive species that disperse rapidly are likelyto find opportunities in newly forming communi-ties.

The effects of climate change on socioeconomicbenefits obtained from forests will likely be influ-enced greatly by future changes in human demands,as determined by population growth,increases inincome, changing human values,and consumer pref-erences.

Outdoor recreation opportunities are very likely tobe altered by climate change. For example, warmerwaters would increase fish production and opportu-nities for some species;but decrease opportunitiesfor cold water species. Outdoor recreation opportu-nities in mountainous areas are likely to be altered.Summer recreation opportunities are likely toexpand in the mountainous areas when warmerlowland temperatures attract more people to higherelevations. Skiing opportunities are likely bereduced with fewer cold days and snow events. Inmarginal climate areas,the costs to maintain down-hill skiing opportunities are likely to rise whichwould possibly result in the closure of some areas.

Key Findings

Carbon storage in US forests is currently estimatedto increase from 0.1 to 0.3 Pg of carbon per yearand analyses suggest that carbon dioxide fertiliza-tion and land use have influenced this current stor-age. Within the next 50 years, forest productivity islikely to increase with the fertilizing effect of atmos-pheric carbon dioxide. Those productivity increasesare very likely to be strongly tempered by local envi-ronmental conditions (e.g.,moisture stress, nutrientavailability) and by human land use impacts such asforest fragmentation,increased atmospheric deposi-tion,and tropospheric ozone.

Economic analyses when driven by several differentclimate scenarios indicate an overall increase in for-est productivity in the US that is very likely toincrease timber inventory, subject to other externalforces. With more potential forest inventory to har-vest,the costs of wood and paper products to con-sumers are likely to decrease,as are the returns toowners of timberland. The changes in climate andconsequent impact on forests are likely to changemarket incentives to harvest and plant trees,andshift land uses between agriculture and forestry.These changes will likely vary within a region.Market incentives for forestry are likely to moderatesome of the climate-induced decline in the area ofnatural forests. International trade in forest prod-ucts could either accentuate or dampen priceeffects in the US,depending on whether forest har-vest activity increases or decreases outside the US.

Over the next century, changes in the severity, fre-quency, and extent of natural disturbances are possi-ble under future climate change. These changes innatural disturbances then impact forest structure,biodiversity, and functioning. Analyses of the resultsfrom climate and ecological models suggest that theseasonal severity of fire hazard is likely to increaseby 10% over much of the US,with possibly largerincreases in the southeastern US and Alaska andactual decreases in the Northern Great Plains.Although the interactions between climate changeand hurricanes,landslides,ice storms,wind storms,insects,disease,and introduced species are difficult

The remaining forestland is in various federal,state,county, and municipal ownerships. Over 50% of thefederal land in forests is managed by the US ForestService. The largest state owner of forestland isAlaska. County and municipal ownerships compriseless than 4% of the total forestland. However, over2.5 million acres of forestland are managed by theselocal governments in Minnesota alone. Over 80% ofthe forestland in federal ownership is found in thewest,while most of the forestland held by states isin the northern part of the US. Of the 472 millionacres in private ownership, over 60% is found in theeastern part of the US. Forest industry ownersaccount for 14% of the forestland in private owner-ship,and these lands are found mainly in the south-ern part of the US (54%).

Forests are an environment in which people recre-ate,such as the National Forests and Parks,and anenvironment in which people live,such as the NewEngland woods,and the conifer forests of RockyMountains. Forests provide recreational opportuni-ties such as hiking,camping, fishing,hunting,birdwatching,downhill and cross-country skiing,andautumn leaf tours. In addition,many rivers andstreams flowing through forests provide fishing,boating,and swimming recreational opportunities.Activities associated with recreation provide incomeand employment in every forested region of the USand Canada (Watson et al.,1998). Forests provideclean air and water, watershed and riparian buffers,moderate streamflow, and help to maintain aquatichabitats. New York City’s water supply is derivedfrom water collected within forested watersheds ina 2,000 square mile area. Forests provide wildlifehabitat. Flather et al.(1999) report that at least 90%of the resident or common migrant vertebratespecies in the US rely on forest habitats for part oftheir life requisites. Forests also provide culturaland aesthetic values,carbon storage,and productsthat can be harvested such as timber, pulpwood,fuelwood,wild game,edible plants,fruits and nuts,mushrooms,and floral products. This wealth fromforests depends on forest biodiversity – the varietyof plant and animal species – and forest function-ing—water flow, nutrient cycling and productivity.These aspects of forests are strongly influenced byclimate.


BACKGROUNDCovering nearly one-third of the US, forests are anintegral part of the vegetative cover of the nation’slandscape (Figure 1). The total area in forests in theUS is 747 million acres,which is about 7% of theforestland in the entire world. US forests are distrib-uted in the eastern and western parts of the US,with small stands of forests in the central part locat-ed mainly along rivers and streams. The white-red-jack pine forests of New England have supported atimber industry and the northeastern deciduousforests of maples,beeches,birches,and oaks providethe colorful autumn landscapes for tourists. TheMid-Atlantic region is rich in tree species from pineforests and coastal wetlands to northern uplandhardwoods such as oak-hickory and maple-beech-birch forest types. The southern forests are also amix of conifer and deciduous forests. Westernforests are primarily conifer forests such as Douglas-fir, fir-spruce, Ponderosa pine and piñon-juniper.

Ownership of forestland varies by region in the US.Over 63% of US forestland is in private ownership.

Current Distribution of Forests in the United States

douglas-firhemlock-sitka spruceponderosa pinewestern white pinelodgepole pinelarchfir-spruceredwoodchaparralpinyon-juniperwestern hardwoodsaspen-birch

white-red-jack pinespuce-firlongleaf-slash pineloblolly-shortleaf pineoak-pineoak-hickoryoak-gum-cypresselm-ash-cottonwoodmaple-beech-birchaspen-birch

Western Forests Eastern Forests

Figure 1. Map of forest vegetation types for theUnited States. Map is from the USDA Forest Service,http://www.srsfia.usfs.msstate.edu/rpa/rpa93.ht SeeColor Plate Appendix

POTENTIAL CONSEQUENCES OF CLIMATEVARIABILITY AND CHANGE FOR THE FORESTS OF THE UNITED STATES

492

493


Climate change is one of several pressures onforests encompassed under the broader term,globalchange. Human activities have altered the vegeta-tion distribution of forests in the US. The arrival ofEuropeans along the eastern coast initiated the har-vest of forests. For example,eastern white pineswere highly prized as ship masts for the EnglishNavy. From 1600 to the mid 1800s,the nativeforests in the eastern US were extensively harvestedfor wood products as well as to clear land for agri-culture and urban uses (Figure 2). While total forest-land area has stabilized in the US since the early1900s,land use shifts still occur where forestland isconverted to agricultural or urban use and agricul-tural or pasture is planted to trees. In some parts ofthe eastern US,new forests have regrown on aban-doned agricultural lands,although forestland in thenortheastern part of the US is still less than 70% ofits original extent in the 1600s. Forestland in theSouth occupies less then 60% of the 1600s extent offorestland,and the establishment of pine plantationshas placed some 20% of the forestland in this regionunder intensive management. While western forestshave remained relatively constant in area since the1600s, recent expansion of urban areas and agricul-ture is fragmenting them. Across the Nation,urbanareas have continued to increase (Flather et al.,1999). This expansion of human influences into therural landscape alters disturbance patterns associat-ed with fire, flooding,and landslides. In addition,human activities can result in the dispersal of pollu-tants into forests. Increasing atmospheric concen-trations of ozone and deposition of nitrogen (N)compounds have profound effects on tree photosyn-thesis, respiration, water relations,and survival.

Human activities modify the species composition offorests and the disturbance regimes associated withforests. Fire suppression has altered southeastern,midwestern,and western forests. Harvesting meth-ods,such as clearcutting,shelterwood,or individualtree selection,have changed species composition innative forests. The average age of forest stands inthe East is less than the average for the West, reflect-ing the extensive harvesting as the eastern US wassettled. Intensive management along with favorableclimates in parts of the US has resulted in highlyproductive forests that are maintained for timberproduction,such as the southern pine plantations.Native and introduced insects and disease species,such as gypsy moth, chestnut blight,Dutch elm dis-ease,have altered US forests (Ayres and Lombardero,2000).Trees have been planted far outside their nat-ural ranges for aesthetic and landscaping purposesin urban and rural areas.

Population levels,economic growth,and personalpreferences influence the socio-economic valuesassociated with forests,and consequently theresources demanded from forests. Per capita con-sumption of wood in the US has been relatively sta-ble over the last several decades and the futuredemand for wood products is projected to followpopulation growth over the next 50 years.Technological development and consumer prefer-ences strongly affect the demand for specific woodand paper products. For example,consumer prefer-ence has influenced the increasing amount of recy-cled material used in fiber over the last severalyears. Though recreational hunting has been declin-ing,the economic impact is significant;Flather et al.(1999) estimated that, for all wildlife,hunters alonespent nearly 21 billion dollars on equipment andtravel in 1996. Other products harvested fromforests are more difficult to assign an economicvalue or to project future demand. Blatner (1997)estimates the harvest of mushrooms fromWashington,Oregon,and Idaho forests in 1992 to bevalued at about $40 million.

Figure 2. Land area changes in forestland. Data are from ForestService Resource Bulletin PNW-RB-168, Forest Resource ReportNo. 23, No. 17, No. 14, the Report of the Joint Committee onForestry, 77th Congress 1st Session, Senate Document No. 32.Data for 1850 and 1870 were based on information collected duringthe 1850 and 1870 decennial census; data for 1907 were also basedon the decennial census modified by expert opinon, reported byR.S. Kellogg in Forest Service Circular 166. Data for 1630 wereincluded in Circular 166 as an estimate of the original forest areabased on the current estimate of forest and historic land clearinginformation. These data are provided here for general referencepurposes only to convey the relative extent of the forest estate inwhat is now the US at the time of European settlement. See ColorPlate Appendix

Forest Land Coverage over the Past 400 Years

494


ty development. Temperature affects fruit and seedyields and quality by influencing factors such as flow-ering,bud dormancy, and ripening of fruit and cones(Kozlowski and Pallardy, 1997). Changes in air tem-perature in autumn and spring can also affect harden-ing and dehardening of tree needles (Guak et al.,1998).Temperature affects ecosystem-level processessuch as soil decomposition and mineralization.Indirect effects associated with warming could belarger than the direct effect on plant growth in sub-polar biomes where warming of permafrost is likely(Mooney et al.,1999).

Shortages or excesses of water offset the rates ofmost important processes controlling the biogeo-chemistry of the major nutrients. In particular, forest-ed wetlands are sensitive to changes in hydrologicregimes. While forest ecosystems generally occupythose regions with low annual water deficits, waterlimitation in space and time is still critical for overallcarbon balances,and is one of the major driversembedded in the models used to predict globalchange effects.

KEY ISSUESThe vulnerability of forests to climate variability andchange is examined by looking at four key issues: for-est processes,disturbance,biodiversity, and socioeco-nomic benefits.

• Forest processes regulate the flux and apportion-ment of carbon, water, nutrients,and other con-stituents within a forest ecosystem. These process-es operate at spatial scales from leaf to landscapeand control responses of forest ecosystems,suchas forest productivity, to environmental factorssuch as temperature,precipitation,and atmospher-ic concentrations of CO2.

• Forests are subjected to disturbances that arethemselves strongly influenced by climate. Thesenatural disturbances include fire,hurricanes,land-slides,ice storms,wind storms,drought,insects,disease,and introduced species.

• Biodiversity refers to the variety of populations,species,and communities. Climate influences thedistribution and abundance of plant and animalspecies through food availability, habitat availabili-ty, and survivorship.

• Forest processes and forest biodiversity areuniquely capable of providing goods such as woodproducts,wild game,and harvested plants,ecologi-cal services such as water purification,and ameni-ties such as scenic vistas and wilderness experi-ences—the socioeconomic benefits of forests.

At the global scale,population,economic growth,and personal preferences influence the demandsfrom forests. Wood consumption rose 64% globallysince 1961. This increase was strongly influencedby rising wood per capita consumption for fuel-wood,paper products,and industrial fiber(Matthews and Hammond,1999). More than half ofthe wood fiber produced globally is consumed asfuel in contrast to the US,where fuelwood compris-es around 14% of total wood fiber consumed(Brooks,1993;Matthews and Hammond,1999). Thedemands on forests globally are expected to changeas the world’s populations become increasinglyurban. More industrial products and environmentalservices are likely to influence the management ofthe world’s forests (Brooks,1993).

CLIMATE AND FORESTSClimate influences the composition,structure,andfunction of forest ecosystems,the amount and quali-ty of forest resources,and the social values associat-ed with forests. Native forests are adapted to localclimatic features. Where summer drought is typicalin the Pacific Northwest,native conifer and hard-wood forests have water-conserving leaves (Shrineret al.,1998). Black spruce and white spruce arefound in the cold-tolerant boreal forests where win-ter temperature extremes can reach –62˚ to –34˚C(-79˚ to –30˚F) (Burns and Honkala,1990). Thepiñon-juniper forests of the Southwest are drought-adapted.

Changes in the distribution and abundance ofplant or animal species reflect the birth, growth,death,and dispersal rates of individuals in a popu-lation. Climate and soil are strong controls on theestablishment and growth of plants. Climate influ -ences the distribution and abundance of animalspecies through changes in resource availability,fecundity, and survivorship (Hansen and Rotella,1999). Changes in disturbance regimes,and com-petitive and cooperative interactions with otherspecies also affect the distribution of plants andanimals. In addition,human activities influencethe occurrence and abundance of species on thelandscape.

Temperature and Precipitation. The spatial and tem-poral distribution of water and temperature are theprimary determinants of woody plant distributionsover the Earth. Air temperature affects physiologicalprocesses of individual plants,and over the long-term,the environmental conditions for seed devel-opment and germination,population,and communi-

495


of nutrient cations (calcium,magnesium,and potas-sium). Low availability of N in the soil limits forestproduction. Increases in forest growth in responseto N deposition have been reported both inScandinavia and the US,although the responsevaries between deciduous and coniferous species(Magill et al.,1997;Aber et al.,1998;Magill et al.,1999).

Ozone is a highly reactive gas. Closely associatedwith the combustion of fossil fuels,ozone is formedin the lower atmosphere (ground-level ozone)through chemical reactions between nitrogenoxides and hydrocarbons in the presence of sun-light. High levels of ozone occur, generally in sum-mer, when warm,stagnant air masses over denselypopulated and highly industrialized regions accumu-late large quantities of nitrogen oxides and hydro-carbons. Thus, the distribution of ozone concentra-tions is very ir regular in space and time.

Ozone leaves the atmosphere through reactionswith plants and soil surfaces along a number ofphysical and chemical degradation pathways.Ozone concentrations can dissipate in a matter ofdays when air masses move away from urban areasthroughout the eastern US,or from western citiesacross more remote forested areas,such as from LosAngeles to the San Bernardino and San GabrielMountains,California. Ozone tends to remain in theatmosphere when it cannot react with material (e.g.vegetation or soils). Thus,ozone concentrations ineastern Maine can be as high as over urban Bostonbecause ozone-bearing air masses have reachedthese remote areas with little loss of ozone as theypassed over the ocean. High concentrations canoccur in relatively remote mountaintop locationsbecause ozone also accumulates at the top of theatmospheric boundary layer where contact withvegetated surfaces is minimal.

U n l i ke nitrogen and S deposition,w h i ch by theirn a t u re are slow and cumu l a t i ve , the effects of ozoneon ve getation are direct and immediate, as the pri-m a ry mechanism for damage is through direct plantu p t a ke from the atmosphere through stomata (smallopenings in the plant leaf through which water andgases pass into and out of the plant). Ozone is as t rong oxidant that damages cell membra n e s ;t h eplant must then expend energy to maintain theses e n s i t i ve tissues. The net effect is a decline in netphotosynthetic ra t e . The degree to which photosyn-thesis is reduced is a function both of dose andspecies conductance ra t e s , the rate at which leave sex ch a n ge gases with the atmosphere (Musselmanand Massman, 1 9 9 9 ) . A n a lyses suggest that curre n t

Changes in these goods,services,and amenitiesare influenced by changes in factors that deter-mine their supply — land area,productivity, man-agement,production technology, quality ofamenities — and their demand — human popu-lation leve l s , economic grow t h , and personal pre f-e re n c e s .

1. Forest Processes

Current Environmental Changes Include Depositionof Nitrogen and Ozone.These environmental changes influence the abilityof forests to respond to changes in climate (Aber etal.,2001). Tree species have been shown to be sen-sitive to air pollutants such as ozone and sulfur (S)(Fox and Mickler, 1996;Taylor et al.,1994;US EPA,1996) and nitrogen (N) deposition has been linkedto soil acidification and cation depletion in forests(Aber et al.,1995;Aber et al.,1998). Total depositionof N and associated acidity in the US have increasedas much as 10-fold over global background levels asa result of human activity (Galloway, 1995;Vitouseket al.,1997;Matthews and Hammond,1999).Combustion of fossil fuels injects N and S into theatmosphere as simple oxides. The N and S oxidesare retained in the atmosphere only for days toweeks,whereas carbon dioxide (CO2) is retained fordecades. Some N and S compounds can be re-deposited on the forest surface either in a dry formor, by dissolution into water, in a wet form (“acidrain”) (Boubel et al.,1994). The shorter residencetime of N and S in the atmosphere results in anintensely regional distribution of deposition,withthe eastern US,and especially the Northeast, experi-encing the highest levels of both S and N(NADP/NTN, 1997).

With the large reduction in S deposition in the lastdecade from the controls imposed by the Clean AirAct of 1990,the importance of the acids in precipi-tation has been reduced. Sources of N vary region-ally. Urban areas contribute to increased N deposi-tion in pine forests in southern California (Fenn etal.,1996). Nitrogen deposition as ammoniumoccurs where fertilizer is applied intensively orwhere livestock are concentrated in feedlots(Lovett,1994). Agricultural lands and feedlots alongthe Front Range of the Rocky Mountains inColorado contribute to increased N deposition inthe alpine and forest ecosystems in Rocky MountainNational Park (Baron et al.,1994;Musselman et al.,1996;Williams et al.,1996).

High levels of N deposition can result in negativeeffects such as soil acidification,causing depletion

496


ozone levels have decreased production 10% inn o rt h e a s t e rn fo rests (Ollinger et al., 1997) and 5% ins o u t h e rn pine plantations (Weinstein et al., 1 9 9 8 ) .Wa rming of surface air, a consistent fe a t u re of theH a d l ey and Canadian scenarios used in this assess-m e n t , is like ly to increase ozone and other air-qualityp ro blems (see Watson et al., 1996 for analysis of pre-vious climate scenari o s ) ,f u rther increasing the stre s son fo rests in areas where air quality is compro m i s e d .

Impacts of Elevated Atmospheric Carbon Dioxideand Climate Change on Forest Processes

Experimental exposure of trees to elevated atmos-pheric CO2 has shown significant changes in physi-ological processes, growth,and biomass accumula-tion (Aber et al.,2001;Mooney et al.,1999). Over awide range of CO2 concentrations,photosynthesishas been increased in plants representative of mostnorthern temperate forests (Eamus and Jarvis,1989;Bazzaz,1990;Mohren et al.,1996;Long et al.,1996;Kozlowski and Pallardy, 1997). Reviews of theextensive CO2-enrichment studies have shown vari-able but positive responses in plant biomass accu-mulation (Ceulemans and Mousseau,1994;Saxe etal.,1998;Mooney et al.,1999). In a review of stud-ies not involving environmental stress,biomass accu-mulation was greater for conifers (130%) than fordeciduous species (only 49% increase) under elevat-ed CO2 (Saxe et al.,1998). The wide range of plantresponses reflects,in part,the interaction of otherenvironmental factors and the CO2 response(Mooney et al.,1999;Stitt and Krapp,1999;Morisonand Lawlor, 1999; Johnson et al.,1998;Curtis andWang,1998). A recent field-scale experiment pro-duced significant (25%) increases in forest growthunder continuously elevated concentrations of CO2

for loblolly pine in North Carolina (Delucia et al.,1999). While positive tree responses to elevatedCO2 are likely to be overestimated if abiotic andbiotic environmental factors are not considered,most ecosystems responded positively in terms ofnet carbon uptake to increases in atmospheric CO2

above the current ambient level.

A significant question is how long these increasedresponses can be sustained. The observed acclima-tion or down-regulation of photosynthetic rates(Long et al.,1996;Lambers et al.,1998;Rey andJarvis,1998) has been ascribed to a physiologicalresponse (accumulation of photosynthetic reserves,Bazzaz,1990),or a morphological response (changesin trees,Pritchard et al.,1998;Tjoelker et al.,1998b).Declining photosynthetic rates have also beenascribed to the result of a water or nutrient stressimposed on pot-grown seedlings where root growth

is limited (Will and Teskey, 1997a;Curtis and Wang,1998). Down-regulation has been shown in low-temperature systems such as the Arctic tundra,where the initial enhancement of net carbon uptakedeclined after 3 years of elevated CO2 exposure(Shaver et al.,1992;Mooney et al.,1999). Down reg-ulation is likely in areas where nutrient availabilitydoes not increase along with carbon dioxide. Arecent review of large-scale field exposures to car-bon dioxide suggested that,though variable,treeresponse to CO2 was sustained over these short-term studies (Norby et al.,1999). They also exam-ined the CO2 response of trees growing near surfacevents of deep geothermal springs,and concludedthat a basal area increase of 26% was sustainedthrough 3 decades of elevated carbon dioxide.

Under enriched CO2 conditions, water use efficien-cy (WUE) has been shown to increase,which resultsin higher levels of soil moisture. These increasedlevels of soil moisture have been shown to be a sig-nificant factor in increased carbon uptake in water-limited ecosystems (Mooney et al.,1999). Whilethere is still uncertainty as to whether stomatal con-ductance decreases under elevated CO2 (Long et al.,1996;Will and Teskey, 1997b;Saxe et al.,1998;Curtisand Wang,1998),WUE increases either with or with-out changes in stomatal conductance (Aber et al.,2001). With constant conductance,the higheratmospheric CO2 concentration results in faster car-bon (C) uptake with constant water loss. If conduc-tance is reduced,a tradeoff is established betweenincreased C gain (which is partially reduced bydecreased conductance) and decreased water loss(also reduced by decreased conductance).Experimental studies have emphasized leaf-levelresponses. A physiological response observed atone scale (i.e.,leaf) does not necessarily imply that aresponse will be observed at the larger scale (i.e.,canopy, watershed). There is some evidence thatthe reduction in stomatal conduction of treeseedlings is not seen in mature trees (Ellsworth,1999;Mooney et al.,1999;Norby et al.,1999).

In the near-term, changes in the physiology of plantsare likely to be the dominant response to elevatedcarbon dioxide,strongly tempered by local environ-mental conditions such as moisture stress, nutrientavailability, as well as by individual species respons-es (Egli and Korner, 1997;Tjoelker et al.,1998a;Berntson and Bazzaz,1998;Crookshanks et al.,1998;Kerstiens,1998). Over the long term,plant specieschanges within forests will have a large influence onthe response of forests (Mooney et al.,1999).

497


Interactions of Multiple Stresses on Forests

It is crucial to understand not only the direct effectsof CO2, ozone,temperature,precipitation,and N andS deposition on forests,but also the interactiveeffects of these stresses. For example,if canopyconductance in forests is reduced in response toCO2 enrichment,then ozone uptake is reduced andthe effects of this pollutant mitigated. Droughtstress has a similar effect by reducing stomatal con-ductance. However, if N deposition leads toincreased N concentrations in foliage and hencehigher rates of photosynthesis and increased stom-atal conductance,then the positive effect ofincreased photosynthesis is partially offset byincreased ozone uptake. Reductions in productionfrom ozone damage could possibly speed the onsetof N saturation in ecosystems and the attendantdevelopment of acidified soils and streams. To proj-ect the effects of climate change and other stresseson forests,their interactive effects on forest process-es must be understood and integrated into ecologi-cal models.

Carbon Sequestration

Globally, an estimated 62-78% of the terrestrial car-bon is in forests (Shriner et al.,1998). NorthAmerican (Canada and US) forests hold 14% of thisglobal total,with large amounts contained in theboreal forests. Estimates of the carbon sequesteredannually as a result of climate and carbon dioxidefertilization in US forests were analyzed recently(Schimel et al.,2000) at a value of 0.08 Pg (Pg=Petagrams where 1 Pg = 1015 grams) of carbon peryear. This analysis focused on the 1980-1993 period.Estimates of carbon sequestration based on forestinventory data were 0.28 Pg of carbon per year, andmost of this increase in carbon storage was estimat-ed to be on private timberland (Birdsey and Heath,1995). Houghton et al.(1999) estimated theincrease in carbon stored per year ranged from 0.15to 0.35 Pg of carbon per year. Schimel et al.(2000)suggest that the larger inventory-based estimatesimply that the effects of intensive forest manage-ment and agricultural abandonment on carbonuptake in the US are probably equal to or largerthan the effects of climate and atmospheric carbondioxide. A comprehensive approach to account forcarbon fluxes to and from the atmosphere is thefocus of the recent IPCC Special Report on LandUse,Land-Use Change,and Forestry (Watson et al.,2000).

Projecting the Impact of Climate Change on Forest Processes

A number of studies have examined the climate con-trols on the distribution and productivity of forestsusing different types of models (Goudriaan et al.,1999;McGuire et al.,1992;VEMAP members,1995).In this assessment,three types of models were usedto project the impact of climate change on forestprocesses. Biogeochemistry models simulate thegain,loss and internal cycling of carbon, nutrients,and water;with these models,the impact of changesin temperature,precipitation,soil moisture,atmos-pheric carbon dioxide,and other climate-related fac-tors can be examined for their influence on suchprocesses as ecosystem productivity and carbonstorage.Biogeography models examine the influ-ence of climate on the geographic distribution ofplant species or plant types such as trees, grasses,and shrubs. Dynamic global vegetation models inte-grate biogeochemical processes with dynamicchanges in vegetation composition and distribution.

An earlier analysis using biogeochemistry modelsand four climate scenarios (different from thisassessment) shows increased net primary productiv-ity (NPP) at the continental scale (VEMAP members,1995). Carbon storage results vary with the mod-eled sensitivity to changes in water availability.When the direct effects of carbon dioxide are notincluded,several analyses indicate a reduction inproductivity (biogeochemistry models:VEMAP mem-bers,1995) or in vegetation density (biogeographymodel:Neilson et al.,1998).

In this assessment,three biogeochemistry models(TEM,Century, and Biome-BGC) and one dynamicglobal vegetation model (MC1) show increases intotal live vegetation carbon storage (3 to 11 Pg C)for forest ecosystems within the conterminous USunder the Hadley scenario (Table 1). Under theCanadian scenario,the models project changes inlive vegetation carbon for forests from a reductionof 1.6 Pg to an increase of 11 Pg C (Pg= Petagramswhere 1 Pg = 1015 grams). The MC1 model simu-lates a decline in vegetation carbon of about 2 Pg(Bachelet et al.,2001;Daly et al.,2000). Results forlive vegetation in all ecosystems and for live anddead vegetation in both forest and all ecosystemsparallel these results. Modeling the changes inspecies groups and fire disturbance are the primaryreasons for the carbon responses in MC1.

In the MC1 projections, regional changes in carbonstorage vary greatly across the conterminous US,andreflect the likelihood of disturbances such as

498


hurricanes,landslides,wind storms,and icestorms. Over geologic time,local, regional,andglobal changes in temperature and precipitationhave influenced the occurrence,frequency, andintensity of these natural disturbances.

Impacts of disturbances are seen over a broadspectrum of spatial scales,from the leaf and treeto the forest and forested landscape. Disturbancescan result in:leaf discoloration and reduction of leaffunction;deformation of tree structure such as bro-ken branches or crown losses;tree mortality orchronic stress resulting in tree death;altered regen-eration patterns including losses of seed banks;dis-ruption of physical environment including soil ero-sion;alterations in biomass and nutrient turnover;impacts on surface soil organic layers and theunderground plant root and reproductive tissues;and increased landscape heterogeneity (patchinessof forest communities). Introduced species (inva-sives) can affect forest ecosystems through her-bivory, predation,habitat destruction,competition,loss of gene pools through hybridization with nativespecies,and disease (either causing or carrying dis-ease). Outbreaks of native insects and disease canresult in similar impacts on forests.

At the ecosystem scale,introduced species as wellas outbreaks of native insects and diseases can alternatural cycles and disturbance regimes,such asnutrient cycles,and fire frequency and intensity(Mack and D’Antonio,1998). Some tree specieshave developed adaptations to survive repeatedoccurrences of certain disturbances over time.Thick bark on some trees allows their survival in

drought or fire (Figure 3). Under the Hadley sce-nario about 20% of current forest area experiencesome level of carbon loss,while the remaining 80%experience increased storage. Under the muchwarmer and generally drier Canadian scenario, closeto 80% of current forest area experiences a drought-induced loss of carbon. Reductions in carbon stor -age are projected in MC1 to be especially severe inthe eastern and southeastern US,where losses exac-erbated by drought or fire approach 75%.

In summary, synthesis of laboratory and field studiesand recent simulation experiments indicate that for-est productivity increases with the fertilizing effectof atmospheric carbon dioxide and that these pro-ductivity increases are strongly tempered by localconditions such as moisture stress and nutrientavailability. It is likely that modest warming couldresult in carbon storage gains in some forest ecosys-tems in the conterminous US. Under even warmerconditions,it is likely that drought-induced losses ofcarbon would occur in some forests,notably in theSoutheast and the Northwest. These losses of car-bon would possibly be enhanced by increased firedisturbance. These potential gains and losses of car-bon are very likely to be subject to changing land-use patterns,such as the conversion of forests toother uses.

2. Forests and Disturbance

Natural Disturbances Impacted by Climate

Natural disturbances,impacted by climate,includeinsects,disease,introduced species, fires,droughts,

Table 1. Changes in Carbon Storage for the Conterminous US

These results are based on simulations by the VEMAP models (TEM,BIOME-BGC,CENTURY, MC1) using thetwo transient climate change scenarios described in the text. Baseline period is 1961-1990. Changes aregiven at decades centered on 2030 and 2095,and for forest ecosystems,and all ecosystems. Values areexpressed in Pg (billions of metric tons).

Hadley Canadian 2030 2095 2030 2095

Forest Ecosystems Live Vegetation 0.4 to 4.0 3.0 to 11.1 -1.5 to 2.9 -1.6 to 10.5 Total (Live + Dead) 0.3 to 3.3 3.2 to 10.5 -2.9 to 2.0 -0.6 to 9.4

All Terrestrial EcosystemsLive Vegetation 0.4 to 4.7 3.3 to 13.2 -1.2 to 3.8 -1.8 to 13.9 Total (Live + Dead) 0.2 to 4.8 3.4 to 14.5 -0.2 to 4.4 -1.9 to 15.0

ground-level fires. In western forests, repeatedground-level fires reduce intermediate height vegeta-tion that can serve as fuel between the surface andthe crown. Thus,these repeated ground-level firesreduce the occurrence of stand-killing crown fires.

Disturbances can be regional or widespread.Landslide processes exhibit very strong geographicpatterns. Pacific coastal mountains are particularlyprone to sliding because of weak rocks,steepslopes,and high precipitation from frontal storms inthese tectonically active areas. Other disturbances,such as insects,pathogens,or introduced species,are widespread across the US. Many disturbancescascade into others. For example,drought oftenleads to insect outbreaks,disease,or fire. Insectsand disease can also create large fuel loads andthereby contribute to increased fire frequency.

Disturbances can be of either natural or anthro-pogenic origin (e.g., fire). Many large forest areashave been affected by past human activities,andcurrent disturbance regimes are profoundly differ-ent from historical regimes. For example, fire sup-pression in the fire-adapted western forests has ledto increased forest density and biomass, changes inforest composition,and increased outbreaks ofinsects and disease (Flannigan et al.,2000;Ayres andLombardero,2000). Some forest types have evolved

to depend on the periodic occurrence of the distur-bance. Long-leaf pine forests are a fire-climaxecosystem that would not exist if fire were not apart of the Southeast.Nearly 1.5 million acres ofprescribed burning and other fuel reduction treat-ments in 1998 were used to enhance forest healthand diversity by restoring fire-dependent ecosys-tems that have been affected by long-term fire sup-pression (USDA Forest Service,1998). Natural dis-turbances interact with human activities such as airpollution,harvest, agricultural and urban encroach-ment,and recreation.

Impact of Climate Change on Forest Disturbances

Climate variability and climate changes alter the fre-quency, intensity, timing,and/or spatial extent of dis-turbances. Many potential consequences of futureclimate change will possibly be buffered by theresilience of forest communities to natural climaticvariation. However, the extensive literature on thissubject suggests that new disturbance regimesunder climate change will likely result in significantperturbations to US forests,with lasting ecologicaland socioeconomic impacts (Dale et al.,2001).

Hurricanes. Hurricanes seriously impact forestsalong the eastern and southern coasts of the US aswell as the Caribbean Islands (Lugo,2000). They 499


Figure 3. Change in live vegetation car-bon density from the historical period(1961-1990) to the 2030s (2025-2035) andthe 2090s (2090-2099) under two climatescenarios. Change in the biomass con -sumed by natural fires between the 20th

century (1895-1993) and the 21st century1994-2100) as simulated by the dynamicvegetation model MC1 under two climatescenarios (bottom two panels) (Bacheletet al. 2001) See Color Plate Appendix

Patterns of Live Vegetation for Different Times and Climate Scenarios

Hadley Canadian

near 20% for the Northeast and small decreases forthe northern Great Plains,with increases less than10% over the rest of the continent. The warmer anddrier Canadian scenario produces a 30% increase infire severity for the Midwest,Alaska,and sections ofthe Southeast,with about 10% increases elsewhere.These results suggest a possible increase of 25-50%in the area burned in the US. Temperature and pre-cipitation are not the only climate-related factorsthat influence fire regimes; for example,lightningstrike frequency was estimated to increase 44%under the Goddard Institute for Space Studies gener-al circulation model scenario (Price and Rind,1994).Other factors such as length of the fire season,weather conditions after ignition,and vegetationcharacteristics also influence the fire regimes.Wotton and Flannigan (1993) found that the fire sea-son would be on average 30 days longer in a doublecarbon dioxide climate as compared to the currentclimate for Canada. Wildfire severity was at least assensitive to changes in wind as to changes in tem-perature and precipitation in a climate change sensi-tivity analysis for California (Torn and Fried,1992).Human activities such as fire policies and land usewill likely also influence fire regimes in the future(Keane et al.,1998). For example,wildfires on alllands in the western US increased in the 1980s after30 years of aggressive fire suppression that had ledto increases in forest biomass.

The second modeling study examined the influenceof climate change on vegetation and the interactionwith natural fires (Bachelet et al.,2001). Theamount of biomass consumed in fire increasedunder future climates in analyses with the dynamicglobal vegetation model,MC1 (Figure 4). In thismodel, fire occurrence,severity, and size are simulat-ed as a direct function of fuel and weather condi-tions (Lenihan et al.,1997;Daly et al.,2000,Bacheletet al.,2001). In the western US,increased tempera-ture,steady to increased precipitation,CO2 fertiliza-tion,and increased water use efficiency enhanceecosystem productivity, resulting in more biomass.The highly variable climate of dry years interspersedwith wet years and the fuel buildup contributes tomore and larger fires in the western landscape. Inthe eastern US under the Canadian scenario, firesare projected to increase in the Southeast as a resultof increased drought stress in forests.

Both approaches suggest the potential for anincreased area to be burned with a changing cli-mate.These analyses are based on the physical andbiological factors influencing potential fire hazard.Factors such as current land management,land use,and ownership are not considered. Harvest activi-

can inflict sudden and massive tree mortality, com-plex patterns of tree mortality including delayedmortality, and altered patterns of forest regeneration(Lugo and Scatena,1996). Because most hurricanedamage is from floods,effects can be removed fromthe actual hurricane in distance (heavy precipitationwell inland from the coast) and in time (delaysinvolved in water movement,and in mortality result-ing from excessive water).Hurricanes can lead toshifts in successional direction,higher rates ofspecies turnover, opportunities for forest specieschange,increasing landscape heterogeneity, fasterbiomass and nutrient turnover, and lower above-ground biomass in mature vegetation (Lugo andScatena,1995).

Hurricane location,size,and intensity are influencedby sea surface temperatures and by regional weath-er features (Emanuel,1999). Sea surface tempera-tures (SSTs) are expected to increase,with warmerSSTs expanding to higher latitudes (Royer et al.,1998;Walsh and Pittock,1998). Climate change isalso likely to influence the frequency of regionalweather events conducive to hurricane formation,although it is not yet possible to say whether hurri-cane frequency increases or decreases (Royer et al.,1998;Henderson-Sellers et al.,1998;Knutson et al.,1998;Knutson and Tuleya,1999).

Fire. Fire frequency, size,intensity, seasonality, type,and severity are highly dependent on weather andclimate. An individual fire results from the interac-tion of ignition agents (such as lightning,fuel condi-tions,and topography) and weather (including airtemperature, relative humidity, wind velocity, andthe amount and frequency of precipitation). Overthe 1989-1998 period,an average of 3.3 millionacres burned annually in the US, varying from 1 mil-lion acres a year to over 6 million,mostly in thewest and southeast. Most of the burned acres result-ed from human-caused fires.

Two modeling approaches were used to look at theimpact of climate change on fire:1) fire severity rat-ings estimated from future climate (Flannigan et al.,2000),and 2) the interaction of vegetation biomassand climate in establishing conditions for fire(Bachelet et al.,2000).

In the analysis by Flannigan et al.(2000),the futurefire severity is projected to increase over much ofNorth America under both climate scenarios andthese results are consistent with earlier analyses(Flannigan et al.,1998). The results show great vari-ation for the US and Canada. The warmer and wet -ter Hadley scenario suggests fire severity increases500


ties and the conversion of forest land to other useswould also alter the amount and kind of fuel.

The rapid response of fire regimes to changes in cli-mate is well established (Flannigan et al.,1998;Stocks et al.,1998),so this response has the poten-tial to overshadow the direct ef fects of climatechange on species distribution,migration,andextinction within fire-prone areas. This possibility ofincreased fire poses challenges to the managementof protected areas such as national parks (Malcolmand Markham,1998) and to the management offorests for carbon storage.

Drought. Droughts occur in nearly all forest ecosys-tems (Hanson and Weltzin,2000). The primaryimmediate response of trees to drought is to reducenet primary production (NPP) and water use,whichare both driven by reduced soil moisture and stom-atal conductance. Under extended severe droughtconditions,plants die. Seedlings and saplings usuallydie first and can succumb under moderate droughtconditions. Deep rooting,stored carbohydrates,andnutrients in large trees make them susceptible onlyto longer, more severe droughts. Secondary effectsalso occur. When reductions in NPP are extreme orsustained over multiple growing seasons,increasedsusceptibility to insects or disease is possible,espe-cially in dense stands (Negron,1998). Drought canalso reduce decomposition processes leading to abuildup of organic matter on the forest floor.

The consequences of a changing drought regimedepend on annual and seasonal changes in climateand whether a plant’s current drought adaptationsoffer resistance and resilience to new conditions.Forests tend to grow to a maximum leaf area thatuses nearly all available growing-season soil water(Eagleson,1978;Hatton et al.,1997; Kergoat,1998;Neilson and Drapek,1998). A small increase ingrowing-season temperature could increase evapora-tive demand and trigger moisture stress.Using thisassumption about leaf area, results from MAPSS, abiogeography model,and MCI,a dynamic global veg-etation model,suggest that increased evaporativedemands will likely cause future increases indrought stress in forests of the Southeast,southernRocky Mountains,and parts of the Northwest(Neilson and Drapek,1998;Bachelet et al.,2000).While earlier forest models,often known as ‘Gap’models (Shugart,1984) also suggested forestdeclines,all of a tree’s current drought adaptations,such as adjusting growth rates or aboveground-belowground allocations of carbon,had not beenincorporated into the analysis (Loehle and LeBlanc,1996). These adaptations buffer the impact of cli -

mate,including drought,on individual trees and for-est stands. The current generation of ecologicalmodels,such as MAPSS,MC1 and others,haveimproved upon these process algorithms,includingthe suggestions from Loehle and LeBlanc (1996).

Wind events. Small-scale wind events,such as torna-does and downbursts,are products of mesoscaleweather circumstances (Peterson,2000). These dis-turbances can create very large areas of damage.For example,an October 25,1997 windstorm flat-tened nearly 13,000 acres of spruce-fir forest in theRoutt National Forest of Colorado (USDA ForestService Routt National Forest,1998),and a July 4,1999 windstorm flattened roughly 250,000 acres offorest in the Boundary Waters Canoe Area ofMinnesota (Minnesota Dept.of Natural Resourcespress release 7/12/99). Although small-scale windevents occur throughout the US,the highest con-centration of tornadoes occurs across the CentralGreat Plains states of Oklahoma, Texas,Kansas,andNebraska.

If climate change increases intensity of all atmos-pheric convective processes,this change will accel-erate the frequency and intensity of tornadoes andhailstorms (Berz,1993). Karl et al.(1995a) foundthat the proportion of precipitation occurring inextreme weather events increased in the US from 501


Figure 4. Simulated total biomass consumed by fire over the con-terminous US under historic and two future climates; Hadley(HADCM2SUL) and Canadian (CGCM1) scenarios. The fire simula-tions are for potential vegetation and do not consider historic firesuppression activities. However, grid cells with more than 40%agriculture have been excluded from the calculations (Lenihan et.al., 1997, Daly et. al., 2000, Bachelet et. al., 2001). See Color PlateAppendix

Biomass Consumed under Two Scenarios of Future Climate

affect Fraser fir (Abies fraseri) survival (Dale et al.,1991).

A key feature of most invasive species is that theyhave the capacity to thrive in disturbed environ-ments through their high reproductive rates, gooddispersal abilities,and rapid growth rates (Vitouseket al.,1996). If climate change results in increaseddisturbances such as fire or drought,these disrup-tions to ecosystems create just the type of environ-ments in which invasive species are likely to expandrapidly. The interactions among introduced species,native communities,intensively managed forests,human activities that fragment ecosystems,increased atmospheric deposition,and climatechange might positively affect the prevalence ofinvaders,but forecasting specific impacts of inva-sions remains problematic (Dukes and Mooney,1999;Williamson,1999).

Insect and Pathogen Outbreaks Outbreaks ofinsects and pathogens can adversely affect recre-ation,wildlife habitat, wood production and ecologi-cal processes.Over the 1986 to 1995 period,these 4native insect species damaged annually the follow-ing acreages: over 4 million acres for western sprucebudworm;less than 1 million acres for easternspruce budworm;less than 2 million acres formountain pine beetle;and nearly 12 million acresfor southern pine beetle (The Heinz Center, 1999).Within this time period,the acreage affected by anyone of these insects could vary from less than halfof the long-term average to three times the long-term average. Nearly 13 million acres of southernforests are affected by a single disease,fusiform rust,and 29 million acres of western forests are affectedby a parasitic plant, dwarf-mistletoe (The HeinzCenter, 1999). Disturbances such as drought andfire influence these outbreaks.

An extensive body of scientific literature suggestsmany pathways through which elevated carbondioxide and climate change could significantly alterpatterns of disturbance from insects and pathogens(Ayres and Lombardero,2000). Elevated carbondioxide and climate change could possibly increaseor decrease the disturbances of insects andpathogens through direct effects on the survival,reproduction,and dispersal of these organisms. Forexample,an increase in the interannual variation inminimum winter temperatures could possibly favormore northerly outbreaks of southern pine beetles,while decreasing more southerly outbreaks(Ungerer et al.,1999). It is also possible that climatechange would alter insect and pathogen distur-bances indirectly through changes in the abundance

1910 to 1990. Karl et al.(1995b) further suggestthat the US climate has become more extreme (interms of temperature and precipitation anomalies)in recent decades. Further, Etkin (1995) found apositive correlation between monthly tornado fre-quency in western Canada and mean monthly tem-perature,and inferred that this relationship suggestsincreased tornado frequency under a warmer cli-mate. Despite the above indirect inferences abouttornado frequencies,and the direct data on thunder-storm trends,there is still inadequate understandingof tornado genesis to directly forecast how climatechange will affect the frequency or severity of wind-storms in the next century (Peterson,2000).

Ice Storms. Ice storms,also known as glaze events,result when rain falling through subfreezing airmasses close to the ground is supercooled so thatraindrops freeze on impact (Irland,2000). TheNational Weather Service (NWS) defines an icestorm as an occurrence of freezing precipitationresulting in either structural damage or at least 0.25inch of ice accumulation. Ice accumulation canvary dramatically with topography, elevation,aspect,and areal extent of the region where conditionsfavor glaze formation. While ice storms can occur asfar into the southern US as northern Mississippi andTexas,the frequency and severity of ice stormevents generally increases toward the northeast(Irland,2000).

Depending on forest stand composition,amount andextent of ice accumulation,and stand history, dam-age can range from light and patchy to total break-age of all mature stems (Irland,1998). Even thoughthe weather conditions producing ice storms arewell understood,it is not known how changes in cli-mate will affect the frequency, intensity, location,orareal extent of ice storms.

Introduced species. Climate,as well as human activi-ties,largely determine the potential and realized dis-tributional ranges of introduced species (Simberloff,2000). Unsuitable climate at points of arrivalrestricts the survival of a great majority of intro-duced species (Williamson,1999). In warmer partsof the US,introduced species comprise a larger frac-tion of the biota (Simberloff, 1997). Where climatecurrently restricts invasives, changes in temperatureor precipitation may facilitate increased growth,reproduction,or expansion of their ranges. Forexample,laboratory studies of balsam woolly adel-gid (Adelges piceae), growing under various temper-ature conditions,provided the basis for simulationssuggesting that temperature-induced changes in thepopulation dynamics of the insect significantly502


of their natural enemies and competitors. In addi-tion, climate,and elevated carbon dioxide,influ-ence the susceptibility of trees to insects andpathogens through changes in the chemistry ofplant tissues (Ayres and Lombardero,2000).

The short life cycles,high mobility, reproductivepotential,and physiological sensitivity to tempera-ture suggest that even modest climate change willpossibly have rapid impacts on the distribution andabundance of many forest insects and pathogens.Beneficial impacts could possibly result wheredecreased snow cover increases winter mortality.Detrimental impacts could possib ly result whenwarming accelerates insect development and facili-tates dispersal of insects and pathogens into areaswhere tree resistance is less. Detrimental impactscould also possibly result from interactions of dis-turbances. For example, warming could increaseoutbreaks in boreal forests that would tend toincrease fire frequency (Ayres and Lombardero,2000). Already, the impact of insects and disease inforests is widespread. In 1995, over 90 millionacres of forestland in the US were affected by afew species:southern pine beetle,mountain pinebeetle,spruce budworm (eastern and western),spruce beetle, dwarf mistletoe, root disease,andfusiform rust. Thus,there are potentially impor-tant ecological and socio-economic consequencesto these beneficial and detrimental impacts (Ayresand Lombardero,2000;Ayres and Reams,1997).

3. Biodiversity

Land Use Impacts Species, Communities, and Biomes

Global ch a n ge encompasses a number of eve n t so c c u rring at the continental scale, i n cluding cl i m a t ech a n ge , land use ch a n ge , species inva s i o n , and airp o l l u t i o n . Global ch a n ge has and will like ly contin-ue to affect the abundance and distribution ofplants and animals which , in turn , will have consid-e rable ra m i fications for human economics, h e a l t h ,and social we l l - b e i n g . O rganisms provide go o d si n cluding material pro d u c t s , fo o d s , and medicines.In addition, the number and kinds of species pre s-ent affect how ecosystems respond to globalch a n ge . It is the responses of individual org a n i s m sthat begin the cascade of ecological processes thatthen manifest themselves as ch a n ges across land-s c a p e s ,b i o m e s , and the globe (Hansen et al., 2 0 0 1 ,Wa l ker et al., 1 9 9 9 ) .

Humans modify the quality, amount,and spatialconfiguration of habitats. A number of natural

community types now cover less than 2% of theirpre-settlement ranges (Noss et al.,1994).Examplesinclude:spruce-fir forests in the southernAppalachians;Atlantic white-cedar in parts ofVirginia and North Carolina; red and white pine inMichigan;longleaf pine forests in the southeasterncoastal plains;slash pine rockland habitat in south-ern Florida;loblolly/shortleaf pine forests in thewest gulf coastal plains;and oak savannas inOregon. For the species dependent upon ecosys-tems that have declined in area,such habitat losscan reduce effective population sizes, genetic diver-sity, and the ability of species to evolve adaptationsto new environments (Gilpin,1987). The areainvolved in land use shifts can dwarf the land areainvolved in natural disturbances. For example,thearea of harvested cropland went from 292 millionacres in 1964 to 347 million acres in 1982,andthen down to 293 million in 1987. For compari-son,the total area burned in fires at 1 to 6 millionacres annually is much less than these land areashifts in and out of agriculture. Though the forestremains standing,only five species of insects areestimated to defoliate about 21 million acres eachyear, on average (The Heinz Center, 1999).

Land use change alters the spatial pattern of habi-tats by creating new habitats that are intensivelyused by humans or by reducing the area and frag-menting natural habitats. These changes increasethe distance between habitat patches and reduceoverall habitat connectivity. Native forests havebeen converted to agricultural and urban uses,notably in the eastern and midwestern parts of theUS. In some cases, forests have regrown on aban-doned agricultural lands. Recent expansion ofurban areas and agriculture are fragmenting west-ern forests. Nationally, urban areas have doubled inarea between 1942 and 1992 (Flather et al.,1999).While urban areas increased most rapidly in theSouth and Pacific Coast regions,the most influen-tial land use change has been the increase inhuman population density in rural areas,particular-ly in the Rocky Mountain and Pacific Coast regions.This expansion alters disturbance patterns associat-ed with fire, flooding,and landslides.Roadways andexpansion of urban areas have fragmented forestsinto smaller, less-contiguous patches.

Loss of habitat and degradation of habitat qualitycan reduce population size and growth rates,andelevate the chance of local extinction events(Pulliam,1988). The steadily increasing number ofspecies listed as threatened and endangered in theUS is currently at 1,232 (USDI, Fish and WildlifeService,2000). Factors contributing to species 503


association with climate warming (Abraham andJefferies,1997). The breeding dates of both amphib-ians and birds in Great Britain have shifted one tothree weeks earlier since the 1970s in associationwith increasing temperatures (Beebee,1995;Cricket al.,1997).

The primary focus for this analysis is on the conti-nental-scale response of forest vegetation as reflect-ed in climate-induced changes in the distributions ofbiomes,community types,species richness,and indi-vidual tree and shrub species. Vegetation responsesto projected future climate change were assessed byreviewing the available literature and by using theclimate predictions of global climate models asinput to a set of different vegetation simulationmodels. Paleoecological studies of climate changeimpacts on forests provide information about pastresponses. However, their results are limited withrespect to the future because the current size, age,and species composition of temperate forests hasbeen strongly impacted by human activities,andsecond,global temperatures are predicted toincrease at an unprecedented rate (Dale,1997).Although vegetation models incorporating land-usedynamics are under rapid development at localscales,the current state of knowledge does notallow for integrating the effects of land use at thecontinental scale.

The vegetation models used in this analysis includethe following. Two statistical models project the dis-tribution of individual tree species with resultsaggregated into community types. For the easternUS,the DISTRIB model,which projects potentialchanges in suitable habitat of 80 tree species, wasdeveloped from 33 environmental variables and thecurrent distribution of each tree species (Iverson etal.,1999). The results for 80 species under each cli-mate change scenario were also aggregated to exam-ine potential changes in forest types in the easternUS (Iverson and Prasad,2001). Shafer et al.(2001)developed local regression models that estimate theprobability of occurrence of 51 tree and shrubspecies across North America based on 3 bioclimaticvariables:mean temperature of the coldest month,growing degree days,and a moisture index.Changes in species richness of trees and terrestrialvertebrates based on energy theory (Currie,1991)were also analyzed under the different climate sce-narios. Species interactions and the physiologicalresponse of species to carbon dioxide are notincluded in these statistical models. In contrast,theMAPSS biogeography model (Neilson,1995) projectsbiome response to climate as change in vegetationstructure and density based on light,energy, and

endangerment include habitat conversion, resourceextraction,and exotic species (Wilcove et al.,1998).The spatial distribution of such factors results inspecies at risk being concentrated in particularregions of the US,especially the southernAppalachians,the arid Southwest,and coastal areas(Flather et al.,1998),which can be traced back tohuman population growth and attendent land-useintensification. Land-use intensification has alsobeen found to affect animal communities broadly.Along a gradient of increasing land-use intensifica-tion in forested regions of the eastern US,nativespecies of breeding birds are increasingly under-rep-resented and exotic species over-represented(Flather, 1996).

Species that take advantage of anthropogenic habi-tats,such as some deer, goose,and furbearer species,have greatly expanded in recent years (Flather et al.,1999). Some species of deer (e.g.,white-tailed deer)are now so abundant that the primary concern iscontrolling their populations. In addition, exoticspecies have become established and greatlyexpanded their ranges in the US (Drake et al.,1989).

Impact of Climate Change on Biodiversity

Changes in the distribution and abundance of plantor animal species reflect the birth, growth,death,and dispersal rates of individuals in a population.When aggregated,these changes manifest as localextinction and colonization events,which are themechanisms that determine a species’ range. Whileclimate and soils are strong controls on the estab-lishment and growth of plants,the response of plantand animal species to climate change will be theresult of many interrelated processes operating overseveral scales of time and space. Migration rates,changes in disturbance regimes,and competitiveand cooperative interactions with other species willaffect the distribution of plants and animals.Because of the individualistic response of species,biotic communities are not expected to respond asintact units to climate change. Community compo-sition responds to a complex set of factors includingthe direct effects of climate,differential species dis-persal,and indirect effects associated with changesin disturbance regimes,land use,and interspecificinteractions (Peters,1992).

Some of the best evidence that species’distributionsare correlated with changing climates is found forplants (Webb,1992). Recent observations of somespecies suggest a response to historical changes inclimate. Breeding ranges of some mobile species,such as waterfowl,have expanded northward in504


water limitations (VEMAP Members,1995). Thepotential natural vegetation is coupled directly toclimate and hydrology, and rules are applied to clas-sify vegetation into biome types. The model con-siders the effects of altered carbon dioxide onplant physiology. The analyses for species,commu-nities,and biomes used an equilibrium climate sce-nario based on the transient Canadian and transientHadley scenarios. The baseline scenario was theaverage climate for the 1961-1990 period. The “cli-mate change”scenario was the average of the pro-jected climate for 2070 to 2100. The results ofthese analyses are given below. The implications ofspecies dispersal and land use change are discussedin the context of these analyses.

Tree Species. The potential distribution of treespecies under climate change is modeled usingboth statistical models (Iverson and Prasad,2001;Shafer et al.,2001) and both climate scenarios usedin the National Assessment. These statisticalapproaches assume that there are no barriers tospecies migration. Results should be viewed asindications of the potential magnitude and direc-tion of range shifts under a changed climate andnot as predictions of change.

For many of the eastern tree species,their possibleranges shift north (Iverson et al.,1999;Iverson andPrasad,2001). Under both climate scenarios,therange of sugar maple (Acer saccharum) shifts outof the United States entirely (Figure 5). White oak(Quercus alba) remains within its current rangebut is reduced in importance in the southern partsand increases in the northern parts of its range. Atotal of 7 of the 80 eastern tree species are project-ed to decline in regional importance by at least90%:bigtooth aspen (Populus grandidentata),aspen (Populus tremuloides),sugar maple,north-ern white-cedar (Thuja occidentalis),balsam fir(Abies balsamea), red pine (Pinus resinosa),andpaper birch (Betula papyrifera). The ranges ofmost species are projected to move to the north,with the ranges of several species moving north by60 to 330 miles (100 to 530 km). For somespecies,such as aspen,paper birch,northern white-cedar, balsam fir, and sugar maple,the optimum lati-tude for their occurrence moves north of the US-Canadian border.

When integrated into community types,southernforest types expand while higher elevation andnorthern forest types decline in area (Figure 6).The oak-hickory type is projected to expand inarea by 34% primarily to the north and east(Iverson and Prasad,2001). The oak-pine type is

projected to expand in area by 290% throughoutthe Southeast. Area of spruce-fir and aspen-birchtypes is projected to decline by 97% and 92%respectively. These types are replaced mainly byoak-hickory and oak-pine forests. The loblolly-short-leaf pine type is also projected to be reduced by32% and shifts north and west,being replaced in itscurrent zone by the oak-pine type. The longleaf-slash type is projected to be reduced by 31% inarea.

In the western US,the potential future ranges formany tree species are simulated to change,withsome species’ ranges shifting northward into Canada(Shafer et al.,2001). Simulated future ranges forWestern hemlock (Tsuga heterophylla) andDouglas-fir (Pseudotsuga menziesii) (Figure 7) areprojected to decrease west of the Cascade

505


Projected Changes in Distribution of Sugar Maple

Figure 5. Projected distribution for sugar maple under current cli-mate and the Hadley climate scenario and for the eastern UnitedStates, using statistical models developed by Iverson et al. (1999).The Predicted Current map is the current distribution and impor-tance value of sugar maple, as modeled from the regression treeanalysis. Importance value is an index based on the number ofstems and basal area of both the understory and the overstory.Predicted Hadley is the potential suitable habitat for sugar mapleunder the Hadley climate scenario. These potential maps imply nobarriers to migration. The Difference map represents the differ-ence between Modeled Current and Predicted Hadley maps. ThePotential Shifts map displays the modeled current distribution,along with predicted potential future distribution (using the Hadleyscenario) and the overlap where the species is now and is project-ed to be in the future. See Color Plate Appendix

Mountains with, for example,species typical todayof Glacier National Park expanding to the southeastinto Yellowstone National Park (Bartlein et al.,1997). Contrasts between eastern and westernrange shifts can be seen in the simulated futurerange of Paper birch (Betula papyrifera) whoseeastern US range limit contracts northward butwhose western US range limit expands southward(Figure 7).

Community Richness. The relationship betweenspatial patterns of climate and richness of trees andterrestrial vertebrate species were used to predictchanges in species richness under climate changeacross North America (Hansen et al.,2001;Currie,1991;Currie, 2001). The climate relationships withspecies richness were stronger for temperature thanprecipitation. Because the climate-richness relation-ships,including where maximal richness occurs,dif-fer across taxonomic groups,the impacts of climatechange will likely also vary across the groups.

Across all scenarios,tree richness is projected toincrease in the cooler regions:northern US,thewestern mountains,and near the Canadian Border(Currie,2001). Because species richness of birdsand mammals is currently highest in moderatelywarm areas and decreases in hotter areas,the modelprojected species richness for these taxa declines asair temperatures increase. Those scenarios wherewarming is projected to occur results in a decreaseof bird and mammal richness—-a 25% decrease inlow-elevation areas in the Southeast, for example.Warming in colder areas (such as mountainousareas) is projected to result in highly variableincreases in richness (11% to >100%). Cold-bloodedanimals such as reptiles and amphibians would ben-efit from increased air temperatures. Consequently,under the warming climate scenarios,species rich-ness for these taxa is projected to increase fromabout 11% to 100% over the entire conterminousUS.

Flather et al.(1998) identified regions in the US thathave high concentrations of the currently designat-ed threatened and endangered species. Species rich-ness was used to explore the potential impact of cli-mate change on taxonomic groups within thesehotspots of species endangerment. The projectedrichness results were overlaid on maps of currenthot spots for threatened and endangered species,and the species richness changes within each taxo-nomic group were evaluated. Within thesehotspots,species richness for reptiles and amphib-ians increased,whereas bird and mammal richnessappeared to be much reduced in many of them,especially in the East.

Mountains. The potential future range of Westernhemlock extends into mountain ranges throughoutthe interior west while Douglas-fir expands east ofthe Cascades and Sierras as well as northward alongthe west coast of Canada into Alaska. The potentialfuture ranges for subalpine conifers such asEngelmann spruce (Picea engelmannii),Mountainhemlock (Tsuga mertensiana),and several speciesof fir (Abies species) are much reduced in the west-ern parts of the US;however these subalpinespecies expand to the north along the west coast ofCanada and into Alaska. While many of the moremesic and higher elevation species shift northwardinto Canada,the potential future range of Ponderosapine (Pinus ponderosa) expands within the interiorwestern US.

The complex topography of the western US,com-bined with its seasonal and regional variations in cli-mate,strongly influences potential future shifts inthe ranges of tree species and their likely future abil-ities to successfully disperse to new habitat (Hansenet al.,2001;Shafer et al.,;2001). Species range shiftsin the western US are simulated to occur in alldirections whereas in the eastern US,the shifts tendto be northward as temperature increases and west-ward as precipitation increases. In the western US,several conifer species associated with moderatelymoist climates shift south and east along the Rocky

506


Figure 6. Projected forest communities under (a) current climate, (b) the Hadley climatescenario, and (c) the Canadian climate scenario, based on the results of individualanalyses of 80 tree species shifts (see Prasad and Iverson, 1999-ongoinghttp://www.fs.fed.us/ne/delaware/atlas/index.html) See Color Plate Appendix

Current and Projected Forest Communities in the Eastern US

a. b.

c.

507


Biome Shifts. The impact of potentialclimate change on the biogeography ofthe US is examined using six scenariosthat varied in degree of warming(Bachelet et al.,2001). Across the sce-narios, forest area is projected todecrease by an average of 11%,with arange of +23% under the moderatelywarming scenarios to –45% under thehottest scenarios. Northeast mixedforests (hardwood and conifer) are pro-jected to decrease by 72% in area in theUS,on average,as they shift into Canada(Neilson et al.,1998). The range of east-ern hardwoods is projected to decreaseby an average of 34%. Although thesedeciduous forests shift north, replacingNortheast mixed forests,they aresqueezed from the south by Southeastmixed forests or from the west by savan-nas and grasslands,depending on thescenario. Southeast mixed forests areprojected to increase under all scenarios(average 37%). This biome would remain intactunder moderately warm scenarios,but would beconverted to savannas and grasslands in the Southunder the hotter scenarios.

Alpine ecosystems are projected to all but disappearfrom the western mountains,being overtaken byencroaching forests. Under the climate scenariosstudied, wet coniferous forests in the northwestdecrease in area by 9% on average,while the extentof interior western pines change little. Responses inboth wet conifer and interior pine forests showwide variations among scenarios,with expansionsunder newer transient scenarios (Bachelet et al.,2001).Arid woodlands also are projected to expandin the interior West and Great Plains,encroachingon some grasslands.

Likely Species Shifts, Dispersal, and Land UseImpacts. The locations and areas of potential habi-tats for many plant and animal species are likely toshift as climate changes. Potential habitats for treesfavored by cool environments are likely to shiftnorthward. The habitats of alpine,subalpinespruce/fir, and aspen communities are likely to con-tract dramatically in the US and largely shift intoCanada. Potential habitats that are likely to increasein the US are oak/hickory, oak/pine,ponderosa pine,and arid woodland communities. Most of theseresults were evident in the results from three inde-pendent models:the mechanistic model MAPSS andtwo statistical models. These results were also rela-tively robust across the several climate scenarios

analyzed,including the Hadley and the Canadianscenarios. The statistical models are highly defensi-ble for this application because of the strong andextremely well-documented relationship betweentree species performance and environmental condi-tions,especially climate and soil.However, statisticalmodels do not incorporate a direct CO2 effect,which enhances water-use efficiency. Even so,it iswell accepted that climate and soils are strong con-trols on the establishment, growth,and reproduc-tion of many tree species.

How well these species actually track changes intheir potential habitats will very likely be stronglyinfluenced by their dispersal abilities and the distur-bances to these environments. Good analyticalmodels for dispersal are few. Some native specieswill very likely have difficulty dispersing to new

Figure 7. Simulated distributions and scenario agreement for Betulapapyrifera and Pseudotsuga menziesii (after Hansen et al., 2001).Estimated probabilities of occurrence for each taxon simulated withobserved modern climate (left panel). Comparison of the observeddistributions with the simulated future distributions under future cli-mate conditions as generated by the Hadley (HADCM2) andCanadian (CGCM1) scenarios for 2090-2099 (middle panels). Grayindicates locations where the taxon is observed today and is simu-lated to occur under future climate conditions; red indicates loca-tions where the taxon is observed today but is simulated to beabsent under future climate conditions; and blue indicates loca-tions where the taxon is absent today but is simulated to occurunder future climate conditions. Scenario agreement (right panel).Light purple indicates locations where the species is simulated tobe present under the future climate of either the HADCM2 orCGCM1 scenario; dark purple indicates locations where the speciesis simulated to be present under both future climate scenarios. SeeColor Plate Appendix

Paper Birch and Douglas Fir Tree Distributions under Future Climate Change Scenarios

508


habitats because of the rapid rate of climate changeand the varying land uses along alternate migrationroutes. For example,aspen communities are cur-rently being reduced by conifer encroachment, graz-ing,invasive species,and urban expansion. Weedspecies that disperse rapidly will likely to be well-represented in these new habitats. Hence,thespecies composition of newly forming communitieswill likely differ substantially from those occupyingsimilar habitats today.

4. Socioeconomic Impacts

Current Supply and Demand for Amenities, Services,and Goods from Forests

Forests and the forestry sector provide theresource base and flow of amenities, goods,andservices that provide for needs of individuals,communities, regions,the nation,and export cus-tomers. Harvested products include timber, pulp-wood,fuelwood,wild game, ferns, mushrooms,floral greenery, and berries. Recreational opportu-nities in forests range from skiing, swimming,hik-ing and camping to birding,and autumn leaf tours.Forest land is also managed to provide clean waterand habitat for wildlife.

Environmental conditions and available technologyas well as social,political,and economic factorsinfluence the supply of and the demand for forestamenities, goods,and services. Factors that influ-ence supply of these services include forest area,forest productivity, forest management,productiontechnology, and the quality of amenities. Factorsthat influence demand include population,econom-ic growth and structure,and personal preferences.Changes in these factors also influence supply ofand demand for forest amenities, goods,and servic-es. Land-use pressures alter the amount and type ofland available for forest reserves, multiple-useforests,commercial recreation, forest plantations,and carbon storage (Alig,1986;Leemans et al.,1996;Solomon et al.,1993). Changes in forest speciescomposition, growth,and mortality alter the possi-ble supply of specific types of wood products,wildlife habitat,and recreation. Assumptions aboutchanges in human needs in the US and overseas alsoaffect the socioeconomic impacts from climatechange on US forests (Joyce et al.,1995; Perez-Garciaet al.,1997). Clearly, forest changes caused byhuman use of forests could exceed those impactsfrom climate change (Dale,1997).However, climatechange could impact many of the amenities, goods,and services from forests (Bruce et al.,1996),includ-

ing productivity of locally harvested plants such asberries or ferns;local economies through land useshifts from forest to other uses; forest real estate val-ues;and tree cover and composition in urban areas,and associated benefits and costs. In addition, cli-mate-change impacts on disturbances such as firecould increase fire suppression costs and economiclosses due to wildfires (Torn et al.,1999). In thisassessment,it was only possible to explore theimpact of climate change on wood products andrecreation in more detail.

North America is the world’s leading producer andconsumer of wood products. The US has substantialexports of hardwood lumber, wood chips,logs,andsome types of paper (Haynes et al.,1995). The USalso depends on Canada for 35% of its softwoodlumber and more than half of its newsprint. Past,current,and future land uses and management aswell as environmental factors such as insects,dis-ease, extreme climate events,and other disturbancesaffect the supply of forest products. These factorscan alter the cost of making and using wood prod-ucts,and associated jobs and income in an area.Changes in economic viability of wood productioninfluence whether owners keep land in forests orconvert it to other uses. Demand for wood prod-ucts,consumption,and trade is strongly influencedby US and overseas population,economicgrowth,and human values.While demand fortimber needed to make US wood products isprojected to grow at about the same rate aspopulation over the next 50 years,the demandsfor products and the kinds of timber harvestedwill be affected by technology change and con-sumer preferences. For example,use of recy-cled paper slows the increase in the amount oftimber harvest needed to meet increasingdemands for various paper products.

The combinations of resources,travel behavior,and population characteristics vary uniquelyacross the regions of the US. Participation in out-door activities is strongly related to age, ability anddisability, race,education,and income. These factorsinfluence the types of recreation opportunities avail-able and in which people participate.Approximately 690 million acres of federal lands areused for recreation,of which 95% are in the West.State and local governments manage over 54 millionacres,of which 30 million (55%) are in the East(Cordell et al.,1990). Most of the downhill skiingcapacity is in the western US while most of thecross-country skiing capacity is in the East.The num-ber of people participating in recreation is expectedto continue to increase for many decades. The

509


Analyses of these particular scenarios indicate thatforest productivity gains increase timber inventoriesover the next 100 years. This increased wood sup-ply results in reductions in log prices,which,inturn,decrease producers’profits. At the same time,lower forest-product prices mean that consumersgenerally benefit (Figure 8). The net effect on the

importance of recreation opportunities near urbanareas is rising as US preference shifts from long-dis-tance vacations to frequent close-to-home trips(Cordell et al.,1990). A little over 14% of privatelands are open to recreation,but this amount isdeclining as lands are converted to other uses oraccess is restricted. Projected increases in per capi-ta income will likely contribute to higher demandfor snow-related recreation,although land- andwater-based activities will continue to dominatetotal recreation patterns (USDA Forest Service,1994).

Potential Impact of Climate Change on Timber andWood Product Markets, and Recreation

Adaptation in forest land management and timbermarkets. The possible degree and uncertainty inthe flow of value from forests — goods,services,and amenities — is influenced by the combined(and individual) uncertainty in changes in climate,forest productivity, and the economy. Comparisonsof earlier with more recent analyses indicate thatdifferences in assumptions,modeling structure,andscope of analysis such as spatial scale significantlyaffected the socioeconomic results. For example,an early study of Scandinavian boreal forests con-cluded that increased warming would benefit high-er latitude regions (Binkley, 1988); yet when theglobal trade patterns were analyzed,other regionswith lower production costs benefited most (Perez-Garcia et al.,1997). Early analyses (e.g.,Smith andTirpak,1989) that did not include economic forcesconcluded there would be significant damage to USforests. When active forest management was includ-ed in a previous economic analysis,timber marketsadapted to short-term negative effects of climatechange by reducing prices,salvaging dead anddying timber, and replanting species appropriate tothe new climate (Sohngen and Mendelsohn,1998).The total of consumer and producer benefit (sur-plus) remained positive.

For this assessment,the dynamic optimizationmodel FASOM (Adams et al.,1996,1997;Alig et al.,1997,1998) was used to evaluate the range of possi-ble projected changes in forest land area,timbermarkets,and related consumer/producer impactsassociated with climate change (Irland et al.,2001).The range of scenarios considered alternateassumptions about 1) climate (the Hadley and theCanadian scenarios),2) forest productivity (the TEMand CENTURY biogeochemistry models),and 3)timber and agricultural product demand (deter-mined by population growth and economicgrowth).

Figure 8. Prices for standing timber under all climate change sce-narios remain lower than a future without climate change (base-line). Prices under the Canadian scenario remain higher thanprices under the Hadley scenario when either the TEM or theCENTURYmodel are used. (Irland et al., 2001). See Color PlateAppendix

Average Price for Standing Timber in US Forests

Change in Timber Product Welfare from 2001 to 2100

Figure 9. Increased forest growth overall leads to increased woodsupply; reductions in log prices decrease producers’ welfare (prof-its), but generally benefit consumers through lower wood-productprices. Welfare is present value of consumer and producer surplusdiscounted at 4% for 2000-2100. (Irland et al., 2001). See ColorPlate Appendix

more tropical areas. Effects on fishing opportunitieswill likely vary with warming waters increasing fishproduction and opportunities for some specieswhile decreasing habitat and opportunities for coldwater species.

Recreation is likely to expand in mountainous areaswhere warmer temperatures attract more people tohigher elevations. Skiing is an important use offorested mountain landscape and is sensitive to theclimate in the mountains. Competition within theskiing industry is strong,with successful ski areasattracting customers by providing high-speed lifts,overnight accommodations,modern snowmakingand grooming equipment,and other amenities.Small ski areas often cannot compete and manyhave closed or have been annexed by adjacent larg-er areas (Irland et al.,2001). Climate change willlikely alter the primary factors influencing the abili-ty of a ski area to make snow;namely temperature,water availability, and energy. Higher winter temper-atures could possibly increase the amount of snowmelting,the number of rain events,and decrease theopportunities for snowmaking while increasing theneed for machine-made snow. The efficiency ofsnowmaking declines as temperature warms. Thecost of making snow is 5 times greater at 28˚F (–2˚C) as at 10˚F (–12 ˚C). The annual electricityusage at Maine’s Sunday River ski area is approxi-mately 26 million kilowatt-hours (Hoffman,1998),ata cost of nearly 2 million dollars per year and mostof this energy is used for snowmaking. On theother hand,higher winter temperatures and morerain events together could possibly result inincreased water availability for snowmaking andperhaps increased visitation with fewer extremelycold days. Changes in the geographic line of persist-ent subfreezing winter temperature could possiblyalter the location of winter recreation by affectingthe feasibility of snow-making,such as in the south-ern Appalachian Mountains. While the impacts ofclimate change on the US ski industry remain specu-lative,ski areas operating in marginal climates arelikely to be seriously affected.

ADAPTATIONS: FORESTMANAGEMENT STRATEGIES UNDER CLIMATE CHANGEA major challenge in developing strategies for cop-ing with the potential effects of climate change onforest processes and subsequent values is that themagnitude and direction of such changes at thelocal level remain highly uncertain. In addition,

economic welfare of participants in both timber andagricultural markets was projected to increase in allscenarios from between 0.4 to 0.7% above the cur-rent values (Figure 9). Land would likely shiftbetween forestry and agricultural uses as these eco-nomic sectors adjusted to climate-induced changesin production. Although US total forest productiongenerally is projected to increase in these analyses,hardwood output is higher in all scenarios whereassoftwood output increases only under moderatewarming. The extent of these changes varies byregion. In these analyses,timber output increasesmore in the South than in the North,and sawtimbervolume increases more than pulpwood volume.

While previous studies and this analysis differ in thedegree of market and human adaptation,one generalconclusion is that timber and wood product mar-kets will likely be able to adjust and adapt to climatechange (Irland et al.,2001). Assumptions aboutchanges in population,land use,trade in wood prod-ucts,consumption of wood products, recreation pat-terns,and human values are highly uncertain on acentury time scale. For example,if human needsfrom forests increase over the next 100 years andimports are limited,the socioeconomic impacts ofclimate change on forests would be greater than ifneeds are low or products can be imported fromareas where climate increases forest growth. Thus,assumptions about change in human needs in theUS and overseas,and about climate change effects inother parts of the world,are likely to be the majorfactors that determine socioeconomic impacts onthe US.

Recreation. Outdoor recreation will likely bealtered as a result of changes in seasonality of cli-mate,and air and water temperatures (Irland et al.,2001). Secondary impacts of environmentalchanges,such as increased haze with increased tem-peratures,and degraded aquatic habitats underchanging climates,will also likely affect outdoorrecreation opportunities. Because recreation isextremely broad and diverse in its environmentalrequirements (Cordell et al.,1999),it is difficult togeneralize about the impact of climate changeacross recreation as a whole (Wall,1998). Change inbenefits to consumers,as measured by aggregatedays of activities and economic value, vary by typeof recreation and location (Loomis and Crespi,1999;Mendelsohn and Mackowiki,1999). In some cases,recreation in one location will be substituted forrecreation in other locations. For example,tempera-ture increases will likely extend summer activitiessuch as swimming and boating in some forest areas,and substitute to some degree for such activities in510


511


Mitigation measures,such as irrigation,mighttemporarily support specific gene pools untilnew and stable environments are identified.Recovery efforts can focus either on managingthe state of the system immediately after the dis-turbance,or managing the ongoing process ofrecovery. Recovery can be enhanced by addingstructural elements that create shade or othersafe sites necessary for reestablishing vegetationor that serve as perches for birds (and thusplaces where seeds would be dispersed).Alternatively, late successional species might beplanted to speed up succession.

There will likely be surprises in how changes inclimate alter the nature of forest disturbances andthe forests themselves. A monitoring programcould determine the influence of climate changeon forests and the natural disturbance regimes.Programs for monitoring the impact of forest dis-turbances currently exist for insects,pathogens,and fire. However, few surveys quantify the extentand severity of damage from wind,ice storms,andlandslides. Further, reserve areas such as wilder-nesses,are not currently monitored. Informationfrom monitoring programs could then be used toupdate risk assessments in management plans andprescriptions in an adaptive management sense(Walters,1997) or to assess the regional vulnerabil-ity of landscapes (O’Neill et al.,2000). A risk rank-ing system could identify aspects of the forestmost susceptible to severe repercussions from dis-turbance under a changing climate.

Human Adaptation to ForestChanges under an Altered Climate

While recent studies have suggested that humanactivities associated with markets will likely allowfor adaptation to ecological changes by changingland use practices,production and use technolo-gies,and consumption patterns,it may be impor-tant to examine the breadth of possible adapta-tions. As forest conditions change costs of goods,services,and amenities,investors,producers,andconsumers will likely shift investments and con-sumption decisions. However, it is possible thatseparate effects on producers and consumers oftimber products,especially in different regions,could be large and in opposite directions.

Helping human communities to adapt to changingforest conditions could include reducing potentialsocioeconomic impacts by mitigating carbonbuildup in the atmosphere. Carbon buildup in theatmosphere can be slowed by changing forest man-

potential climate-induced changes in forest process-es must also be put into context of other human-induced pressures on forests,which will likelychange significantly over future decades and cen-turies. Finally, such strategies must be considered interms of their economic viability. Strategies for cop-ing with climate change could include:1) active for-est management to promote forest adaptation to cli-mate change,and 2) assistance to urban and ruralcommunities to adapt to changing forest conditions.

Active Forest Management underClimate Change

Current forest management capabilities provide ini-tial guidance on how to manage forests under afuture changing climate. The value of the forest,anychanges in natural disturbance regimes,and theavailable environmentally and economically accept-able management options influence the copingstrategies for forests. One way forests may be aidedin adapting to climate change is to take steps todecrease other forest stresses,such as atmosphericdeposition. Strategies for coping with disturbancesin forests will vary regionally.

If climate change results in alteration of such distur-bance regimes as fire,drought,or insects and dis-ease,managers could try to cope with these impactsby influencing forest ecosystems prior to the distur-bance,mitigating the forest disturbance itself, manip-ulating the forest after the disturbance,or facilitatingthe recovery process. Prior to the disturbance,theecosystem could be managed in ways that alter itsvulnerability or ability to enhance recovery from adisturbance. For example,trees susceptible to ice orwind storms could be removed,as is common incities. Density and spacing of tree planting could bealtered to reduce susceptibility to drought. Speciescomposition could be changed to reduce vulnerabil-ity of forests to fire,drought,wind,insects,orpathogens. Management could be designed toreduce the opportunity for disturbance to occur.Examples include:limiting the introductions of non-native species;burning restrictions;and prescribedfires to reduce fuel loads. While manipulations offuel type,load,and arrangement could protect localareas of high value,fuel management may not becomplicated for larger landscapes. Some distur-bances,such as fire,insects,disease,and drought canbe managed through preventive measures,orthrough manipulations that affect the intensity orfrequency of the disturbance. For example, fire,insect,and disease controls are examples of manag-ing to reduce the impact of a disturbance for high-valued forests.

Integration of Climate and Land Use Change intoBiodiversity, Ecosystem Structure, and HigherTrophic Level Models. Climate and land use inter-act in ways that influence biodiversity, implying thatthese factors cannot be considered in isolation(Hansen et al.,2001). Climate and land use jointlyinfluence species distributions through alteration ofdispersal routes and changes in habitat throughchanges in disturbances such as wildfire, flooding,and landslides. Land use may modify the local cli-mate through changes in transpiration, cloud forma-tion and rainfall,and increased levels of drying.Land use models include socio-economic variablessuch as human population size,affluence,and cul-ture (Alig,1986); however, the influence of climateon land use and land use impacts on climate are notconsidered. Global and local climate projectionssimplify the feedbacks between land use, vegetation,and climate. Integrated models of land use and cli-mate are needed to predict the interactions of thesetwo influences on biodiversity.

Understanding the Interactions of Current ForestDisturbances and Climate. Basic information on thedisturbance — frequency, intensity, spatial extent —and the climatological conditions that initiate thesecurrent forest disturbances are currently not avail-able for a number of forest disturbances. For exam-ple,the genesis of ice storms is well-understood,butthe nature of the ice accumulation in relation tostorm characteristics is not well-understood.Similarly, the understanding of what specific clima-tological conditions lead to the formation andoccurrence of small-scale wind events is inadequate.Analyses suggest that future climate variability maypush fire and hurricanes outside of the well-studiedhistorical conditions that initiate them and controltheir intensity and frequency. For most distur-bances,there is a need for better understanding ofthe interactions between climate variability, distur-bance frequency, and spatial patterns.Once the rela-tionships between weather, climate,and forest dis-turbances have been quantified,the occurrences ofsuch disturbances could be predicted to help mini-mize their impact.

Quantifying the Impact of Disturbances on ForestStructure and Function. A key aspect of managingthe forest before,during,or after disturbances is anunderstanding of the disturbance impacts onforests. For disturbances such as fire,the impact hasbeen studied extensively. For others such as icestorms,insects,diseases,and introduced species, amore complete understanding of impacts is neces-sary before we can manage for the disturbance,or

agement and forest industry technology toincrease net carbon sequestration. Carbon storagein forests could be increased by reducing the con-version of forests to other land uses,setting asideexisting forests from harvest,or reducing forestfires (Birdsey et al.,2000;Sampson and Hair, 1996).Carbon storage could also be increased by con-

verting other land uses to forests and by enhanc-ing forest management. Improvements in forestindustry technology could enhance sequestrationby allowing substitution of wood products whereappropriate for products requiring more energyand carbon emissions in production and use,andincreasing the use life of products and recycling ofproducts allowing more carbon accumulation inforests (Sampson and Hair, 1996).

CRUCIAL UNKNOWNSAND RESEARCH NEEDS Linkages between Forest Processes, Air Quality,and Climate Change. It is crucial to begin tounderstand not only the direct effects of CO2,ozone,temperature,precipitation,and nitrogen andsulfur deposition on forests,but also the interac-tive effects of these stresses. Physiological impactspropagate through forest ecosystems by alteringcompetitive interactions among individuals andspecies,litter characteristics,and soil processes.These impacts and interactions feed back to affectphysiological processes, nutrient cycling,andhydrology. These interactions stress the need forintegrated ecological research to improve ourunderstanding of these forest dynamics.

Responses of Terrestrial Animals to ChangingClimate, and Trace Gases. The response of treesto climate,soils,and other biophysical controls isrelatively well-documented,in comparison to otherplants or animals. For native species,and the inter-action of introduced species and natives,under-standing the combined ef fects of climate,carbondioxide concentration,and nutrients such as nitro-gen on terrestrial vertebrates and invertebrates isseverely lacking. The response of vertebrates andinvertebrates to soils and other biophysical con-trols is also lacking. Research is needed to com-bine models of impacts of climate change andother factors on forest pests (insects andpathogens,introduced species) with models ofrange and abundance changes of host species.

512


C L IM A TE CHANGE IMPAC T S ON THE UNITED STAT E S

Documents

Transcript of C L IM A TE CHANGE IMPAC T S ON THE UNITED STAT E S