ABTRACT tc - Education Resources Information Center · 9. Jointed ruler, early 20th century. 10....

re- 100-194

TI

-INSTITUTION

SPONS

PUS DATECOLT R ACTNOTE

SE ,4124 :4

Climate change to the Tear 2000: A Sury 1 cf,Ex ertOpinion. s

Department -of Agriculture, Washington, O.C. ;Institute-for:the Future, Picnic Park, Calif.;National Defense-UnivA, Washington, D.C.; NationalOceanic and Atmospheric Administration (LOC),Rockville,,Bd.Adw.anced Research Projects'Agency (DOD) NashiiD.C.,Feb 78BDA-903-77-C-0914137p.;'Colored grapes and charts may not eproducewell

AVAILABLE PROM Superintendent of Documents, U.S. Government PrintOffice, Washington, D.C. 20402 (Stock Number.008-020-00738-21 No price gucted)w

PRICE NP -$O.83 HC-57.35 Plus Postage.RIPTORS *Climate Otintrol; Climatic Factors; Earth Science;

*.Global Approach; Bicker Educatibn; *Beteorology;Natural Sciences; *Research; *Research Methodology;science Education 4

ABTRACTThis survey of expert opinion was conducted by the

National. fe se University, Wadkington, D.C. tc avant fy tte14411hood of Significant changes in climate and their practicalconseguendes. The major objectives of the.study,.are embcdied in fourtasks. This publication presents the results Of the first task only:the definition and estimation of the 10elihood of climate changes,during the next 25 Y44rs, and the.Conitruction cf climate scenariosof the year 2000 It-includeS an abstract, foreword, summary, threechapters and four appendices. Chapter cre describes the methodologyused in analyzing the information collected. Chapte=r two containsnarrative and stati§tical descriptions of five climate scenarios,ranging from large global cooling tc large global warming. Chapterthree summarizes the aggregatdd probabilistic data of the climatepanelists and compares these data from scenario to scenario, acrosslatitudinal zones, and across three time periods between -now and theyear 2000. (Author/EM)

ig

* * * * ************ *4 ********* Reproductions supplied by EDE are the best that can be made *

* from the original document. *

**** ***********, ********* ******* * *

NATIONAL; INSTITUTE OFSOucoilow

THIS DOCUMENT MAS OPEN nem,-,01/c011 SCTLY Ai oncrivio room

Pomp O oupAINZATMIT *NamAlvan IT potinsop VIEW do oporooto

1,11TATSD DD -NOT OdECIMAIIit'Ywan 6FF ICIAL PiATIONAL INsTiTurr-nOur.OTKotfOOTTON OW "Cum

tA

2

I. Mercury olurnn B rometer, circa1850: d eloped te0 expand scale

I. arid facil tate readihg.

2. Eprty km man dileridar Wall:Clock, mi -19th century, byGilbert.

Metcury Ba °meter, early 19thcentury, u d for measuringortation bar petric pressure.

1

4. Campbell-Stokes SunShine Reeprder,early. 20th *miry, used to record _

duration of iunshicie.\k5. Hygrodqiki la 19th century, hatid

calculator for determining.relativehumidify an, water Content ofambient air.

6. Pyranometer used for measuringtotal global radiation :.

11-. 7.'Shi6's Binnacle and Gimbaled Compass,18th century, 4..ited for nautical

-navigation.

8. Field Cornpass, contemporary.

9. Jointed ruler, early 20th century.10. Aneroid Barometric Element.

11. Gold Bourdon Instrument.

12. Misr aneous meteorological 'tiler-eters.

13. Prismatic cell, used to demonstratedispersion of white li4ht.

14. A Student's Volt-ammeter, early 20thcentury.

15. Aneroid Field Barometer, contemporary

16. Nautical Sextant, early 18th century,used for finding geographical posi-tion by means of celestial bodies.

17. International Cloud Atlas, contempora

7

Illustrations in this e-xt are from a collection of over 200 antique woodcuts and etchings from thHart Archives, New York, NAY.

Cover design and book layout by th National Defense University Visual CoMmunications andPrinting Division.

CLIMATE CHANGETO THE:YEAR 2000

'A SURVEY OF EXPERT OPINION

CONDUCTED BY THERESEARCH DIRECTORATE

.

OF THENATIONAL DEFENSE UNIVERSITY

JOINTLY WITH THE

U. S. DEPARTMENT OF AGRICULTURE

DEFENSE ADVANCED RESEARdiPROJECTS AGENCY

NATIONPL OCEANIC ANDATMOSPHERIC ADMINISTRATION

INSTITUTE FOR THE FUTURE

FORT LESLEY J. McNAIRWASHINGTON, D. C. 20319

FEBRUARY 1978

DISCLAIMER

Analyses and opinions expressed or.implied hereinare solely those of the authors, based on a surveyof expert opinion, an do not necessarily representthe e-views of the National Defense University, theDepartrwt of Defense, the Department ofAgriculture, the National Oceanic and AtmosphericAdministration, or any other U.S. Governmentagency. This document has been cleared for publicrelease by the Office of the Assistant Secretary of.Defense for Public Affairs for open publication andunlimited. distribution.

TECHNICAL ASSISTANCE PROVIDED BY THEINSTITUTE FOR THE FUTUREMENLO PARK , CALIFORNIA

SPONSORED BY THEDEFENSE ADVANCED RESEARCH

PROJECTS AGENCYARPA ORDER 3359

-CONTRACT NUMBER MDA903-77,C-0914

Requests for additional copies should be sent to the DefenseDocumentation Center, ATTN: DEC-TSR=1, Bldg. 5, CameronStation, Alexandria, Virginia 22314

Library of Congress Catalog Card No 78-600020

DEPARTMENT OF DEFENSE

NATIONAL DEFENSE UNIVERSITYWASHINGTON. D.C. 7031

The National Defense University (NDU), a Joint Chiefs of Staff (JCS) organization, was establishedformally by the Department of Defense on 16 January 1976, The Industrial College dif the Armed' Fore s([CAF') and The National Wai- College (NWC), collocated at Fort Lesley J. McNair, Washington, DC, areprincipal constituent elements of the University, along with the Office of the Presidentiand fourlUniversi ydirectorates.

The mission of the University is to prepare selected military and civilian professionals for high levelassittnments in the formulation and execution of national security policy. At NWC, t4e emphasis is onnational security policy formulation and strategy, while at ICAF the focus is on the management ofresources in the interest of national security. Both programs look to the future to try to anticipate therequirements of national security based on US interests, US objectives, and the employment andmanagement of resources for advancing those interests and objectives.

A corollary of the University's mission is research and the preparation of studies related, to nationalstrategy, national security policy formulation, military strategy, and allocatiOn :and management ofresources for national security.

The Research Directorate coordinates University research efforts. Researchers include students and faculty,as well as senior and associate fellows front all of the uniformed services and frommother governmentagencies. The National Security Affairs Institute, an element of this directorate, provides a forum forrepresentatives of the Executive and Legislative branches of government, distinguished scholars, and othercitizens from the private sector, who ire concerned with aational security matters, to exchange views onsecurity issues. An annual National Security Affairs Conference, jointly sponsored with the Office of theAssistant Secretary of Defense (International Security Affairs), brings together senior government officials

with prominent representatives of the academic and civilian communities. Reseatch \ Directoratepublications include the Proceedings of the NAional Security Affairs Conference, an extensive\ monographseries, and books.

Climate Change to the Fear 2000, a study sponsored Jointly by the National Defense University, theDepartment of Agriculture, and the National Oceanic and Atmospheric Administration, was \conductedunder the aegis of the University Research Directorate. We at the National Defens'e University are deeplygrateful for the professional contributions made by all who participated in this study. Without the generouscommitment of time and the conscientious cooperation of the many individuals involved, such a complexstudy would not have been possible.

R. G. CARD RLieutenant General, USA.President

ABSTRACT

An attempt to quantify perceptions of global climate change to the year .

200 has been the initial focus of an interdepartmental' study at TheNational Defense University. Subjective probabilities for the occurence ofspecified climatic events were elicited by a survey of 24 climatologists fromseven countries. Individual quantitative responses to ten major questions wereweighted according to expertise and then averaged, a method of aggregationwhich preserved the climatologists' collective uncertainty about futureclimate trends. The aggregated subjective probabilities were_used toconstruct five possible climate scenarios for the year- 2000, each having a"probability" of occurrence.. The agg?egated probabilities of contingentevents are compared from scenario to scenario, across zones of latitude, a)ndby time periods.

The derived climate scenarios manifest a broad range of perceptions aboutpossible temperature trends to the end of this century, but suggest as mostlikely a climate resembling the average for the past 30 years.- Collectively, therespondents tended to anticipate a slight global warming rather than acooling. More specifically, their assessments pointed toward only one chancein five that, changes in average global temperatures will fall outsiO,e the rangeof 0.3°C to +0.6°C, although any temperature change was generallyperceived as-being amplified in the higher latitudes of both.hemiipheres. Therespondents also gave fairly strong credence to a 20- to 22-year 'cycle ofdrought in the High Plains of the United States but did not agree on itscauses.

Consequences of the possible climatic changes delineated in the scenarios arebeing considered in subsequent phases of this research. A generalized climateresponse methodology will be demonstrated by its application to crop yielddata gathered from a survey of agricultural scientists. The policy implicationsof the resultant climate/crop scenarios will be examined using a world foodeconomic model.

i i i

TABL. OF CONTaNTS

= ABSTRACT

FOREWORD ix

a

ACKNOWLEDGMENTS xiRESIDENT STAFF xiiiADVISQRY'GROUP _xiii

CLIMATE PANEL xv

SUMMARY, xvii

CHAPTER ONE-METHODOLOGY 1

RESEARCH APPROACH FOR TASK I 1

ANALYSIS OF DATA 3

CLIMATE SCENARIOS 12

CHAPTER TWO-CLIMATE SCENARIOS 17

LARGE GLOBAL COOLING 17

MODERATE GLOBAt COOLING 21

SAME AS TOE LAST 30 YEARS 25MODERATE GLOBAL WARMING 29LARGE GLOBAL WARMING . . . .. . . . .. 33

CHAPTER THREE-DISCUSSFON OF SCENARIOS ANDCLIMATE PROBABILITIES 37

,GENERAL 37GLOBAL TEMPERATURE CHANGES . . 37LATITUDINAL TEMPERATURE CHANGES 39GROWING SEASON 43

/,2

PRECIPITATION 46DROUGHT AND MONSOON FAILURE . . . . 51

CONCLUSIONS ON CLIMATIC VARIABILITY 57

APPEND-I-X A-CtIMATE QUESTIONNAIRE 59

APPENDIX B-OUTLOOK FOR 1977 CROP YEAR &THE PERSISTENCE OF DROUGHT

APPENDIX C-NUMBER OF RESPONSES & AVERAGEEXPERTISE RATINGS

7

101

APPEND/XD-REFERENCES 103

LIST OF FIGURESI-1 SELF AND PEER RATING

1-2 GLOBAL TEMPERATURES

1-3 SAMPLE RESPONSE TO QUESTION I 6'

CUMULATIVE PROBABILITYNNCTION FORQUESTION I

I -5 EQUIVALENT DENSITY FUNCTION FORQUESTION 1 . ...... . ... ... . . .

1-6 ADDING TWO DENSITY FUNCTIONS FORQUESTION I

1-7 NORMALIZED DENSITY FUNCTION FOR TWORESPONDENTS TO QUESTION I 9

1-8 PROBABILITY OE MEAN PIOATHE -HEMISPHERETEMPERATURE CHANGE BY THE AR 2000 ASDETERMINED BY THE PANEL Of CLIMATICEXPERTS 10

1-9 PROBABILITY OF MEAN NORTHERN HEMISPHERETEMPERATURE CHANGE TO THE YEAR 2ooaAsDETERMINED BY THE PANEL OF CLIMATICEXPERTS . ... . ..... . 11

111=3_,

QUESTION VIMID-LATITUDE DROUGHT 14

RELATIVE CLIMATIC INFLUENCE 38

PROBABILITY OF TEMPERATURE CHANGE INTHE NORTHERN HEMISPHERE 40

PROBABILITY OF TEMPERATURE CHANGE . 42

111-4 CHANGES IN HIGHER MID-LATITUDE GROWNGSEASONS BY THE YEAR 2(100 44

III-5

111-6

111-7

B-1

vi

CHANGES IN PREcIPIJATION ANDPRECIPITATION VARIABILITY 47

PROBABILITY OF DROUGHT 52

PROBABILITY OF MONSOON FAILURE 56

DROUGHT PERSISTENCE IN THE UNITEDSTATES 99

LIST OF TABLESI-1 GORRELATI IN BETWEEN SELF AND PEER

RATINGS

1-2 CONVERSION OF EXPERTISE RANKING TOWEIGHTED SCA 4

1-3 'DEFINITION OF TEMPERATURECATEGORIES 12

1-4 PERCENTAGE OF GROCIPEDPROBABILITYDENSITIES LYING IN EACH TEMPERATURECATEGORY 13

1-5 FREQUENCY OF DROUGHT IN U.S. IN1991-2000 14

11-1 LARGE GLOBAL COOLING . . . 18

11-2 MODERATE GLOBAL COOLING 22

11-3 SAME AS LAST 30 YEAS 26

111, 11-4 MODERATE GLOBAL WARMING 30

I I-S LARGE GLOBALWARMING = 34

B-1 PROBABILITY OF YIELD CHANGE DUE TOWEATHER . . . . . .. .... 97

B-2 USDA YIELD ESTIMATES . . 97

B-3 LENGTH OF DROUGHT IN THE UNITEDSTATES ...... . .

C 1 NUMBER OF RESPONSES AND EXPERTISERATINGS

. 98

101.

vii

,

4

,,N

r

so--,

FOREWORDr1

Within the last few years there has beerran increasing/;public_awareness of theimpacts of weather and climate ongrapRind. This concern was highlightedby the severe 197677 winter in the eastern half of the United States andsubsequent drought conditions in portions of the country. Controv ersy hasarisen about whether the climate is merely fluctuating or is in 06g-termtrend of change. Any significant change in climate would- have profoundimpacts upon U S policies and programs with regard to World foodproduction and reserves as well as a wide range of energy-related and otherpolicy matters.

.,

In view of the potentially 'serious implications of climatic change, TheNational efense University in the fall of 1976 initiated an interdepart-mental resigfch*project to qyantify the likelihood of significant changes in;limate and their practical consequences. The major objectives of the studye embodied in the four tasks described in the Summary. This report

presents the results of(

the first task onlythe definition and estimation ofthe likelihood of climatic changes during the next 25 years, and theconstruction of climate scenarios for the year 2000. Findings of theremaining research tasks will be reported later.44 .,

The causes of global climate4change remain in dispute. Existing thlories ofclimate, atmospheric models, and actuarial experience are inadeate tomeet the needs of policyMakers for informatiSitabout future climate, In theong run, research may lead to reliable forecasts of cliniirte. For the present,

wever, policymakers have no recourse but to heed expect judgmentssubjktive and contradictory though They may beabout future worldclimate and . its effects on agriculture and other sectors of the economy.

%--- Informed, expert judgments on the likelihood of change, or the odds forarepetition of some event, are useful t -e decriionmaker weighing the costs,orth

ranef its, and risks of alternative p icies. To marshal the full spectrum of

rt

Vogists.arefully selected panel of catolW e hope the analyses of the-xpet perceptions, we submitte .a., structured climate question0aire to a

probabilistic responses to the questionnaire will lead to a better quantitativeunderstanding of those weather climate issues that are most relevant toimportant public concerns about the future of agriculture and the worldfood situation.

) 1

FOREWORD

x

The mod s operandi for this project was the establishment of a small,interdisci linary staff at The National Defense University to act-as a focalpoint for several major sources of substantive research contributions. Thefirst and seminal source is a standing Advisory Group of eminent scientists,and administrators who provide guidlande and represent many potential usersof the research results.' Another .invaluable source is the Institute for theFuture, Menlo Park, California, which furnishes ongoing advice on futuristicsmethodologies and technical assistance through the-auspiget' of the .DefenserAdvanced.-Research Projects Agency. A third source is the ad hoc ClimatePariel o0'prominent individuals whose subjective assessments :of future .

climate formed tfte basis of this report. In addition, there .re two other ad.. 1

hec groiips of experts concernedwith subsequent phaties of this research, effort One group is furnishing data on the response of crops to cilimate

.changes; the other is preparing to analyze_ the piolicy implications of`climate/crop scenarios for the year 2000.

Chapter I of this report describes the methodology used in analyzing theinformation contributed by the panel of climate experts from the UnitedStates and abroad. Chapter II contains narrative and statistical descriptionsof five climate scenarios, ranging from-large global cooling to large globalwarming. Chapter III summarizes the aggregated probabilistic data of theclimat anelists and compares these datafrom scenario:to scenario, acrossla udinal zones, and across three time periods between now and the year2000. Appendix A is the climate questionnaire sent to the Climate Panel.Because of a limited response, only part of Questicin VII, "Outlook for 1977Crop Year," was analyzed, This information is contained in Appendix 13,along with an analysis of subjective probabilities concerning the persistenceof droughts in the United States. Appendix C provides information on thenumber of respcindents and the average level of expertise for each of thequestions in the questionnaire. Appendix D contains references citedty thepanelists in their answers to the questionnaire.

ACKNOWLEDGMENTS

7g:

ACKNOWLEDGMENTS

Although the .Research Directorate of The National Defense Universityserved as the focal point and "broker" for this climate project, many outsideindividuals generously volunteered thei tirne and knowledge to make itwork.

Mr. Joseph W., Willett, Dir ctor of the Foreign Demand and CompetitionDivision in the U.S. pep rtment of Agriculture, conceived the climateproject. In- 1975-1976, w ile he was a Senior Research Fellow at theStrategic Research Group o The National War College (the predecessor ofthe Reseair Directorate of The National Defense University), Joe persuadedus of the need to pursue this undertaking. Now back`bt-the USDA, Joe hasremained inclose touch with the project and his continuing counsel has beeninvaluable:

Vice Admiral M. G. Bayne, U.S. Navy (Retired), and Lieutenant GeneralR. G. Gard, Jr., U.S. Armythe farmer and current presidents of theuniversitynot only approved this novel research, but supported it withenthusiasm.

The study could not have proceeded without funding and encouragementfrom the Cybernetics Technology Office of the Defense Advanced ResearchProjects Agency. Dr. Robert A. (Young is Director of the CyberneticsTechnology Office,- and Lieutenant Colynel Roy M. Gulick, Sr., U.S. MarineCorps, is our immediate point of contact in DARPA. Among other things,DARPA funds secured the services of the Institute for the Future. Dr, RoyAmara, president of the Institute, and Dr. Hubert Lipinski were alwaysavailable to suggest appropriate methodologies and to provide technicalassistance in the analyfcal phases of the study.

From its inception, the project has employed a small team= of residentinvestigators to seek the judgments and perceptions of experts, to collatedata and prepare reports, andno mean taskgenerally to orchestrate thecontributions of the many far-flung participants in the study. Working withthis team, the most competent and dedicated group with which I have everbeen associated, has bee'n a great personal pleasure. The resident team hasconsistently received exceptional support from numerous administrativeelements of The National Defense University. The advice and skills of manyindividualssecretaries, stenographers, editors, research assistants, artists,illustrators, type composers,'and printershave gone into the publication ofthis report.

xi

ACKNOWLEDGMENTS

xii

At the outset of its entry into the alien field of climate, it was- evident thatThe National Defense University would require expert guidance anddirection. To meet this need, we formed a standing Advisory Group ofscientists and potent4al users of the:reseal-eta. The Advisory Group met at-Fort McNair in December 1976 an June and December F_.1977. At the

W.first meeting we made it clear- ) that we wanted the adviSors, collectively andindividually, to be activilIs, nd to insure that our efforts were leading tovalid and useful results. At each phase we presented our results and futureplans to the advisors. Their lively and constructive criticism were painful attimes, as when we had ,to rework portions of the research. But we aregrateful that this group saved us from committing some major blunders andfrom straying down some dead ends. Any remaining blunders are ourresponsibility. Many advisors gave liberally of their time and knowledge forat most token compensation.

Last, but far from least, there is the ad hoc Climate Panel which provided thegrist for this report. This panel was carefully selected with the assistance ofthe Advisory Group, Pains were taken to include authorities in global,regional and topical climatology, and paleoclimatOlingy, as well as spokesmenfor differing viewpoints about future climate trends. The climate panelists'response to 'Cr questionnair:e provided the judgments, knowledge, and aboveall, the quantitative perceptions on which this report is based. The panelistslabored with much admitted soul-searching for only a nominal honorarium.We are truly in their debt.

The caliber and representativeness of the Climate Panelthe "balance"among specialties and divergent points of viewwere obviously crucial to thevalidity of this study. In November 1977, drafts of Chapters I and II andAppendix B were circulated to the Climate Panel and the Advisory Groupfor comment. The same drafts were discussed the following month at ameeting of the Advisory Group. The written and oral comments on thedrafts, as well as the findings taken in tow, lead us to believe we achievedour goals of high caliber and adequate balance in the panel of climatologists.(I note in passing that Chapter III, which contains cornparisonsof the data inthe individual climate scenarios of Chapter II and a selectton of the panelists'comments, is the responsibility solely of the resident.rssearch staff.)

To those mentioned above, and to all listed below, my`moSt sincere thanks.

ANDREW J. DOUGHERTYColonel, USAFDirector, Research Directorate

ACKNOWLEDGMENTS

RESIDENT STAFF

In order of their appo tmen as .Senior R7esearch Fellows, tie principalinvestigators resident at he National Defense University were

`1111 r.. William R. Gass :r, U.S. Department of AgricultureColonel Vernon M. Malahy, Jr., U -. Air ForceDr. Paul C. Dairy ple, U.S, Army Engineer Topographic LaboratoryColonel Theodor: H. nin Crampt sr-6 .S. Army

The resident staff eceived valuable assistance on a continuing basis from:

Dr. Richard . Felch; USDA (formerly with the National Oceanic andAtmosbhetic Administration)

-Mr. Douglas Le Comte, NOAAMajor Russ II A. Arnbroziak, U.S. Air Force Reserve

In addition, Captain Bernard C. Diesen, Ill, Lieutenant Colonel David S.Lydon, bo USAF, and Mr. Richard C. McArdle, USDA, made specializedcontributions at various phases of the project.

ADVISI RY GROUP

The co potion of the. Advisory Group has changed over time'. Listed beloware the advisors who participated in at least One of the meetings of thegroup. Dr. J. Murray Mitenen, Jr., was the Senior Scientific Advisor to thepha e of the r earch cvered by Ithis report. He and Doctors Gilman andLa dsberg revie cl early drafts of the climate quA-tionnaire. Among other

"es ecial contribL4ors'to the project were those who pretested the question-n ire: Doctors Bierly, Namias, klewman, Thompson, and Walther. Dr. David

Hershfield of the 'U.S. Department of Agriculture also pretested theuestionnaire.

U.S. Government

DR. DAWSON AHALT, Department of AgricultureCAPTAIN WILLIAM S. ARNOLD, USN, Office of the Joint Chiefs of StaffDR. GERALD BARNEY, Council on Environmental Quality`COLONEL DALE BARNUM, Department of the Air ForceMR. JAMES M. BECK, Department of the ArmyCOLONEL MORTIMER F. BENNET, USAF, Office of the Joint Chic

StaffsDTR. EUGENE W. BIER LY, National Science FoundationCOMMANDER DONALD R. CARVER, Department of the NavyCOLONEL WILLIAM E. CUMMINS, II, Department of the Air ForceDR. FERNAND P. DE PERCIN, Department of the Army

ACKNOWLEDGMENTS

t

MR. FRANK EMERSON, Departr*nt of EnergyDR. CHARLES E. FRENCH, Agency fo7-Interngtional Deve opmentDR. DONALD L. GILMAN, National Oceanic and At °spheric Admin-

istrationMR. LINDSEY GRANT, Department of StateCOLONEL PAUL JANOTA, Department of the Air ForceDR. KARL R. JOHANNESSEN, National Oceanic and Atmospheric Admin-

.-artistrationMR. WILLIAM E. KIBLER, Department of AgricultureMR. ANDREW W. MARSHALL, Office of the Secretary of DefenseMR. JERRY MI LSTED, Office of the Secretary of DefenseDR. J. MU RAY MITCHELL, JR., National Oceanic and Atmospheric

AdministrationMR. EMIL NELSON, Department of EnergyCOLONEL ROBERT M. PFEIFFER, USAF'Office of the Joint Chiefs of

StaffMR. LEWIS PITT, National Oceanic and Atmospheric AdministrationDR.. THOMAS D. POTTER, National Oceanic and Atmospheric Admin-

istrationCAPTAIN KENNETH W. RUGGLES, USN, Office of the Secretary of

Defense

DR. NORTON STROMMEN, National Oceanic and Atmospheric Admin-istration

Universities

DR. GEORGE ALLEN, Oxford UniversityDR. WAYNE L. DECKER, University of MissouriDR. OTTO C. DOERING, Purdue UniversityDR. D. GALE JOHNSON, University of ChicagoDR. HEL'MUT E. LANDSBERG, University of MarylandbR. JEROME NAMIAS, Scripps institution of OceanographyDR. JAMES E. NEWMAN, Purdue UniversityDR. M. RICHARD ROSE,-Alfred UniversityDR. THEODORE W. SCHULTZ, University of ChicagoDR EARL R. SWANSON, University of IllinoisDR. LOUIS M. THOMPSON, Iowa State UniversityDR. PENEkOPE HARTLAND-THUNBERG, Georgetown University

Institutions aid Other Nongovernment Agencies

DR. PIE R R CROSSON, Resources for the Futurze Inc.DR. EDWA D DEAGLE, The Rockefeller FoundationDR. MALCOLM GRAY, Aspen Institute for Humanistic StudiesDR. WALTER G. ROBERTS, Aspen Institute for Humanistic StudiesDR. FRED H. SANDERSON, The Brookings InstitutionDR. ERIC WALTHER, Charles F. Kettering Foundation

ACKNOWLEDGMENTS

CLIMATE PANEL

Listed below are the scientists who responded to the climate questionnaire(Appendix A).

DR. HIDETOSHI ARAKAWA, Tokai University, JapanDR. ROGER G. BARRY, University of-ColoradoDR. WALLACE S. BROECKER, Lamont-Doherty Geological ObservatoryDR. REID A. BRYSON, University of WisconsinPROFESSOR WILLI S. DANSGAARD, University of CopenhagenDR. ROBERT E. DICKINSON, National Center for Atmospheric ResearchPROFESSOR HrRMANN FLOHN, University of BonnDR. HAROLD C. FRITTS, University(of ArizonaDR. W. LAWRENCE GATES, Oregon State UniversityDR. JOSEPH GENTILLt-T4e University of Western AustraliaDR. WILLIAM J. GIBBS, Bureau of Meteorology, AustraliaDR. WILLIAM W. KELLOGG, National Center for Atmospheric ResearchDR. JOHN E. KUTZBACH, University of WisconsinPROFESSOR HUBERT H..LAMB, University of East AngliaDR. HELMUT E. LANDSBERG, University of MarylandDR. THOMAS F. MALONE, Butler UniversityDR. J. MURRAY MITCHELL, JR. National Oceanic and Atmospheric

AdministrationDR. JEROME NAMIAS, Scripps Institution of OceanographyDR. REGINALD E. NEWELL, Massachusetts Institute of TechnologyDR. STEPHEN H. SCHNEIDER, National Center for Atmospheric ResearchDR. JOSEPH SMAGORINSKY, National Oceanic and Atmospheric Admin=

istrationMR. MORLEY THOMAS, Atmospheric Environment Service, CanadaDR. HARRY VAN LOON, National Center for Atmospheric ResearchDR. HURD C. WILLETT, Massachusetts Institute of Technology

xv

SUMMARY

111"k

:11

"

Lk, ,

SUMMARY

Will world climate at the end of thiis century be substantially different Promthat of the past two or three----efcades? Some climatologists postulate acontinuation of the cooling trend that began in the 1940's, others contendthat the world is entering a period of rapid warming, and many foresee aperliod of greater climatic variability.

Such conflicting opinions and their implications inspired a study now underway in the Research Directorate of The National Defense University_Fort -

McNair, Washington, D.C. This project, the first comprehensive attempt toquantify perceptions of climatic changes, is a joint effort of the Department

taJ. of Agricultuke, Department of Defense, and the. National Oceanic andAtmospheric Administration. Technical assistance is being proVded by theInstitute for the Future, Menlo Park, California, throulp he DefenseAdvanced Research projects Agency.

The major objectives of the study are embodied in four tasks:Task I: fio define Ad estimate the likelihood of changes in climateduring the next 25 years, and to construct climate scenarios for theyear 2000.Task II: To estimate the likely effects of possible climatic changes onselected-toris in specific v)untries, and to develop a methodology forcombining crop responses and climate probabilities into climate/cropscenarios for the year 2000.Task To evaluate the domes and inte natiogt polio implicationsof the climate/crop scenarios, and to identify the climatic v riables thatare of key importance in the choice of poll y options.Task IV: To transfer the climate/crop research results end a generalizedclimate response methodology to individuals and organizations con-cerned with the consequences of climatic changes in fields other thanagriculture, and to identify areas of research which might refine orextend the findings of the first three tasks.

This report is a summary of Task I, which was carried out by surveying apanel of climatologists.' The salient finding is that the likelihood ofcatastrophic climatic change by the year 2000 is assessed as being small.'tMore specifically, the respoh s to the survey suggest only 1 chance in 10that average global temperatq in the next 25 years will increase by morethan 0.6'C relative to the earl -y- 1,970's. Likewise, there is only 1 chance in10 that it will decrease by more than 0.3'C. The most likely event will be aclimate which resembles the average of the past 30 years, arising primarily X VI I

SUMMAllY

from a balancing of the warming effect of carbon dioxide with the coolingeffect of a natural climate cycle. However, the respondents tended toanticipate slight global warming rather than a cooling.

Most of the clime panelists perceived that any global temperature changeswill- be amplified at higher latitudes, particularly in the Northern Hemi=sphere. This magnification will be less prcfnouncedi- in the SouthernHemisphere because the larger surface area of southern oceans provides morethermal inertia against change.

liiirtvanelists responses reflect fairly strong support for the continuation ofa 20- to 22-year drought cycle in the High Plains of the United States. Thisperception was 'Aempered by the absence of an agreed-upon causalmechanism for the apparent periodicity. For mid-latitude regions outside theUnited States, there was, more uncertainty and less support for cyclicdroughts than was evident for the United States. Similarly, no periodicitywas idehlified relative to frequencY of dr4ught in the Sahel region of Africaor the failure of the Asian Monsoons.

Collectively, the climatologists expressed considerable uncertainty aboutpossible changes in the amount and variability of precipitationuncertaintynot only with respect to the magnitude of changes but in many cases evenwith respect to the direction of change. This uncertainty was particularlypronounced about possible changes in year-to-year variability. There was,however, some tendency to 'associate more precipitation and decreasedvariability of precipitation vith global warming, and less precipitation andincreased variability with global cooling.

The foregoing conclusions are derived from the quantitative responses of theClimat Panel to a set of:questions about significant climatic factors,including variability. The questionnaire (Appendix Al covered the periodfrom the present to the end of the century. The panelists were asked toassigh probabilities to specified climatic changes and to give the rationale fortheir answers. For each question they were also asked to assign a numericalvalue to their ownt expertise and that of other panelists.

Of the 24 climatologists replying to' the questionnaire, 21 provided aquantitative response to at least one question. The answers of the latter to aquestion on global temperature were weighted on a well-defined scale ofexpertise., and then averaged to yield a probability density function for

-changes in global temperature. The range of perceived global temperaturechanges was partitioc4d into five subintervals upon which are based fiveglobal climate scenarios with corresponding "probabilities" of occurrence.Next, each respondent was associated with a subinterval of global tempera-ture, or scenario, according to his perception of global temperature trends.Finally, detalk of each scenario were developed as conditional -probabili-ties" aggregated in a similar manner from the appropriate panelists' answers

to other questions. Thekse__ questions were concerned with the relativeinfluence of selected atmosphNeric components, the latitudinal digtribution oftemperatur changes, the length and variability of growing season, fkleamount of recipitation and its year-to-year variability, and the frequency. ofdroughts and monsoon failures. Chapter II contains the five climatescenarios. In Chapter III the aggregated subjective probabilitiesof contingeytevents are compared from scenario to scenario, across zones ofilatitude, a- ndby time periods.- (Th4 responses to a question dealing with the outlook forcrops in 1977 and with the persistence of drought in the United States are

0discussed in Appendix B.)

II-1J basic method of weighted averages, described more fully in Chapter I, is

Considered appropriate when respondents base( their replies on a commondatabas: This method has a tendency to'preserve and possibly to _overstateuncer. *

The five climate scenarios and the panelists' comments rna a broadrange of perceptions about future climate. The experts' aggreg_ d subjectiveprobabilities do not reflect a consensus on any narrowly defined climaticissue, but a large majority of the climate panelists were in broad agreement,for example, that the average global temperature is not likely -co change morethan half a degree Celsius by the year 2000. Constructed by a standard

!futuristics methodology, the scenario narratives portray reasonable, coher-ent, d consistent possibilities for world climate around the year 2000.

w ver, as is evident from the probabilistic data that accompany the texteach scenario, it is unlikely that any scenario will materialize in all its

aced specifics. Although the scenarios cannot be viewed a alternate andutually exclusive climate forecasts, they do put plausible quantitativeours on climatic charfr-over the next 25 years. The "probability" of a

scehrio provides policymakers with some measure =perhaps the bestavailableof the confidence to be placed in each of a range of possibleclimatic changes, none of which can be predicted by the current state of theclimatologists' art.

The next objective of this climate research project is to demonstrate foragriculture how climat&Information can be combined with climate responsedata to analyze practical implications of possible climatic change. The twopriviaras-ticins behind this effort are: What are the likely impacts ofpossible climate changes on global food production? What are the policyimplications of these impacts? Among the policy questions of particularirdrest are those concerning food prices, food reserve requirements, foodtrade, and related issues. A generalized climate response prediction method-ology has been developed, and estimates of the response of 'selected crops toparametric climate changes have been obtained from a survey of agriculturalscientists. The response methodology will be applied to these crop yield dataand the climate data reported herein to calculate the expected crop yields

SUMMARY

xix

SUM ARV

Kx

associated with eadh climate scenario. A world footheconorn.ic model will bdused to examine policy implications of the resultant climate/crop scenarios__and their "probabilitiesof occurrence.

OfIAPTFR: F

Research A roach For TaskGeneral FiaturesPanelists' Concerns

Analysis of DataSelf and Peer RatingsProcessing of Responses

Clirnaie ScanarioiUse of ScenariosConstructing ScenariosReview of Scenarios -

Nature of cenarios

CHAPTER.METHODOLOGY

RESEARCH APPROACH FOR TASK 1

diineral Fe: u-

The purpose of Task I was to define and estimate the likelihood of changesin climate during the next 25 years, and to construct climate scenarios forthe year 2000. Infprmation was collected from a carefully selected group ofexperts through the use of a structured questionnaire. Ten separate questibnsdealt with Particular climatic variables and/or specific geographic regions-ofinterest. These topics of inquiry were as follows:

average global tem ratureaverage latitudinal temperaturecarbon dioxide and turbidityprecipitation changeprecipitation variabilitymid-latitude drbughtoutlook for 1977 crop year

sian monsoonshel drought

le gth of the growing season

Each question elicited information about three elements: probabilistic forequivalent) forecasts on a particular climatic variable, reasons for quimtita-tive estimates, and self and peer expertise rating. The complete questionnaireis contained in Appendix A.

A panel of climatological experts from the United States and abroad wasselected by the research team, with assistance from the project AdvisoryGroup. The panelists were selected both for their competence in the field ofclirtiatology and for the diversity of views which they represented. The list ofpanelists responding to the questionnaire appears in the acknowledgments.

The questionnaires were sent to 28 panelists and 24 were returned. Of these,21 contained quantitative information. Appendix C lists for each climatequestion .the number of panelists who submitted quantitative estimates andthe average of their expertise

1

2

Panelists Concerns

Most resixindepts, as well as some Of the inviie panelists who'declined to--Participate, voiced some degree. of apprehension or concern about the

uestionnaire and the um (end possible abust) of the information. derivedfrom their respOnses. These concerns centered on the following issues:

the lack of sufficient actuarial experienbe, comprehensive. theories, or,adequate models to support the quantitative estimates given in the

alquestions,

the possible su ppression of the full range of uncertainty accompanyingresponses,

the risk of being an un ting pa science by consensus.

The following comments by panelists reflect these concerns:

To the best of my knowledge, there exist, in general, no tech-plques formaking climate fotecasts that have demonstrated skill in.the sense that theforecasts are betide- than a forecast of the Jong-term average statistics.Knowledge of even the long-term average statistics (means, variances,extremes, conditional probabilities, etc.) would be most useful for somepurposes, but even this data is not readily available.

I think .that the strongest message to come from your questionnaire willbe that we lack the basis for predicting even the grossest aspects ofclimate. ,

We possess no skill for forecasting beyond a short period, other than thatwhich probabilities based on a frequency distribution can provide. Onlya deterioration of elimate will fife the imagination of the experts'.Prophets become known for their prophecies of doom. A prophecy ofstatus quo or improvement would not be interesting.

There is a good deal of guesswork involved, due to uncertainties, aboutfeedback mechanisms, the importance bf aeross:51s,. the general circulationin the atmosphere and oceans, and many other factors.

I feel that one of the most important outcomes of your study could be aclear statement of our present ignorance. That in itself should clearlyindicate the need for contingen plans.

In the preparation of this report, the project team has given considerableattention to the foregoing concerfis in analyzing the data and aggregating therange of viewsand the expressed qualificationsprovided by the respond-ents. Realizing th'at confident predictions of climate are beyond the state of

the art, the woject team has proceeded on the assumption that expertprobabilistic judgments, properly qualified, -constitute ,the best availableguidance for those who mitst make pblicy in mattisrs affected by climate.The climate data in the rbpori bespeak uncertainty and a' wide range ofperceptions. In the-descripti9n of the methodology and the presentation ofthe analysis 'and results, appropriate a eatrhave been introduced to avoidmisunderataridisq,_,

ANALYSt OF DATA

Self and Peer Ratings

An interesting and tifieful feature of the quasfionnaire was the conoept.of selfand peer ratings. Figure 1-1 is an excerpt of the instructions provided at theend of each question and designed to assess the respondents' expertise.

Figure 1.1

SELF AND PEER RATING

Using the self-ranking definitions. provided in the instructions, pleaseindicate your level of substantive expertise on, this major question.

5 4 3 - 2 - 1

Again using the self-ranking guide, please identify those other respondentswhom you would rate as "expert (5)1 or "quite familiar (4)" in {heir

s an Swer to this particular question'.

The categories from 5 to 1 (expert, quite faMiliar, familiar, casuallyacquainted, and unfamiliar) were carefully defined in the questionnaire.Table 1-1 shows a sample of the degree of correlation between self and peer

ratings for five ..respondents on Question I. The general agreement between

.self and peer ratings is fairly evident by a scan of..the two right-hand columns

in the 'table A detailed analysis of the correlation between self ratings andthe mean of peer ratings shows it to have a value of 0.52 at a significancelevel of 0.007: This is considered-arfairly high correlation.

A simple .averaging of self and peer ratings for each respondent on eachquestion, rounded to the nearest integer valUe, provided a weighting that wassubsequently used in aggregating responses. The particular weighting scalethat was used is shown in Table I-2._ Levels of expertise falling below"familiar" ("casually acquainted" and "unfamiliar ") ware not used in theprocessing. Of the three levels shown in Table 1-2, the "expert" category was

MET4-0 DOLOGY-le;

Table 1-1 e zl ,t1

CORAELATION BETWEEN SELF AND PEER RATINGS(Examples cfrom Question I)

Respondent Rating

frequency pf Peer RatingsQuite

Expert Familiar

A

4

ExpertExpert

ul e a

Quite familiarFamiliar

104

343

weighted twice as heavily as the "quite familiar". category and the "quitefamiliar- was wellhted twice as heavily as "familiar.' In effect this reflectsthe largely empirical and intuitive notion-thit an expert's opinion is worthabout twice as much as one who is _"qpitie famtliar," which in turn is worthtwice as much as an individual who is-rank "familiar with a topic.. _

Table 1-2

_CONVERSION OF EXPERTISE RANKING TO WEIGHTED SCALEExpertise

ExpertQuite familiarFamiliar

Weight

2

Processing of Responses

The general schema for processing the information from the que onnaireswas as follows:

tabulate each respondent's probability density function with respect tochange about a particular variable at a given time, or, derive theprobability density function from graphical information provided bythe respondent.

multiply each probability density function by the appropriate expertiseweight (as described earlier).

add the weighted density functions of respondents.

divide the weighted and aggregated density functions byexpertise weights to norrnaliz the group response.

sum of

combine the panel's responses on each climatic variable into a set ofscenarios spanning the range of uncertainty or range of conditionsdescribed by the respondents.

A

AL

Question I, dealing with possible changes in global mean temperatu're , was a

pivotal question because perceptions of global mean temperature greatlyinfluence perceptions .with respect to the climate variables treated insubsequon>1 questiorts.

-

duestion I is based on Figure 1:2, a ptot of historical changes in annual meantemperature during the past century. Each respondent was asked to provide

three esitimates of the future 'course of possible changes in globaltemperature to The year 2000. The first estimate was to be a temperaturepath to the year 2000 such that thee was only 1 chance in 10 that theactual path could even lower. The second estimate was to be a path withan ev)n chance t _t temperature coutd be either lower or higher; and the

. ,

third was a ppth based oh 1 chance in 10 that it Could be higher..!

GLOBAL TEMPERATURE

METHOD

Historical record oVehanges in annual mean temperatureduring the past century for the latitude band, 0-80°N.

0.8

0.6

U 0.40

rere

.2

ti

0.21860

The period 18801

1880 1900 1920Year

4 is the zero reference .base,

1940 1960 1980 2000

From Mitchell, NOAA

For the purpose of this study, "global temperature"mean temperature between and Kr north latitude.

rp

ed as equivalent to annual'5

IET L Y

Figure 1-3 shows a gam* rp© o Question I. by a single respondent.Each of the three estimates could be drawn in any furtttional fotrn desired.Percentiles of 10, 50, or 90 can be read off for any year b_ etweer the` "present" (the end of the plot in Figure 1-2) and the tear 2000.

,

'Figure 1-3

SAMPLE RESPONSE TO QUESTION IActihal example of a sfnWeresponse to the inarlurtions

0.8

0.6

0.4

ar

0.2

-0.21860 1880 1900 1920

The period 1880 -18$4 is the hero referente base. Year1940 1960 980 2000

6

The processing of responses will now be illy strated using the answers to thisquestion by single respondent. Figure 1-4 is 4 plot, of the informationshown irk Figure 1-3 for the year 2000, converted to a cumulative probabilityfunction in which the ends of the function have been extended beyond the90th percentile and below the 10th percentile in a linear approximation. Forexample, the respondent has indicated a '10 percent chance that thetemperature will change by 0.04 °C or less, a 50 percent chance that it willchange b 0.2°C or less, and a 90 percent chance that it will change by0.47°C r less. Similar values can, of course, be obtained for any otheryear from igure 1-3.

-"These temperature changes ) -are in relation to the zero reference base period,1880-1884, as shown in Figure 1-2.

Figure I=

CUMULATIVE PROBABILITY FUNCTION FOR QUESTION I

-0.5 0.0Temperatu e Cha

The period 18801884 is the zero reference bas-IN ,

/Figure 1-5EQUIVALENT-DENSITY FUN TION FOR QUESTIONL I

0:5 r

°C) by the Year 20'001.0 1.5

1:0 0.0Temper

0-1884 is the zero reference base.

0.5 1.0ure Change 1"0) by the Yea _000

1.5 2.0

The next step is to convert the cumulative probe nctron into anequivalent density function by taking the first derivitive ri.f the plot inFigure 1-4. Since the plot consists of tvdp straightline segments, we havebasically two degrees of freedom, or two leVels in the density function,which is shown in Figure 1-5. The area under_athe curve intercepted by anyparticular temperature rangels eciAl to the probability of occurrence of thatparticular temperature range, and the total area under the curve in Figurel,,5is unity.

Figure 1-6 shows unweighted densityefunctions from each of two:Yespond-enti. The two functions are next weighted by the appropriate expertiseweights, added and then divided by the sum of the weights to obtain thecombined and normalized density function fo-r the two respondents. Againthe area under the curve of this combined and normalized density function,shown in Figure 1-7, is equal to unity:

Figure 1-6

ADDING TWO DENSITY FUNCTION S FOR QUESTION I

2.50

1.r50

0

in- 1.00

0

0.50

I

QUI E FAMILIAR

I I I_ L., _ I __J I I I

-1 0 -0.5 . 0.0 0.5 1.0Temperature Change 1°C) by the Year 2000

The period 1880-1884 is the /Aro reference base. - r

1.5 2.0

The procedure outlined above is repeated for the responses of each .of theother panelists. Figure 1-8 is a plot of the aggregatokl normalized responses ofthe full panel for the year 2000. An analogous procedure yields probabilitydensity furictions of mean global temperature change for the years 1975,1p8o, and 1990._ The information contained in the probability densityfunctions is shown in Figure 1-9 as extensions to Mitchell's original curve.The extensions on the curve show the 10th, 50th and 90th percentilerfdr

METHODOLOGY.

each year from the "present" to the year 2000. Intermediate percentiles arealso plotted: Thus, Figure t-9 is a summary of the aggregated responses ofthe panelists with respect, to global temperature.

NORAALIZED DENSITY FUNCTION FOR TWO RESPONDENTS TO QUESTION I

; ? --....--.- .;-.:. ',.:.-..-A,....,..,

,:: , 1

0.0 0_5 ._ . 1.0 1.5 2.0Temperature Change (QC) by the Year 2000

The period 1880-1884 is the zero reference base.

-1:0 -0.5

1

In aggregating the responses by the method of weighted averages, it has been_ assumed that the respondents are drawing from the same general information

base and, therefore, that their information is highly dependent. In such casesof information dependence among respondents, it is customary to use themetho of weighted averages to aggregate respo -es. All responses are usedand w ighted by the respondents' expertise as pe ceived bylhemselvesandtheir peers. The shape and range of the aggreg eci curves are not acutelysensitive, to the weighting. system used. The-met od is -conservative" in thesense, that the_derived propability curves tend to be broad and to overstateuncertainty as a result of° the additive treatment of the individUalisubjectiveprobabilities. Had the responses been based on independent information, a;multiplicative treatment of the` "-individual probabilities would have beenmore appropriate, and the derived probability curverould have shown lessdispersion. 9

METHODOLOGY

'Figure 1-8

BABILITY*OF MEAN NORTHERN HEMISPHERE TEMPERATURE CHANGEE YEAR 2000 AS DETERMINED BY THE PANEL OF CLIMATICEXPERTS.

2.00

0-50

Scale A q0 -0.5 00 0.5 1.5 20 -

Scale 8 4 .01?-:-51

0.0 0.5 1.0 1.5.

---' Temperature Change 1°C) by h Year 2000

Scale A is based on the period 1880=1884 as the zero reference base (see Figure l-21.Scale B is based on the period 1965-1969 as the zero reference base (see the end point on Figure l=2).

10

METHODOLOGY

Figure 1-9

PROBABILITY OF MEAN NORTHERN HEMISPHERE T APERATURE CHANGETO THE YEAR 2000 AS DETERMINED BY THE PANEL OF CLIMATIC EXPERTS

1.0

0_8

0.6

0.2

0

0.2

Change

in annualtemperature,0-80 N.latitude

i'C)mean

860 1880 1900 1920 1940 1960 1980 20Yt!ar

The period 1880-1884 s th e fern reference haw.

0.95'1.9)

0.8

0.8

0.46%5)0.4

0,2

0.05' (I)

9,200

11

METH°ILY

12

CLIMATE SCENARIOS

Use of Scenarios

A convenient procedure for dealing with a range of uncertainty when it isnot possible to construct quantitative models is through the use of scenarios,which may be considered plausible sequences of events or trends.Scendriosdescribe interconnectionsperhaps even causal processesand highlight,where possible, decision points. In a sense, a scenario is a possible "slice offuture history."

Constructing Scenarios

1 In the present instance, since responses on global temperature are pivotal Insetting the stage for other climate variables, the plot in Figure I-8 carrbe usedas a basis for dividing the perceived temperature range into a number ofcategories. These categories then become the bases for constructingscenarios. The number of categories (and scenarios) is, in a sense, arbitraryand can be three or five or even a larger number, if desirable. Table 1-3 showsthe perceived temperature range divided into five categories. They' rangefrom large global cooling to large global warming. Associated with eachtemperature range is a probability of occurrence where, in fact, thetemperature ranges,were selected to make these probability ranges symmetri-cal.

Table 1-3

DEFINITION OF TEMPERATURE CATEGORIES

TemperatdreCategory

Change in Mean Northern HemisphereTemperature from Presene'by theYear 2000 Probability

Large cooling

Moderate cbolingSame as last 30 yearsModerate warm

Large zwarrni

"Piesent- ternthe average tern

0.3°C to 1. :older0.05°C to 0.3° colder

0.05 C,colder to 0.25°C warmer0.25°C to 0.6°C warmer0.6°C to 1.8°C warmer

0.100.25

X0.300_25 "

0.10

rature is defined as the end point on Mitchell's graph (Figure 1-2)erature for the five year period ending in 1969.

In order to pro=ss information with respect to other climate variables, it isuseful to group espondents with respect to thek five temperature ranges,according to here the bulk of each respondent's' probability densityfunFtion lies. Table 1-4 is a matrix showing each of the five temperaturecategories arrayed as rows and the 19 respondents in 5 groups arrayed as

columns of the matrix. As will be noted in Table 1-4, the bulk of eachgroup's probability density functions lies along the diagonal element of the'5x5 matrix (one-respondent at each end, three and four at the intermediateranges, aria ten in the middle range).

The results of the information collected under Task I have been embodied ina set of five scenarios desetibed in Chapter II, with more detailed discussion

-.and comparisons in Chapter III.*

-The scenarios are labeled in accordance with the global- temperaturecategories4n Table 1-3. One purpose is to provide an integrated summary ofperceptions of climatologists cm climate change and variability to the year2000. An equally important purpose is to provide a point of departure forstructuring questions in Task II and Jo trace the impact of such possibleclimatic changes on food production and on the choice of policy options.

METHODOLOGY

Table 1-4

PERc5NTA E,OF GROUPED PROBABILITY DENSITIES LYING INEACH TEMPERATURE CATEGORY

TemperatureCategories

Number o spondents1 4

Large cooling 99 12

Moderate cooling 1 68 10

Same as last 30 years 202 31Moderate warming 22 44 20Large warming 15 -80

The procedure for creating scenarios corresponding to the five globaltemperature categories is as follows:

Each respondent is first assigned to a global temperature category, asdescribed in Table 1-4.

Responses within each temperature category are combined for all otherclimatic variables (except for precipitation and precipitation variability,where all responses were available**).

Responses are integrated into a narrative, supporteciby summary tables.

he responses to Question VII, "Outlook for 1977 Crop Year,"are not included in the,scenarios. That portion of the informatigh for which expertise levels were consideredadequate has been processed and is shown in Appendix B. Included is an analysis ofsubjective probabilities concerning the persistence of droughts in the U.S.

"For questions on precipitation and precipitation variability only, information waselicited from each respondent based on conditional assumptions with respect to globaltemperature.

METHODOLOGY

The processi -steps for Questions II through X are identical to those forQuestion I except that, of course, in these other instances, density functionsor equivalents are provided directly by the respondents and need not bederived through the use of cumulative probability.

Figure I-10

QUESTION VI MID-LATITUDE DROUGHT Frequency of Drought

Timeperiod

."F re.quent"-i.e.,similar to early tomid-1930's and earlyto mid-1950's

"Average"-i.e.,similar to thefrequeocy overthe longest periodof record available

"Infrequent -

i.e., similar fo1940's and 1960's

TotalProbability

US

Other mid-latitudes US

Other mid-. latitudes US

Other mid-latitudes US

----

Other mid-latitudes

--7---- --1977to1980

1.0 1.0 .

1981

to1990

1.0 1.0

1991

to

20001.0 1.0

The sequence of steps is illustrated by using Question VI, which concernsmid-latitude drought. Figure 1-10 is an excerpt from Question VI. Table I-5illustrates how responses for one of the time periods (i.e., 1991 to the year2000) were weighted and aggregated in the Moderate Warming scenario. Theprocess outlined for Question VI is repeated for each of the other questions.

Table 1-5

FREQUENCY OF DROUGHT IN U.S. IN 1991-2000

Respondents Aydgnedto ryloderate Wdi UHF lt] Expertise Frequent Average Infrequent

A.. 3 0.25 0,60 0.25ft 3 0.60 0.20 0.20

5 0.60 020 0.20

Vkitflt][Iti!ti ,Ives dfr 0.25 0.2114

Review of Scenarios

In June 1977, the project Advisory Group recommended that an ad hocpanel review early drafts of the five scenarios for internal and mutualconsistency. Accordingly, project staff met in July with six climatologists atthe National Center for Atmospheric Research at Boulder, Colorado. Thereviewers paid particular attention to the large and moderate warming andcooling scenarios, i.e., those constructed from the smaller data bases. Thedetails and the 'conditional probabilities of. these end, scenarios, therefore,reflect the judgments of more people than the limited number of panelistswho responded to the questionnaires along the lines of these scenarios. Thereview process, which essentially strengthened the data bases of the endscenarios, resulted in significant changes to only one of them, the largeglobal cooling scenario.

Nature of Scenarios

Each scenario seeks to describe average climatic conditions as they mightexist in a period of years around A.D. 2000. The conditions do not referspecifically to the year A.D. 2000; the climate of that year is likely to differfrom the scenario projection to an extent consistent with normal year-1°-year climate variability. Some indication of the course of climate changesLatween the present time and the end of the century is also given in thenarrative, and in the tables appended to each scenario.

Each scenario is assigned a "probability of scenario." This "probability" is a

derived value based on the panelists' probabilistic temperature forecasts anda weighting scheme to take into account each respondent's expertise as ratedby himself and his peers. Therefeire, it reflects the range of judgmentsexpressed by the climate panel ancPthe strengths of their beliefs, as well astheir level of expertise.*

This probability should not be construed as the likelihood that the totalscenario will actually materialize in the future. The correct interpretation ofthe -probability of scenario" involves the following considerations:

'A "probabil of 025, for example, does not mean that there was universal agreementthat the seen. fn question would occur with probability 0.25. Nor does it mean that25% of the panelists "voted" for that particular temperature change to the exclusion ofother changes. Roughly speaking, the "probability" 025 is an amalgam of the proportionof panelists who gave some credence to that particular temperature change, the strengthof their individual "beliefs" in the change (then- individual probabilities of occurrence)and their individual expertise.

METHODOLOGY

15

METHOD OGY

(11-' The "probability" is essentially a measure of the confidence, expressedcollectively by the climate panel, that the global tdmperature changebetween circa 1970 and circa A.D. 2000 will lie in the range indicated by thescenario. This measure of confidence bears an unknown relationship to theprobability that the scenario will actually occur.

(2) It was assumed that the global temperature change indicated by thescenarios has a negligible probability of being greater than +1,8'C (the upperlimit of Large Warming) or less than 1.2'C (the lower limit :of LargeCooling). In this respect, the five scenarios, taken together, are considered tobracket all realistic outcomesi.e., the probabilities of the five scenarios sumto unity.

(3) Details are given. , in each scenario which elaborate on the scenario inrespects other than stipulated global temperature change= These areconsidered .'by the climate bane! to be reasonable inferences --about futureclimatic developments that are consistent with the global temperaturechange. These/ details by no means exude other possible developments,Hence, they are not necessarily to be construed, individuallyor incombination, as having a probability as high as that indicated for thescenario as a whole. Conditional probability information, given in the tablesincluded with each scenario, can be combined with the overall probability ofthe scenario to assess the absolute level of confidence to be placed in futureevepts specified in the scenarios. For example, one can find the overall"probability" of a specified event (e.g., "frequent" drought in the U.S.0 forthe period 199.1=2000) by first calculating for each scenario the product ofthe "probability" of the scenario and the conditional probabilityof theevent, for that scenario, and then summing the products for all five scenarios.

16

CHAPTER TWO

E CENARIOS

Larva Global exsoling

Moderate Glokal

Same at the Last 30 Years

Moderate Global Warming,.

Laibe Glottal Warming

.

4

W.

I

CHAPTER TWOCLIMATE SCENARIOS

This chapter contains descriptions of five climate scenarios, ranging fromlarge global cooling to4orge global warming.The last subsection of Chapter Idescribed the nature of these scenarios, including the correct interpretationof the "probability of scenario- and of the other probabilities associatedwith the scenarios.

In the text and tabled ofthis chapter:the latitudinal zones are as defined inthe climate questionnaire, Appendix A: -Qolar latitudes," 65 to 90';ThAher mid-latitudes,- 45' to 85'; "lower mid-latitudes,/' 3Q to 45'; and-subtropical latitudes," 10 to 30',

LARGE GLOBAL COOL,

The global cooling trend that began in the 1940's accelerated rapidly in thelast quarter of the 20th century, The average global temperature reached itslowest value of he past century a few years before the century ended, Bythe year 2000, the mean northern hemisphere temperature was about a6'Ccolder than in the early 1970's and climatic conditions showed a-strikingsimilarity to the period around 1820. Climatologists explained this largeglobal cooling in terms of natural climatic cycles, partly solar induced andpartly attributable to several major volcanic eruptions that occurred between1980 and 2000. Although most climatologists had expected 'a continuedincrease in carboij dioxide to be reflected in global warming, this warminginfluence was overwhelmed by the natural cooling in theperiod.

While temperature decreased over the entire globe; the largest decreasesoccurred in the higher latitudes of the north-ern hemisphere. The north polarlatitudes, marked by an expansion of arctic sea ice and snow cover

"Statements concerning some details of this scenario reflect a higher degree of certaintythan was expressed by the climatologists who participated in this study. See the attachedtables for thc4 range of unaertainty. See also the discussion in subparagraph (3) at the endof Chapter 17

CLIMATE SCENA I S

Table.11-11

0

LARGE GLOBAL COOLINGPROBABILITY OF SCENARIO: 0.10MEAN N?FiTHERN HEMISPHERE TEMPERATURE CHANGE SINCE 1969: between 03 and 1.2C colder

PROBABILITY OF TEMPERATURE CHANGE BY LATITUDE

(Compared with 1970=75).

Poldr

Higher mid latitude'Louver 1511(1 idhttirieSlibtropical

SoSo6picalLoorvi mid Idottirlr-,Hmher mid 1,150.1t1,'

Polar

U

c? ,

UMA

in. -,

U

--q---U0_ .J

_Lo

ULO .

rp .

UIn r,ci E6! r`T-,

0

C)

'z-L,

--,.___

N

U

Err) ,,

0.2 06 0.1 0.1

01 0.5 0.3 0,1

0.4 0.4 0.20,5 0.5

0.5 0,50,5 0.4 0.1

0.6 0.3 0.1

....'-- 06 03 0.1

Growiny sedsor, higher rThdr11r! lantudwi Protkriffitty dri incwow (decrease) in thr:length of the wowing seasonexceeding 10 days is 0.0 (0,9}; probability iii if1Cfea5e (decrease) ri the variability of the length of the growingseason in r-o('cess of 26. is 0.8

PROBABILITY OF PRECIPITATION CHANGE BY LATITUDE-

(Compared with 1941=70)ANNUAI GROWING SEASON

CV cp.

C

Cc,A)

Ui \

Higher rnT. lId itode 02 0 . 5 0.3 0.2 0.5 I 0.3Lower mid latatrZe- 03 05 0.2 0.3 0.5 0.2Subtropical 02 0.5 ,0.3 0* 0,4 0.4

PROBABILITY OF PRECIPITATION VARIABILITY'CHANGEIBY LATITUDE

(Compared with average for theprevious 25-year period

I-1101er lahhideI._CIVIder Mei l)tittieirStiStropic,r1

ANNUAL

,r, = 0 -1

GROWING SEASON

,==

G3 r,A ,t, c "- (N

-i

0:3 0 6 0:10.3 05 0.203 05 02 0.3 0.5 0.204 0 4 0 2 0.4 0.4 0.2

18

A CLIMATE SCENARIOS

Table IIB

LARGE GLOBAL COOLING

RELATIVE IMPOH ANCE OF CARBoN DIOXIDEAND TURBIDITY (PERCENT) DURING THEPERIOD 19/5 2000

PROBABILITY OF MIDI_ ATITIriF 1)0ou( HT'

Other Me..1-1_,th lrr

PROBABILI/ TY OF SAHFL DROUGHT'

PROBABILITY OF MONS_ N FAILUF E

IN01 Fe=^,e)t; t

I e

11f

Othe

'Fr(!r

.1 \/..e

req t

I ,tv

0(

0,6

05

0 4

0.5

0.5

15 05

Ut

20

=

50

0

.-,

10

05

04

0.1

0.4

0.4

0.4

1991-2000

LL

07

0.7

0.2

0.2

U.1

0.1

0:6 0:3 0.1

0.5

0.5

05

01

0.1

0.1

0.5

0.6

0.5

0.3

frud dverage? to the fret vet the lorylest1960, .

1940-50 oriel 1965 13 %Milk I Vt.! al.,: kill [WI Ifrf.q-p.ivrit hFe

C.) ifI,IV;I I-r1alr, ,ilift!(jUritt

Ilia'util,u hr 1!,(Ito.ric1

0.1

0,1

19

CLIMATE SCENARIOS

20

(especially in the north Atlantic sector), 'had cooled by about 2°C since theearly 1970's.** The northern higher and lower middle latitudes cooled byslightly more than 1°C. The subtropical latitudes in both hemispheresshowed a 0.5'f d r'rease ,in average temperature, while the remainder of thesouthern latitUdes showed a 1°C decrease. The large global cooling trend wasalso reflected in a significant decrease in the lengthof the growing season in,the higher middle latitudes and a substantial increase in the variability in thelength of the growing seasonjrom year to year. v\-

By the year 2000, it s also raining less in the higher middle andsubtropical latitudes, ough precipitation amounts in the lower middlelatitudes changed little or possibly increased slightly.

Precipitation also became more variable. The westerlies showed a pro-nounced shift from the higher middle to lower middle latitudes. This shiftbrought brief, yet severe, "hit- and -run" droughts as well as severe cold spells(including early and late killing frosts) in the lower middle latitudes. Thehigher middle latitudes, particularly Canada, from which the westerlies andtheir associated storm tracks were displaced, suffered an increased incidenceof long-term drought and winter cold. In the subtropical latitudes, thesubtrOpical highs tended to displace the tropical easterly rainbelt and, hence,increased the incidence of long periods of hot, dry weather. The center andintensity of the Asiatic monsoon changed *emetically between the late1970's and the turn of the century: The frequency of monsoon failure innorthwest India increased to such an extent that the last decade of the 20thcentury bore a resemblance tiD the period from 1900 to 1925. Droughts werealso more frequent in the Sahel region.

--e climatologist vv4)o inclined to this scenarioireasoned that the north polar regionswould GNI only about 0.5'C considerably less than the cooling in the middle northernlatitudes.

CLIMATE SCENARIOS

MODERATE GLOBAL coo:LING*

The global cooling trend that began in the 1940's continued through the lastquarter of the 20th century. By the year 2000, mean northern hemispheretemperature had decreased by approximately 0.15°C compared to the early1970's. Climatologists explained this trend principally in terms of a naturalcooling cycle, moderated by the warming effects of increasing amounts ofcarbon dioxide in the atmosphere. The cooling cycle was partly solar inorigin and partly associated with an increase in volcanic activity.

While temperature decreased over the entire globe, the largest temperaturedecreases occurred in the higher latitudes of the 'northern hemisphere.Specifically, the polar latitudes of the northern hemisphere cooled by 1°C;the higher middle latitude by 0.4°C; the lower middle latitudes by 0.3°C;and the subtropical latitudes by 02°C. The southern hemisphere, with itsmore zonal circulation and larger ocean area, cooled more uniformly andslowly; the average cooling in that hemisphere was about 015°C. The extentof the cooling in the higher middle latitudes was not sufficiently large tocause a significant change in the mean length or interannual vana ility of thegrowing season.

The growing-season precipitation as well as annual precipitation levelsremained unchanged in the lower middle latitudes but decreased slightly inthe higher middle and subtropical latitudes. Annual and growing=seasonprecipitation.-variability increased slightly cornoar4 to the 1950=75 period,with the stro4elt tendency toward increased variability in the subtropicallatitudes.

Drought conditions again plagued the mid-latitude areas of the UnitedStates, corroborating the 20-to-22-year drought cycle hypothesis. In theother mid latitude areas of the world, there were intermittent droughtconditions comparable, to those of the 1970's. Droughts were also morefrequent in the Sahel region, as was monsoon failure in Asia.

'Std T. r1(:ii!rmliiil nit detail t :ea) iutfhet cleTee of certainty

than iiy cleihitoliiiicie, Ifl tilL 'ittl 1 trM iitt.;.$(:)1t!(.1

tA)IPS,tOr c)f y. ,W.t) th fi Ill (3) ;lit thf'

of CH.11.111!I I, 21

CLIMATE SCENARIOS

Table II-2A

'MOBPR

ERATE GLOBAL COOLINGBILITY OF SCENARIO: 0,25

F N NORTHERN HEMISPHERE TEMPERATURE CHANGE SINCE 1969: between 0.05' and 0 colder

PROBABILITY OF TEMPERATURE CHANGE BY LATITUDE

uLS) C _.

)

o' ,_0 cu

c_

s

--,7 2 -0 . , .

.- U u ;?._,...'

Polar 01 0.6 0.1 0.1 01Northern 'Higher mid-latitude 0.1 0.4 0.4 0,1hemisphere Lower rnaldattrurie 0.1 0.1 0.7 0.1

Subtropical 0,2 0.7 0,1

Subtropical 0.2 0.6 0.1 0.1Southern Lower rnirld 'rude 0.1 0.2 0.3 0.3 0.1hemisphere Higher mit1-1 i 0.1 0,2 0.3 0.3 0.1

Polar Or/ 0,2 0.2 0.3 0.1

'Growing season n higher middle latitudes- Prexceeding 10 days i 0.1 (0.2); prohahihry of an itseason in excess of 25% is 0.b0.1).

(decrease) in the length of el .,/grow inc- season1r r r in he vdrionllity of the length of the lrowing

PROBABILITY OF PRECIPITATION CHANGE BY LATITUDE

nparecl with 1941=70)

Higher midatiturieLower mid-latitudeSubtropical

0202

2"

NNUAL-

GROWING SEASON

c .

A 0 A

0.5 0.3 0.2 0.5 CO0.6 0.2 0.2 0.6 0,2

.0.5 0.3 0.2 0.5 0.3

PROBABILITY OF PRECIPITATION VARIABILITY CHANGE BY LATITUDE

(Compared with average fur theprevious 25 =yrear period )

Highe r-raid -1.atiturle

Lower mirt-latiturleSubtropical

22

ANNUAL GROWING SEA-SON

-',

,,--

Ct,0 :

LOLT rN ry

0.2.

0.3f

0.60.13

0.5

0.2

0.20.2

0.3

0.30.4

0.60.60.5

0.1

0.1

0,1

Table II-28

CLIMATE SCENARIOS-,

MODERATE GLOBAL COOLING

RELATIVE IMPORTANCE OF CARBON DIOXIDEAND TURBIDITY (PERCENT) DURING THEPERIOD 1975-2000

20

1977 80 1981-90 1991-2000

E . .Y a n

a,- .-LL E

PR ABILITY OF MID ATITUDF DROUGHT'

i,

U W(J States 0,13 0.3 0.1 0.3 0.6 0,1 0.6 0.3 0,1At-

Other Mid-Latitude 0.0 0.4 0.5 0.4 0.1 0. 5 0.4

PROBABILITY OF SAHEL DROUGHT 0.4 0,5- 0.1 0.4 0.2 0.5 0,4 0.1

PROBABILITY OF DNSOON FAILURE-

Northwest India 0.4 0.1 0.3 0.5 0.2 0.4 0.6

r I 11,0 0.5 0.4 0 1 0.3 0.5 0.2 0.5 0.4 0.1

Other n n Asia 0.5 0:4 0.1 0,3 0,5 0,2 0.4 (15 0,1

`Frequent -,irnildr to (NO,/ to rred-1930.r. ,tntl gat I o need-1959, avun sirmlar to ttlfo fret-Went: v(- the longest1960,ctrlperit1 of record dViiilable: ill fret/own t

**Fret:view- simi)pr to 1940. 0 arrrf 1965-73 oenodrecord available, uzfreillient lat to 1954.65 10

dVal I dlitt.! mfrtrftttfr

irridar to the frequency over the Ionr- ad of

rmlar trr 1900,5 t > =rr trl ,IV(Thigt? woldr ti) the frpqm,r tty cp.! o )(I of r,cor(1-Hrliodr try 1930 601),!rwrI

23

SAME AS THE LAST 30 YEARS*IsThe global eoolirig trend that began in the 1940's,leveled Out in the 1970's.Average global temperature in the last quarter of the 20th century increasedslightly; thus, temperatures were more consistent with those in the periodfrom 1940 to 1970. By the year 2000, mean northern hemispheretemperature t d risen approximately 0.1,'C compared to the early 1970's.Climatologi s explained that the warming effects of the increasing amountsof carbon ioxide in the atmosphere had balanced a natural cooling cycle.Temperatu increases were nearly uniform throughout the north n andsouthern hemispheres, with slightly more warming in the northern -emi-sphere than in the southern. No significant changes in the mean length orinterannual variability of the growig season were noted.

The annual precipitation levels as well as the growing-season precipitationremained unchanged from the 1941-70 period. Also unchanged was thevariability Of annual precipitation. However, a small shift toward increasedvariability in th growing season was detected.

Drought conditio again plagued the mid-latitude areas of the UnitedStates, corroborating the 20-to=22-year drought cycle hypothesis. In othermid-latitude areas of the world, drought conditions recurred also, but not tothe same extent as in the United States. On the other hand, favorableclimatic conditions returned to India and other parts of Asia. 'Monsoonfailures became more infrequent. Also, the Sahel region, which had sufferedsevere drought from! 1965 to 1973, returned to average weather conditions.

'Statements concerning some details of his scenario reflect a higher degree of certaintythan was expressed by the climatologists whii participated in this study. See the attachedtables for the range of uncertainty. See also thtp discussion in subparagraph (3) at the endof Chapter I.

CLIMATE SCENARIOS

25

CLIMATE SCENARIOS

Table II:31

SAME AS THE LAST 30 YEARSPROBABIJL Y OF SCENARIO. 0.30 MEAN NORTHERN HEMISPHERE TEMPERATURE CHANGE

SINCE 1969: between 0.05 colder and 0,25C warmer

Y OF TEMPERATURE CHANGE BY LATITUDE

Pmldr

H;t1hnr [(In() i.itir.1(1,(`

tmv,((t ryi.0 1,1T1rJ(In

hi

hi r1fi ill t111103'

Pr

0! 0.1

0C33 3:t_I

CI7,/

.73/

Cr.!

Cl.?

III 5333

Ci 33:

0 -1

0.1 0.2 0E1

0.1 02 0.5

03 0.101 0.3 0.4

1 0:0 0 d0 1 0.1 011

(.31

LC,

0:1

010.1

0.1

0.1

0.1

0.1

nr, 5 Inor mr tfl Fur, i r Prni)111.11( (1' jr III{ III r!crri2 or (ht. it!thiti of tir0vvirir %ciriorlexc,,q1mq 10 cf,,y,, 0 2 10.11. prni)11.11 t, tit In( r I 13.i3,.,) in th. L313,..11)111ty of thr IrTuitti of the (jr °vying5e.Iison ext: of 25'. 0 1 (0 11


((Iompard with 1941-70)ANNUAI GROWING SEASON

A

ii

0 .2

0

0 2

(=)

0.6

(1.(I

0.2

0.2

0.2

c=a

0.2

.2

Fiqiree rel(I 1:il,tudemet,Lit,rede

.10

2-E.4

C-,

U

0.60.60.6

0,20.20.2


Worriffirrof wrfh ovo,Eq.! for Owmoviow, 25 yor rwimd) ANNLLAI GROWING SEASON

.33

cc,

L.3) LTI

111.15..r 0 2 0 21lftiJij 11111 110010 0 I( I, 0 0.2 0.7 U.M.1)11,i0J, .11 0 I, 0 0 .3 (.1.6 1

26

,33

CLIMATE SCENARIOS

Table II-3B

SAME AS THE LAST 30 YEARS

RELATIVE IMPORTANCE OF CARBON DIOXIDEAND TURBIDITY (PERCENT) DURING THEPERIOD 1975=2000

50 10 10

1977,430 1981 =90 1991 =2000

w

Z.supaa.,

,IJ

ccvma

ceCO

I?

t,ctv

c

PROBABILITY OF MID-LATITUDE DROUGHT"

United States 0.5 04 0.1 0.2 0.6 0.2 0.5 0.4 0.1/Other Mid-Latitude 0.4 0.5 0.1 0.3 0.6 0.1 0.4 0.5 0.1

PROBABILITY OF SAHEL DROUGHT' 0.2 0.6 0.2 0.2 0.7 0.1 0.2 0.7 0.1

PROBABILITY OE MONSOON FAILURE"'

Northwest India 0.3 0.6 0.1 =0.2 0.6 0.2 0.2 0.5 0.3

Other India 0.3 0.6 0,1 0.2 0.6 0.2 0.2 0.5 0.3

Other Monsoon Apia 0.3 0.6 0.1 -0.2 0.6 0.2 0,2 0.6 0.2

nt-similar to early to mid-1930s and early to mid-1950s. -v g- -similar to the frequency over the longestperiod of record available; infrequent similar to 1940s and 1960s.

'Frequent -sirnIIAr to 1940.50 and 1965-73 penoth average- similar to the frequency over the longest perirecord available; infrequent similar to 1950.65 period.

'Frew -similar to 1900-25 Period r trr Ow frequency over the longest period of recordavailable; infrequent- similar 1930 60 period

27

CLIMATE SCENARIOS

isz

MODERATE GLOBAL WARMING*

The global cooling trend that began in the 1940's was reversed in the lastquarter of the 20th century. By the year 2000, mean northern hemispheretemperature had risen by approximately 0.4°C, compared to the early1970's. Climatologists explained that this increase in temperatures was dueprincipally to the warming effects of increasing amounts of carbon dioxidein the atmosphere, which predominated over a slow, natural cooling effect.

While average global temperature increased moderately, the largest tempera-ture increases came in the higher latitudes. The northern hemisphere warmed-slightly more than the southern hemisphere due to its greater land area andthe larger thermal inertia of the southern oceans. In the northernhemisphere, the polar latitudes warmed by 1.2°C; the higher middle latitudesby 0.5°C, the lower middle latitudes by 0.3°C; and the subtropical latitudesby 0.25°C. In the southern hemisphere, average temperatures over the polarlatitudes increased by 0.65°C; the higher middle latitudes by 0.4°C; thelower middle latitudes by 0.3°C; and the subtropical latitudes-by 0.2°C. Theincrease in global temperature was reflected in a modeller increase in thelength of the growing season in higher middle latitudes, but no significantchange in the interannual variability of the growing-season was noted.

Annual precipitation levels increased slightly in the higher middle latitudesbut showed little change for lower latitudinal bands. Growingseasonprecipitation also increased slightly in the higher middle latitudes andsubtropical regions but remained unchanged in the lower middle latitudes.Both annual and growing-season precipitation variability remained essentiallyunchanged except for a slight increase in the variability of growing-seasonprecipitation in s 'bbtropical latitudes.

Drought conditions again plagued the mid-lap ude areas of the UnitedStates, corroborating the 20-to-22-year drought cycle hypothesis. Climaticconditions were somewhat- more favorable in tbe Asiatic region and insubtropical North Africa. The frequency of monsoon failure, especially innorthwest India, resembled more closely the long-term average; so did thefrequency of drought in the Sahel region.

'Statements concerning some details of this scenario reflect a higher degree of certaintythan was expressed by the climatologists who participated in this study. See the attachedtables for the range of uncertainty. See also the discussion in subparagraph (3) at the endof Chapter I. 29

CLIMATE SCENARIOS

Table 'II-4A

MODERATE GLOBAL WARMINGPROBABILITY OF SCENARIO: 0.25MEAN NORTHERN HEMISPHERE TEMPERATURE CHANGE SINCE 1969: between 0.25° and 0.6°C warmer

'PROBABILITY OF TEMPERATURE CHANGE BY LATITUDE

(Compared with 1970-75)

.-- gJ-0 L

tr,E

0,E

L1 u ,E

--, ti

go

ei ?'=

P l'

POW 0.1 0.1 0.2 0.2 0.2 0.2Northern Higher thid-latitude' 0.1 0.3 0.4 0.1 0.1hemisphere Lower mid-latitude 0.1 0.5 0.3 0-1

Subtropical 0-1 0.6 0.2 0.1

Subtropical 0.1 0.6 0.2 0-1Southern Lower mid-latitude - 0.1 0.5 0.3 0-1hemisphere Higher mid-latitude' 0.1 0.3 0.5 0.1

Polar 0.1 0.2 0.5 0.1 0,1

'Growing season in higher middle latitudes: Probability of an Increase (decrease) in the length of the growing seasonexceeding 10 days is 0.4 l0.21; probability of an increase (decrease) in the variability of the length of the growingseason in excess of 25% is 0.1 (0.21.


(Compared with 1941-70)ANNUAL GROWING SEASON

60

EC

0u

8 A

Higher mid - latitude 03 0.5 0:2 0.3 0.5 0.2Lower mid-latitude 0.2 0.6 0,2 0.2 0.6 0.2subtropical 0.2 0.6 0:2 0.3 0.5 0.2


(Compared with average forprevious 25-year period)

Higher mid IititutfLower mid lat,te,t1Subtropical

30

leANNUAL GROWING SEASON-

. ,-,

c .

Li ./\

0.2 0.6 0.2 0.2 0.6 0.20 2 0:6 0 2 0.2 0.6 0,20,2 0 6 0 2 0,3 0.5 0.2

CLIMATE SCENARIOS

Table 114B

MODERATE GLOBA-L WARMIN,G

;El 2C.) ID

0_F.

cr,

0.Y

0: a

RELATIVE IMPORTANCE OF CARBON DIOXIDEAND TURBIDIYY (PERCENT) DURING THEPERIOD 1975-2000

6 15 5 lir 10 10

1977-80

z

LL IL

PROBABILITY OF MID-LATITUDE DROUGHT'

United States

Other Mid-Latitude

0.6 0.3 0.1 0.2

PROBABILITY OF SAHEL DROUGHT" 0.3 0.4 0.3 0.3

PROBABILITY OF MONSOON FAILURE"

Northwest India

Other India

Other Monsoon Asia

0.4 0.3 0.3

19B 1991-2000

E

n'm

CTtb

LL

_

..,n...CT(I)

...-

0.2 0.6 0.5 0.3 0.2

0.4 0.3 0.3 0.4 0 =3

0.4 0.3 0.2 0.5 0.3

*FreqUerit -similar to early tc mid-1930s and early to MT-1'1950S, average similar to the frequency overlongest period of record available, infrequent -511-1111dr to 1940s and 1960s.

he

* *Fret:men =timilar to 1940-60 and 1966-73 periods; agesimilar to the frequency over the longest period ofrecord available; infrequent to 1950-65 period.

* *Frequent -similar to 1900 25 period: v edge similar to the frequency vet the longest period of recordavailable infrequent to 1930 60 period.

LARGE GLOBAL WARMING*

The global cooling trend that began in the 1.940's was dramatically reversedin the last quarter bf the 20th century: By the year 2000, the mean northernhemisphere temperature had increased by about 1°C compared to the early1970's. Climatologists explained that this_ trend was'due principally to thewarming effects of the increasing amounts of carbon dioxide in theatmosphere:

. .

While temperature increased -over the.entire globe, temperature increaseswere more pronou,nced at higher latitudest: The subtropical latitudes warmed,on the average, by 0.8°C; the lower .middle latitudes by 1.0°C; the highermiddle latitudes by 1.4°C; and the polar latitudes by a remarkable 3.0°C,compared to the early 1970's. ymmetry prevailed as similar temperaturechanges were observed in both

7,1

oth t e northern and southern hemispheres. Theincrease in temperature was acc Mpanied by a significant increase in thelength of the growing season in ,he higher middle latitudes, as well as by asubstantial decrease in the variability from year to year in the length of thegrowing season.

Precipitation levels generally increased, especially in the subtropical andhigher middle latitudes. In the °lower middle latitudes there was little netchange of precipitation. Annual precipitation variability decreased slightlycompared to the 1950-75 period; precipitation variability during the growingseason similarly decreased in the higher middle latitudes, but increasedslightly in the lower middle and subtropical latitudes.

The warming trend also ushered in more favorable climatic conditions inIndia and other parts of Asia. These conditions were similar to those of the1930-60 period. Monsoon failure was infrequent, especially in northwestIndia. But in the mid-latitude areas of the United States, extending from theRockies to the Appalachians, drought conditions similar to the mid-1930'sand the early-to-thid-1950's prevailed. In other mid-latitude areas of theworld, notably Europe, the probability of drought declined. The increasedlevels of precipitation also returned the Sahel region to wetter weatherconditions.

"Statements concerning some details of this scenario reflect a higher degree of certaintythan was expressed by the climatologists who participated in this study. See the attachedtables for the range of uncertainty. See also the discussion in subparagraph (3) at the endof Chapter I.

CLIMATE SCENARIOS

33

,

CLIMATE SCEN eS

Table

tARGLG.OBAL WARM NGF7:172WCENARIO: 0.10MEAN THERN HEMISPHERE TEMPERATURE CHANGE SINCE 1969: between (1.6 and 1,8°C warmer

PROBABILITY OF TEMPERATURE CHANGE BY LATITUDE I(Compared with 1970-75)

UIP

o 7-: 8

ULP0 I,6Do

ULc)DE66

u)

'-Eul6 3

u)

E6 -.- 3

U

ggi

3

U

m

,4

0t-

C9

Northernhemisphere

Polar

Higher mid -latitudeLower Mid-latitudeSubtropical

a

.

0.10.1

0.1.

0.50.8

0.1.-0.50.70.1

0.10.4

0.2

0.2 0.6

xx

Southernhemisphere

.SerbtroPicalLower mid-IritirdeHigher mid-latitudePolar sx

0.1

0.1

0.80.50.1

0.1

0.1

0.20,50 1

0.20.4

0.1

,

0.2 (10.5

*Growing season in higher..ig..er middle latitudes.. P obabeity of an increase (decrease in the length of growing seasonexceeding 10 days is 0.8 (0.0): probability of an mei-ease (decrease) in the variability of the length of the growing

- season in excess of 25% is 0.0 (0.7),


(Compared with 1941-70)ANNUAL GROWING SEASON

1, 0-U

exn a-95_..

c ,

orC 89-0

0 t's

Higher mid-latitudeLower mid -latitudeSubtropical

040303

050.505

010202

0,30.30.4

0.50.40.5 :

0.20.301


(Compared with average for theprevious 25-year pdiod )

Higher mid latitudeLower mid latitialeSubtropical

34

ANNUAL

020 2

02

Lii

(ROWING SEASON

0 5 0305 0,305 0

LiiLN

C

"Cu _

Ciu

0.2 0.5 0.30.3 0.5 0.20,3 0.5 0.2

CLIME SCENARIOS_

Table 115B -

LOGE GLOBAL WARMING

RELATIVE IMPORTANCE OF-CARBON DIOXIDEAND TURBIDITY tPERCENT) DURING THEPERIOD 1975-2000.

1977-80 198190 199 2000

'Emcr

LtLt

.cr

E

aE

L i_

C.

PROBABILITY OF MID-LATITUDE DROUGHT

United States

Other Mid-Latitude

0.6

0.5

0

0.3

0.1

01

0.6

0.5

0.3

0.3

0.1

0.2

0.7

0.3

0.2

0.3

0_1

0.4

PROBABILITY.OF SAHEL DROUGHT ", 0,1-= 0.8 0.1 0.1 0.7 0.2 0.1 0.6 0.3

PROBABILITY OF MONSOON PURE...-----.,

Northwest India

Other India

Other Monsoon Asia

0 1

0,1

0.1

0.8

0.8

0.8

0.1

0.1

0.1

O.1

0 1

0.1

0.6

076

0.6

0.3

0.3

0.3

0.1

0.1

0.2

0.2

0.2

0.8

0.7

0.7

Frequent similar to early to mid-1930s and early to mid- 1950s, average- similar to the frequency over the longestperiod of record available; infrequent similar to 1940s and 1960s.

'Frequent-similar to 1940=50 and 1965 73 periods average= to the frequency over the longest period ofrecord available. infrequent- similar to 1950-65 period

=frequent-similar to 1000,25 period, average- similar to the frequency over the longest period of record available;infrequent -similar to 1930,60 period.

35

.141116,thiriSsOrk.t

andU.S. and Other Mid:Latitude 6roughtSahelian Drought and Failure of Asian Monsoons ,

Conclusions on Climatic Variability

CHAPTER THREEDISCUSSION OF SCENARIOSAND CLIMATIC PROBABILITIES

GcNERAL

. This chapter summarizes the aggregated probabilistic data of the climatepanelists-and makes comparisons of these data7from scenario to scenario,across latitudinal zones, and ajoss three time periods between_pow and theyear 2900. The data received in respiinse to the climate questiOnnaire(Appendix A) are related primarily to trends in gross climatic parametersrather than the interannup variability of the parameters. In the ensuingdiscussion, particular attern is paid to the limited data that bear on thepaneli ts' perceptions about the important question of future climaticvaria ility. Also included are some of the climatologists' numerous corn-me s giving rationale for their answers and eNpressipg caveats andreservations regarding their responses.

GLOBAL TEMPERATURE CHANGES

The climate scenarios of the preceding chapter are structured around theresponses to the first questionGlobal Temperature of the climate ques-tionnaire (Appendix A). Table 1-3 and Figure 1-8 of Chapter I show the"globW- temperature range (expressed as the change in mean NorthernHemisphere temperature from present by the year 2000) for each of the fivescenarios and the "probability" associated with each scenario. The total spanof temperature changes in all the,,yenarios is 3'C, ranging from 1.2' coolerto 1.8' warmer. The three middle scenarios encompass a range of less than1"C (from 0.3' cooler to 0.6' warmer), and have an aggregate "probability'of 0.8. This general consensusthat there will be no radical change in globaltemperature by 2000 A.Diand the slight group bias toward global warmingare considered to be major findings of this study.

Respondents whose probability estimates tended toward the two warmingscenarios explained their reasoning primarily_ in terms of the likely long-termdominance of the CO, warming effect. This explanation is reflectedquantitatively in Figure III-1, which summarizes the responses to QuestionIll (Carbon Dioxide, Turbidity, and Climate) of the questionnaire. Ingeneral, panelists who inclined toward global cooling hypothesized that the 37

SCAROS - PROBABILITIES

-figure IVA

38

RELATIVE CLIMATIC INF UE

CARBONDIOXIDE

VOLCANICDUST

SMOKE

50

0%

FLUORO-CARBONSartdOTHERGASES

50%

0%

50

OTHERPARTICLES(AEROSOLS)

0 I

These five graphs summarize the aggregrated responses to _ion ill (Carbon Dioxide,Turbidity, and Climate) of the climate questionnaire The climatologists were asked toindicate the relative influence (in percent) of the indicated atmospheric components onglobal climate over the next 25 years. The vertical bars reading from left (Large Cooling)to right (Large Warming) correspond to the five climate scenarios.

SCENARIOS & PROBABILITIES

wanting effects of CO2 might' not materialize to the extent suggested bythose supporting a strong watming trend, or that the CO2 warming effectswould be overshadowed by a long-term, solar-induced cooling trend. Severalrespondents also commented on the possible cooling effects of volcarticactivity, noting, however, the difficulty of predicting the level or timing ofsuch activity. The middle scenariothat of little change in the mean annualglobal temperatureis predicated primarily on the warming effects of CO2balancing the effects of natural cooling_ Some panelists commented on thepossible effects on climatic change of dust and other natural andanthropogenic particles, but there does not appear to be a consensus whethersuch particles have a net warming or cooling effect; also, their effectprobably tends to be more regional than global.

LATITUDINAL TEMPERATURE CHANGES

Perhaps more pertinent than a given change in average global temperature isthe nonunifolmily of the associated temperature changes at differentlatitudes. An analysis of the respondents' estimates of likely temperature

itlchanges by itudinal zones (Question II of Appendix A) indicates thatglobal temperature fluctuations are expected to be far more pronounced inthe polar regions than in low latitudes. In other weds, the poles areperceived as more sensitive to climate change. A number of the respondentsjudged that the poles may experience a change at least several fold largerthan the global average (see Figure 111-2). For instance, in the Large GlobalCooling Scenario (0.3° to 1.2°C cooler), the probability for a 1.0' to 3.0°Ccooling in the northern polar latitudes is 0.9, and 0.6 for a 1.0° to 2.0°Ccooling in the northern higher middle latitudes. By contrast, northernsubtropical temperatures are perceived to drop by only 0.0w to 1.0°C withprobability 1. Similar observations about differential cooling in the higherlatitudes hold for the Moderate Cooling Scenario (0.05° to 0.3 °C cooler).The question arises as to what extent such an increase in latitudinaltemperature gradients could be a mechanism for greater year-to-year climaticvariability. The views of several of the panelists are exemplified by onecomment to the effect that while "a temperature change per se does notimply increased climatic variability, there is some physical basis for sayingthat a general cooling would imply increased baroclinic instability andtherefore increased temperature variability and vice versa."

The picture of increases in latitudinal temperature gradients that can beidentified in the two cooling scenarios does not hold in the two warmingscenarios. In the Moderate Global Warming Scenario (0.25° to 0.6°Cwarmer), the probability is 0.6 for a 1.0' to 3.0 °C warming in the northernpolar latitudes as compared to a probability of 0.6 for a 0.0° to 0.5°Cincrease in the subtropical latitudes. In this scenario, as well as in the lesslikely Large Global Warming Scenario, the perceived differential warming 39

-SCENARIOS & PROBABILITIES

40

Figure III-2

.PROBABILITY OF TEMPERATURE CHANGE IN NORTHERN HEMISPHERE

POLARLATITUDES

1.0

.8

.2

'COOLING iglfw- WARMING

HIGHERMIDDLELATITUDES

1.0

.8

.6

.4

.2

COOLING -Atti WARMING

LOWERMIDDLELATITUDES

1.0

.2

COOLING -At =iiirr- WARMING

1.0

.8

SUB- 6TROPICALLATITUDES

.2

COOLING -At WARMING

DEGREESCELSIUS

1_1_ L _1_ 1 1 1 1 11 1 1 1 1 1 I _I_ _I i t

2.0-3.0 C. 1.0-2.0 C 0-1.0 C 0-1.0 C 1.0=2.0 C 2.0-3.0 C 3.0=5.0 CCOOLING WARMING

These four graphs'summarize the aggregrated re; II of the climate questionnaire.Depicted for therndicated zones of latitude are the climatologists' probabilities of change in annual temperature by the year 2000,compared to 1910-15. For a given zone of latitude and band of temperature, the vertical hays correspond to the fiveclimate scenarios as in Figure III-1. If the probability of a temperature change in a scenario is zero, the bar for thatscenario is absent.

SCENARIOS 81-PR

reduces 'latitudinal tempeature gradients. Some panelists associated a

reduction of these gradients with a decrease in climatic variability. Onerespondent, quoted later at 'great length, commented that "it seemsintuitively reasonable" that re ced latitudinal temperature gratlients couldtcause "less variability due -to roclinic instabilities and blocking pressurepatterns." As will be seen later, thestRtences about climatic variabilitylessvariability in -a warming regime and more variability in coolingaresupported by the climatologists' responses to other questions.

Figure III-3 provides a more aggregated view of temperature changes bylatitude. Here, the partitioning of the climatologists into the five scenarioshas been eliminated. In each latitudinal zone, the frequency distribution oftemperature changes is attributable to the Climate Panel as a whole since theprobability of,a given temperAure change was calculated by multiplying the;probability of that temperature change in each scenario by the overallprobability of the scenario and summing the resulting products over the five'scenarios. The flat and wide frequency: distributions for the polar regions,especially the Northern Hemisphere, reflect the wide range of opinion andthe high degree of uncertainty relative to the warming or cooling issue. Asone moves from polar to subtropical latitudes, the probability density plotsshow a progressive decrease in temperature range and a correspondingpeaking in the temperature intervals of liklle or no change. In the subtropics,the respondents' estimates in aggregate indicate a 0.6 to 0.7 probability thatthe temperature change will be less than 0.5 °C warmer or cooler than atpresent; in the lower middle latitudes that probability is only slightlyless-0.5 to 0.6. In the higher middle latitudes, estimates of temperaturechange fell within ± 2.0°C, with a probability of about 0.45 for a change ofless than 0.5°C warmer or cooler and about 0.75 that the change will be lessthan ± 1.0° C.

1Several respondents commented that temperature cliges are likely to besomewhat less in the Southern Hemisphere than in'the-gorthern Hemispherebecause of the thermal inertia provided by the proportionally greater oceansurface. Note that in Figure 111-3 the temperature range' in the polar latitudesis somewhat less for th61,Southern Hemisphere than for the Northern. Alsonote that, as in Figare 1-8 of Chapter 1, the graphs for all the latitudinalbands in both hemispheres show a slight skewness toward warming. _

I LITI gs

41


42

. Figure 111,3

PROBABILITY OF TEMPERATURE CHANtE (Acmwcoted ll_:ross Scenaj los)

0.4

0 3POLARLATIT

NORTHERN SOUTHERNHEMISPHERE

WARMER

0.4

HIGHER 01MIDDLELATITUDES

02

0.1

COL DER

HEMISPHERE

LDER WARMER

WARMER COLDER WARMER

0.4

LOWER 0

MIDDLELATITUDES

0.2

0.1

0

0.4

SUB__ 03

TROPICALLATITUDES

02

DEGREE:CI#. .

/°_, .0 s= D. L-CL LC) CDI'L LL-L C_L D D. CD ..- CD ID LL-L C.) Ln Ln CD

tn L'I) LI'LLA '7 D-,,-L, Ln r--) c-,r r_-_-) c-2 7

o o. ,r, o ti-i o cir Ln c:r c 1 c:c ci e-r c-71 al CD DI roc-c-i -- rn In c---.1 , Cl r--3 r:-, c-)

Ler

1-Lie,e qtd111-- ci ,1 I ml 6ii1 111 -.,' 1 .1

lel,iy of ,1 reiiireq, if il 1111,' Ili fl 1,1LILl'IdLI olL 111.1- ILI 1

1 LL ci,uit fc, oil itidenilti.11 ..1.,or:Litiiiii ,.-,./Irli mile ot the live rilrri 1

fIL A r « 11« liii mrm ()IAItlf Ali111,111Alih)(11t` A t's.AIDIFA

If


-GROWING SEASON

Question X of Appendix A dealt with temperature-related parameters:changes in the mean length and variability of growing seasons in the highermiddle latitudes during the next 25 years as compared to the present. It wasone of the two questions that dealt directly with the subject of climatevariability, in this case interannual variability in the length of the growingseason. Numerical data from 15 panelists were processed for Question X.

As can be seen from Figure III-4, only in the Large Cooling and LargeWarming Scenarios were there high probabilities (0.9 and 0.8, respectively)for a "significant" change-10 days or morein the length of the presentgrowing season. Moreover, it was only in these two scenarios that thepanelists perceived a large probability for a "significant" change (25 percentor more) in the standard deviation of the length of the growing season: 0.8for an increase in variability under the Large Cooling Scenario and 0.7 for adecrease under the Large Warming Scenario. The aforementioned probabili=ties are ascribable to two climatologists who inclined to the extreme globaltemperature scenarios.

With regard to the middle three scenarios, a majority of the panelists tendedtoward high probabilities (0.7 to 0.8) for no significant changes in the meanlength of the growing season and its interannual

The following comments on the length and variability of growing 3sons

were made by two panelists disposed toward global cooling:

The changes in the variability or standard deviations Of temperature are avery clear function of both the double and the secular solar-climaticcycles. The peak of temperature variability was reached in 816 (thefamous year without a summer, that blew both hot and very cold), t'-ia180 year counterpart of 1996, the low point on my curve in Question 1.

I think it is more likely that the present meridional trend in atmosphericcirculation :(with longer growing seasons) will prevail for a few years, tobe replaced later on by a return to zonal circulation, with shorter growingseasons. Over the 25 years the change might well even out.

Another adherent of global cooling based his response on his curve forQuestion I and the "'corresponding :nectations of change of frequency ofblocking and the high variability from year to ycair which goes with it.

Two panelists wcommented that:

lt;c1 toward the glohat temperature :scenario

the gro.mg season raiidl iiicr nor lh irf 4(1 1, piobahly 43


Flgore- III-4

CHANGES IN HIGHER MID=LATITUDE GROWING SEASONS BY THE YEAR 2000

PROBABILITY OF CHANGE IN LENGTH OF GROWING SEASON

A INCREASE10 DAYS

1.0

4

2

CHANGE10 DAYS

DLGR EASE10 DAYS

COMPOSITEINCREASE /DECREASE

0

1 <C1-

LC MC S MW L

PROBABILI FY OF CHANGE IN STANDARD DEV AT N OF LENGTH OF GROW NG SEASON

Theso graphs !--itimenariie the awlr owaterl ne--,pon,:iin to OuestIo X (Length of firowmg St: i . The top row of graphspertain$ to p.!rci2r)tions rut changes by the ,ear 2000 in the lerl k ow .5ea..5orb the bottom row to changesitby 2000 A.r . In The standard deviation ot the loncithi of tho growing seasoli h changes being relative to the present),The thr,T,=,f-rtfltls of -F-,ifirlificatif" c hantiles art. 10 (Tlyfi for lengtres of cliovving st:a1;Orrs and 25 percent for the standardil,vidtioic, In 0001 graph 01, vF!I bc,11 hob, eulifi,i from 1,0 (Tdrqc 1,0tflimi) to right ( Lafcr VVarbninti) c:00-espuritl tothe flyo cliriTtro ,,,a:orTif lib, Flgurp, III ,IA pr, ebb, rho cliffijt()Icmvitli 0c0rIltrffi;rttr(iihi0.111 rlie tor III of corwerifit)001 Curtfirot)hs, In Firlurrr III 4/3 the :.tart' dot,' ,rue con(leie,!il into ,,ingle iir dpir,,, OW tr,rrTIFI! tit Wrilf:01 IS LISP(1 in sfilisetitiumtfitifire!,. Fur 0 cilyt.rti !--,C0,1)0r1(0 irr FirJr0r, III 4B, the proimbillty of 0 SlcjIlltir0;11 rirr:r `.05(r iri 0 p0r0111t0tUr is lepreSplitt0d bythe rwight of CH: 10,01+"1 t);11 rilt0ir=00rircl by ieddulti rip m011 tho left 11,10{1,J:0k., Tho Ifrobdblity (0 ,i ,,i(lificdrif fIcUrihit! ISfnrr0,-Mt,(1 fry U IiortIlit of the. llItili,r I011 1 ulltrrI 0AVII WI fry right -hand scale; The li ight of tho blank spoco boff:yeerithe ,ipl)er drici lo,yi!r lust , w01,-i!bry. [11,.! p101),II)Th./ of rui ,,i(plificdrif LII,iiimw in 614 oardiriettl.

44


little in 30' =40 N. Anomalies of sprit ] and fall season are little known,but can be significant.

Not knowing what is meant by "present,- the questiorlisvague. Referringto Question I, it would 4ape,..t- that 25 years will encompass variations inboth directions.

Quoted below are two adherents of global warming:

If one plots growing season vs. mean summertime temperature atmid-continental stations (both of which vary with latitude), one candeduce a rule of thumb that a change of t corresponds to about -t 10days in growing season. In 25 years the mean hemispheric temperatureshould rise about l'C, and at middle and higher latitudes the correspond=ing change should be several times larger. Thus, I would foresee a greaterthan 10-day increasein growing season at middle and high latitudes.

Consistent with a general warming to be anticipated with the CO,increase, which would be rargest in the high latitudes, I anticipate a higher-probability of a lengthened grow/it-1g season than that inferred fromclimatological averages. I doubt that thd'interannual variability o thegrowing season length would change by more than 25 percent but suchchanges as there may be are slightly more favored to be in the direction of

decrease than otherwise because the CO z increase itself might beexpected to discourage very low temperatures a bit more than now

Two other panelists stressed the difficulty of making any predictions aboutgrowing seasons:

Climate dues not necessarily change in the same sense all along a parallelof latitude. This, added to the fact that we have no way of telling how itwill change locally, makes it impossible to answer this question.

In my view there is no t,tistically significant evidence to support asystematic change of the growing season. There willObe fluctuations,particularly on the short side associated with volcanoes, but they cannotbe forecast.

Notwithstanding the par -lists' collective uncertainty in Question X about,the growing-season manifestation of temperature variability, one can infer atendency to associate greater variability of temperature with extreme globalcooling and less variability with extreme warming. A similar inclination toassociate more variability with cooling less with warming emerges in thefollowing discussion on precipitation.

45


-Wm

46

PRECIPITATION

Questions IV and V of Appendix A dealt respectively with Changes in thevolume and variability of precipitation for three bands of latitude. (QuestionV and Question X, on growing seasons, ere the only Ones t directlyaddressed climatic variability.) The two questions on precipitation wereunique in that they were couched in terms of conditional probabilities:panelists were asked for probabilities of change under the assumption ofthree given ranges of temperature changes over the next 25 years.

The answers to both questions on precipitation have been summarized inFigure III-5 to facilitate comparisons of the responses on precipitation,volume and precipitation variability. The respondents' estimates of theprobabilities of change in precipitation volume and variability inclioa e a highlevel of uncertainty not only about the amount of the change but n manycases even about the direction of the change; this is particularly tale withrespect to possible changes in interannual variability. Keeping this uncer-tainty in mind, the following cross=scenario and cross latitude highlights canbe identified.

Changes in Precipitation Volume

In Question IV the panelists were asked to provide probabilities of changesby the year 2000 in mean annual and growing-season precipitation relative tothe "normal" period of 1941-70. The thresholds for "significant" increasesand decreases in precipitation volume were defined as ± 10 percent.

For each combination of global temperature scenario and zone of latitude,the highest probability was for no major change in annual precipitation (i.e.,less than 10 percent change). The highest "no change" probabilities (0.6 to0.7) were found in the "Same" Scenario; in other words,those who opted forno major temperature changes were, expectedly, the most confident that nomajor precipitation changes would occur. But even in the extreme warmingand cooling scenarios, the aggregated responses indicated a probability ofabout 0.5 that the precipitation change would be less than 10 percent.

The aggregate responses of the panelists suggest that the highest probability(0.3 to 0.4) for an increase in annual precipitation would occur under awarming scenario, especially in the higher middle latitude and subtropicalzones. Conversely, a decrease in precipitation was associated with the coolingscenarios. A number of respondents commented that this pattern seemsreasonable, based on analogs of previous warm periods, as well as

precipitation results from limited numerical experimentation with generalatmospheric circulation models. But many panelists also noted the limitedscientific basis for assessing the probabilities of precipitation changes under

Figure III 56


CHANGES IN PRECIPITATION AND PRECIPITATION VARIABILITY

PROBABILITY OF CHANGEIN VOLUME OFPRECIPITATION(10% CRITERION)

-I

PROBABILITY OF CHANGEIN STANDARD DEVIATIONOF PRECIPITATION)2);') CRITERION)

GROWING SEASON ANNUAL

HIGHERM DLELA ITUDES

LOWERMIDDLELATITUDES

0

2

3

.4

.5

.4

3

2

0

0-.1

2

.3

4

.5

.4

.3

2

0

0

1

3

4SUBTROPICALLATITUDES

.4

3

1

ANNUAL GROWING SEASON

0

CC

U

LU

LU

Ci

The group of graphs on the It pertams to perceptions of chanties by the year 2000 in mean nualind mean rir t.

season prectpdation, for three tone,-; of latmid, relative to the 1941-.10 "normal" pattern nues\horrIV). The group onthet net pertains to perceptionc of changes over the next 25 year; In the standard deviations of the volume of i',Inntialand growmg-season precontahon, relative to the previous 25-year p000d (Uuestion V). The orobabilitir)s. of "salmi-tram"merease_c or decreases of the two ptimmeter,, (,it lortt 10 tot volume of precipitation and 25=',., for the standard devia-tion) and prob,Milltiwt of no iltjnifiCdr)t 01,111(it! thin 10 rnrd 25',, re.tpectively) are displtiyed in the format(les/Jibed for Fawn) III IF 47


48

alternative temperature scenarios. One panelist stated: "A careful answer tothis question would require a detailed statistical study. Such a study couldand should be done. To my knowledge it has not been done.-

To summarize, amid the indications of collective uncertainty about thevolume of annual precipitation, there was some tendency to associatesignificantly increased precipitation with the two warming scenarios andsignificantly decreased precipitation with the two cooling scenarios,except in the lower mid-latitudes, which exhibit a different pattern. In thelatter zone, the probabilities of significant decreases were about 0.2 in eachscenario. (In the other two bands of latitude, these probabilities show aregular diminution as one goes from large cooling to large warming.) Also, inthe lower mid-latitudes there were somewhat higher probabilities ofsignificant increases for the two cooling scenarios than in the other twozone of latitude.

The results on growing-season precipitation changes were similar to those forannual precipitation except in the lower mid-latitudes again. For theselatitudes, the probabilities of significantly decreased precipitation are lowbut increase monotonically in going from large cooling to large warming,reaching 0.3 in the latter scenario. This unexplained behavior is the reverseof that for annual and growing-season precipitation in the other two zones oflatitude.

For all five scenarios, the pattern of annual and growing-season precipitationprobabilities is markedly similar in the subtropical and higher middlelatitudes. If the departure of the lower mid-latitudes from this pattern is real,thatNsta\, if it reflects a tendency of nature to maldistribute changes inprecipi tion, then it may have some bearing on the question of precipitationvariability.

Changes in Precipitation Variability

A number of prominent climatologists have stated that world weatherpatterns in the near future are very likely to be more unstable, more variablefrom year to year, or from one short period of years to the next. In anattempt to get some probabilistic estimate of t)ikely changes in precipitationvariability, the panelists were asked in 0u' stion V to provide, underalternative assumptions of temperature change, conditional probabilities ofchanges by the year 2000 in the standard deviation of annual andgrowing-season precipitation in t,--ree bands of latitude. The threshold for asignificant change in variability was 25 percent of the standard deviationsassociated with the past 25 years.

It should be noted that many of the panelists again commented on the lacksufficient actuarial experience, comprehensive theories, or adequate


models to support their estimates. Given this caveat, the aggregate responsesof the panelists indicated a relatively high probability (mostly on the orderof 0.5 to 0.6 with a range from 0,4 to 0.7) that the change in the standarddeviation of the mean annual precipitation, would be less than 25 percent,irrespective of temperature trends. The second highest set of probabilities(on the order of 0.3 to 0.4) was assign to a significant increase inprecipitation variability under conditions of large cooling. The probabilitiesof significant decreases in variability do not exceed 0.2 except in most of thewarming cases, for which they rise slightly but still remain less than 0.3. Inthe two warming scenarios, significant idcreases and significant decreases ofprecipitation variability are nearly equiprobable in the range o 0.2 to 0.3. Inthe two cooling scenarios, by contrast, the probabilities r significantincreases of variability are higher than those for significant d creases by afactor of 1.5 to 2.0, but these probabilities are all less than 0.4 except forone case (growing-season precipitation in the subtropical latitudes).

Within each scenario, latitudinal differences in the probabilities of annualprecipitation variability are negligible, In the case of growing-seasonprecipitation, the probabilities for significantly increased variability growslightly larger with descending latitude in each scenario, while the smallerprobabilities for significantly decreased variability are about equal across thezones of latitude in each scenario.

Comparing the variabilities of annual and growing-season precipitation, onenotes slightly higher probabilities for significantly increased variability ofgrowing-season preupitation in the lower mid-latitudes from scenario toscenario. The same tendency is more marked in the subtropical latitudes.

The only obvious correlation between precipitation variability and trends inthe volume of precipitation is a tendency to associate greater variability withthe perception of decreased precipitation in the two cooling scenarios. Thereis a weaker tendency to associate decreased variability with increasedprecipitation in the two warming scenarios. These tendencies are manifestedvisually by the fact that the plots of the probabilities for changes in annualprecipitation volume and variability are approximate reflections (mirrorimages) of each other in the three zones of latitude. This mirror-imagerelationship exists to a lesser degree between the plots for changes ingrowing-season precipitation volume and variability.

The collective uncertainties about precipitation variability that are apparentin the panelists' data are also evident in the verbal comments on Question V.Below are some paraphrases of these comments, which run the gamut ofwhat could be said about variability:

Less variability with WM r-ningMore variability with coolingMore variability, regardless of temperature cliang-2

49


* No radical changes expectedes No cause to believe anything about change in the variability' Of any

meteorological element

Following are a few extracts from the actual comments about climaticvariability in general and precipitation variability in particular. The firstcomment is by a panelist who said he had no basis for predicting even globaltemperature trends.

A temperature change per so does not imply increased climatic variability.There is some physical basis for saying that a general coolirfg would implyincreased baroclinic instability and therefore increased temperaturevariability and vice-versa. On the other hand, one- might also expectgeneral cooling to be associated with less atmospheric water vapor, aweaker hydrologic cycle, and reduced precipitation.

The proposed associations between changes of temperature and ofprecipitation variability are difficult areas, on which the necessaryknowledge is further from adequate than with the preceding questions.This produces the wide spread of probability assessments which I give forthe higher middle latitudes. It may we be that the phasing ofprecipitation variability is more closely correlated with the periods ofchanging temperature than with the periods after a temperature anomalyhas been established; if so, the precipitation variability is probablygenerally greatest during cooling periods and least during warming periods.

The argument concerning van ility over time derives primarily from thesuggestion that variability see-- s to have been larger during periods ofclimatic deterioration, Le., cold Since my inference is for tempera-tures as warm, or warmer, than now by 2000, the precipitation variabilityshould not tend to increase. Precipitation variability is likely to increaseoverall in association with decreasing amounts, especially for convectiveregimes.

There seems to have been less variability during the 1945-60 period whenit was warmer than average, but I do not know whether this can beattributed to the temperature regime or not. It seems intuitivelyreasonable that a decrease in the equator-pole temperature gradient couldcause a more -summertime" condition and less variability due tobaroclinic instabilities and blocking pressure patterns.

. decreasing temperature trend will bring more variability, so the next25 years in mid latitudes should be noticeably more variable than the last25, In the subtropics probably not as much.

50


The 1940-1970 period was unusually uniform in the, perspective of the lastcentury. There is thus a high probability of more variability regardless oftemperature change .

I do not expect any radical changes. The variability is induced byextreme events. A single tropical storm rintall influences standarddeviations for a considerable interval,

. The standard deviation may not be a very good estimate of the sort ofvariability that might have practical implications, For example, proba=bility of certain extreme events might be more important.

DROUGHT AND MONSOON FAILURE

The last comment in the preceding section suggests the examination ofcertain extreme eventsdrought and monsoon failurein the context ofclimatic variability. Questions VI, VIII, and (X of Appendix A wereconcerned with mid-latitude drought, Asian monsoons, and Sahel Ian

droueL -r,,N3tively. In addition, Question VII (Outlook for 1977 CropYe! guantitative perceptions of the persistence of drought in theUnit (see Appendix MI

Mul a :;evere droughts do not lend themselves to discussion in stricttern interannual climatic variability. Nevertheless, in a broader sense,dre and monsoon failure are indicators of climatic variability insofar asthe\ manifestations of spatial and temporal fluctuations in precipitation.

U. -Id Other Mid Latitude Drought

The approach in Question-11./1 was to ask the panelists for probabilities of thefrequpncy of drought in the United States and other mid-latitudes."(fro nt was defined on a one-year basis in terms of crop yields: "Acombination of temperature and precipitation over a period of severalmonths leading to a reduction in yield of the major crops to a level less than90 percent of the yield expected with temperature/precipitation near thelong-term average values."' "Frequent" drought was defined as "similar toearly to mid 1930's and early to mid-1950's, "infrequent" drought as"similar to 1940's and 1960's," and "average" drought as "similar to thefrequency over the longest period of record available." Judgments cm theoccurrence of drought ,\./ere oskecr for each of the time periods 1977-80,1981 90, and 1991-2000.

'As noted in the footnote t

(lefinitions of r mght, munePutt their ! 111-11(:_7';t on infinite number ofIhhh tic -' r.1 mrilet"Iy -t-re 51


Figure 111-6

PROBABILITY OF DROUGHT

1977,80 1981 YO 1991-2000

PROBABILITYOF U.S.DROUGHT

.2

.3

4

5

.4

.2

0LC MC S MW LW

.2

3

PROBABILITY .4OF OTHER 5

MID-LATITUDEDROUGHT 4

.3

2

.1

0

PROBABILITYOFSAHELDROUGHT

0

1

4

5

4

3

LC 5 MW LWIi e top two rows ttlarrarottiri?rt the titiorecti iIci rewonses to Ciortqtion VI (Mid-Latitude Drought): The bottom row does

meow tor titles? ion IX (Sahel Droutint). For each tune pellod, the vertical bars reading from (ft (Large Cooling)to r eiht (Lame Wai ming) coo ettpond to the five climate minim iii For each !,,ceriario and taue tarn lull , the probabilities

-1;(4.,goeig" and l'inficquent- dotoght are given, rowectively, my tho heights of the lower or (Ind the upper bal.fir tt height if the Honk tp.tice hotween thette two hams measurer-, the probability tat -average- drought.

52


Analysis of the panelists' probability estimates (Figure III-6) of thefrequency of drought occurrences in the United States during the next 25years sungests fairly strong su6nortalthough by no means unanimousfor aquasi-20-year -periodicity, but the cause of this periodicity was clearly indispute among the panelists. Thus, in all the scenarios the probabitities of-frequent" U.S. droughts are fairly high (0.5 to 0.7) for the first and thirdperiods-1977-80 and 1991-2000. By contrast, the combined probabilities of"average" and "infrequent" U.S. droughts are high (0.6 to 0.8) in theintervening 1981-90 period-, ---\ except in the Large Warming Scenario. Onepanelist who strongly supporte this scenario gave for each of the three timeperiods probabilities of 0.6 to 03 for "frequent" U.S. droughts, reasoningthat 4,000 to 8,000 years ago "when the earth was generally a few degreeswarmer than now . . . it was markedly more rainy in the subtropics and alsorainier at mid latitudes in some regions but drier in others. It is notable thatin the central United States, in the lee of the Great Divide, it was generallydrier and the prairie extended nearly to the Appalachians."

For mid-latitude regions other than the United States, data on the expectedfrequency of drought are incomplete, and the responses suggest moreuncertainty and somewhat less support for the quasi-20-year periodicity thanwas evident for the United States. The explanatory comments by some ofthe panelists specifically noted that the pattern of repetition is not the samefor all mid-latitude regions. Looking at the probabilities for the terminal1991-2000 period (Figure III=6) one sees an association of more frequentdroughts with the two cooling scenarios and a slight tendency toward moreinfrequent droughts in the Large War ming Scenario. These associations areconsistent with and isomewhat stronger than the previously discussedperceptions that global cooling will tend to be accompanied by drierconditions and greater preCipitation variability, while global warming willtend to he accompanied by the opposite precipitation conditions.

Below are some of the panelists' comments on the estimation of droughtfrequency in the United States and other mid-latitude regions. Thesecomments reflect considerable credence in cyclic droughts, but notnecessarily in a connection of the cycles with solar--actiLyyty.

. I am not convinced that a solar-drought effect has been demonstrated.

I think that the eVidenc:! is mounting that there is a 22-year cycle indroughts and that it is related to sunspot activity. However, predictions ofthe relative intensity and regional distribution patterns lie beyond ourscientific knowledge.

While I do not believe that sunspots are related to drought, there is a

statistical behavior cif drought that 3r_rggrsts a pattern of repetition

53


54

Droughts seem to be related to solar influences ... Power spectrumanalysis shows a very notable 22-year rhythm and a weak one near 11'yearS.

. . I feel confident in saying that the risk of drought in the mid-latitudetends to wax and wane quite appreciably in a 20-year cycle. Thisjudgment on my part is the basis for the rather wide swings of probabilityshown in the table

The 20-year periodicity seems to be a well-documented climatic feature,not only as to Midwestern U.S droughts, but also in North AMericanwinter temperatures, English summer temperatures, and polar meantemperatures. The expected warming towards the end of the centurywould seem to increase the probability of drought

... The evidence seems to suggest that there is some rough 20= to 22-yearperiodicity in droughts in various parts of the High Plains, but that boththe length and location of these droughts varies considerably There isless known about the periodicity in other locations than the High Plains,and in fact, the whole issue of 22-year drought may be nothing more thana statistical coincidence at one location on earth. I do not believe that thesunspot hypothesis is demonstrated, because there is insufficient physicaland statistical evidence to lead to that certainty. Yet, I think it is

intriguing and one that requires considerable persistent study to look forpossible causal links.

There appear to he quasi-periodic recurrences of drought in 'manymid-latitude areasU.S. High Plains, 20 years and South Africa, 20yearsalthough physical explanations are not yet forthcoming.

I am extrapolating the 20 to 25-year cycle in spite of the fact that we doriot understand it (and I do not subscribe to the sunspot theory) . .

Other panelists alluded to specific mechanisms and other factors os the basesfor their probabilities of middatitude droughts:

I would assume that a strongly meridional circulation would continueinto part of the' 1980's, bringing few droughts to the United States andnir ') other mid latitudes. By the 1990's, the circulation may well revertto , zonality, with more frequent droughts in the UMted States andfewer ,.noughts in the more zonally patterned regions, e.g., Europe, USSR,and China.

I cannot put meaningful figures here without making latitude, meridian,,end recurrence distinctions, i.e., your term ''frequent'' applies rather tori,1;iirrent i,r r istr ue in certain regions (e.g., our Midwest) but at the sameUnit persistent rh nco in other reciionS our East Coast). In North


America, drought prevalence correlates highly negatively between EastCoast and Central Plains by decades.

The quasi-20-year periodicity (which I do not think attributable to thesolar cycle) is treated as best established and most important in the UnitedStates. In the United States, as elsewhere in middle latitudes, the droughtfrequency seems to me likely to remain higher than in the early part ofthe century owing to a longer-term increase of "blocking" frequency,which I associate with the present tendency of the natural climate towardsNorthern Hemisphere (and low latitudes) cooling,

Two panelists who did not furnish numerical estimates for Question VI madethe following comments:

I have no way of estimating the frequency of droughts except from a

frequency distribution, which varies from one region to another.

I know of no basis for answering the question, Certain statistical factscould be determined from past records, but to my knowledge this has notbeen done. It could most easily be done for tbe--tiSA arm Europe; dataelsewhere may be hard to assemble. The would give an"average" drought frequency (and range) but would pride no basis for

.different estimates in the three [time] periods,

Sahelian Drought and Failure of Asian Monsoons.

Questions VIII and IX, on Asian monsoons and -drought in the Sahel, usedthe same approach as Question VI (Mid-Latitude Drought), with appropriatechanges in the definitions of "frequent," "average," and "infrequent,"Monsoon failure and Sahellan drought were implicitly defined in terms'ofthe reference periods used to specify the three levels of frequency. About asmany panelists provided numerical estimates on the Sahel as did onmid latitude drought, but fewer responded c,n monsoon failure. One panelistwho did venture an estimate on monsoons .commented that "for other thanmonsoon scholars, this is a real guessing game."

In Figures III-6 and 111=7, the panelists' probability estimates of thefrequency of drought in the Sahel or failure of the Asian monsoons do notshow the periodic pattern that was indicated for the United States,general, the panelists' comments on Questions VIII and IX reflected aconsistency with their rationales for responses on other climatic elements.Thus, one sees, (si-!cially for the 1991-2000 period, the previously notedgeneral tendency to associate dry and more variable ("frequent" Saheliandroughts and monsoon failures) with giobal coohaq, and a weaker tendencyto associate wet and less variable (''--infrequent" Sahelian droughts andmonsoon failures) with rjiohul warm 79. These generalizations about Figures

55


56

Figure 111-7

PROBABILITY OF MONSOON FAILURE

1977 -80 1981 - 90 1991 2000

NORTHWESTINDIA

OTHERINDIA

0

.1Lu

.2 Luecuc

.3

4 Lu

LIJ

.4

Lu

1 LIJCC

0 LL

0

.1

.2

OTHERMONSOON .5ASIA .4,

3

0

a

raphs summarize the aggregrated responses to Question VIII (Asian 'Monsoons). For each time periodthe ver-.

tical -bars reading from left ( arrietooling) to right (Large Warming) coires'pond to the five climate scenarios. Ear eachscenario and =time period, the probatillities of "frequent", "infrequent", and "average" monsoon failures are displayedin the format of Figure 111-6.

sctNAmoseg PROBABILITIES

III-6 and III-7 are subject to obvious exceptions and to almost completeuncertainty in the Moderate Warming Scenario. In this scenario there wereeither insufficient data to warrant processing or virtually equal probabilitiesfor -frequent,:' "average," and "infrequent" occurrences of Saheliandrought and monsoon failures.

CONCLUSIONS ON CLIMATIC VARIABILITY

The five climate scenarios and associated data in Chapter I I bear out andpartially quantify climatologists' conflicting perceptions of global tempera-ture and precipitation trends. Less evident is the degree to which the climatescenarios and data support the more commonly accepted view that the earthmay be in for -a period of increased olimatic variability, whatever thetemperature trend.

Given the mate panelists' diverse comments about variability and sometendency to ssociate greater variability of precipitation and length ofgrowing season witr- global cooling, but less variability with warming, itcannot Le said that the responses to the climate questionnaire corroboratethe existence of general agreement about the onset of increased climaticvariability. As -partial reflections of the panelists' perceptions of futureclimate, these data could not provide direct, quantitative evidence ofincreasing claiatic variability. The direction' and magnitude of trends areobviously in dispute. The nature and extent of future climatic variability aremore obscure and hence less amenable to quantification. It may be that thepanelists were unaccustomed to thinking about variability in terms of thestandard deviation (for which, in any case, historical data are not generallyavailable). On the other hand, the cutoffs for a "significant" change invariability (±25 percent of the standard deviation) may have been too high.That is, a loWer cutoff might have elicited higher probabilities of stillsignificant variability.

One climatologist who reviewed the climate scenarios has an interestingconjecture about the tendency of the panelists not to predict significantchanges in climatic variability. The conjecture is that some respondents mayperceive that the world has already entered a period of increased variability,and, for that reason, do not expect any further major increase in the next 25years. This conjecture y be tenable for Question X on growing seasons,where the panelists were o compare future variability with that of the-preSetit.- It is less tenable for Question V on precipitation variability,where the climatologists wer asked to reference their estimates to theprevious 25-year period.

57


58

One panelist pointed out the importance of distinguishing climate variabilityand fluctuations with respect to'clifferent time scales:

There are two pertinent time ranges of variability, between weeks,months, and seasons within the year (like the temperature and rainfallextreme variations of the past yearsevere droughts of a season or two,reversing the next year, what I call hit-and-run droughts, which pertain toshifting of the wave pattern in strong zonal circulation), and long-termdroughts over periods of years (like our Midwest droughts of the 1930'sand 1950's which go with weak zonal circulation or strong blockingpatternsstrong nonseasonal systemspersisting from year to year). I feelthat at the present time the westerlies are shifting from the higher middletowards lower middle latitudes, bringing with them more of the

rtemporary but severe hit-and-run type of drought and severe cold(probably early and late killing frosts), whereas the higher middle latitudes(particularly Canada) from which the westerlies are being displaced, andthe subtropical latitudes (from which the subtropical high tends todisplace the tropical easterly rainbelt) are likely to suffer increasedincidence of the long-term type of drought and cold (Canada) and heat(subtropics). Thus, I find Question V hard to answer, but I have done sospecifically with respect to the short-term month-to-month type ofvariability, the hit-and-run variety which ikreflected in the length of thegrowing season.

Another respondent questioned use of the standard deviation in certainareas:

The standard deviation is not an adequate measure of variability ,wherethere is zero or near-zero precipitation. In semi-arid climates (and much ofthe world's grain fields have a semi-arid climate) the frequency elistribtion d e precipitation is very skewed ... and other measure's,- -e.g.,Maund s index of variability should be used.

Along this line of using other measures to get more directly at the impact ofclimate fluctuations, one recalls the panelist who commented (page 51) thatthe probability of certain extreme events might be more important than thestandard deviation in estimating the sort of variability which has practicalimplications. The latter panelist also pointed out a serious 4ifficulty in

dealing with any aspect of future climate by observing that "future climatic-changes, especially if related to human activities, need not follow the sameprobabilities as in the past

Given the generally accepted econo c social, and political importance ofclimatic fluctuations and variability, a lonal efforts should be made todevelop methoefblogies that can quantify climatologists' perceptions of past,current, and future climatic variability. Such efforts could shed light on theresearch needed to clarify what nature will do, as contrasted to whabcertain

,14experts believe it will do.

APPENDIX ACLIMATE QUESTIONNAIRE

INSTRUCTIONS FOR QUESTIONS

A. Attadied is a set of 10 major questions, some of which have several pacts. The times_ pan forthe questions varies fern the relatively near term (the outlook for the 1977 crop eason) to theclimate by the end of this century. All individual answers will be held in strict confidence. Theaggregation and quantitative analysis of the responses will be made available to all Participants andwill be included in the final report.

We would appreciate your answering all of the questions in their present form. Your subjectiveresponses may be used to generate another set of questions and to build a set of future climatescenarios. If, in reviewing and answering' hese questions, you feel strongly that a particular questionshould be rephrased or additional questions included, you are invited to add your comments oradditional questions on extra pages.

B. In questions referring to latitudinal belts, the following definitions apply:

POLAR latitudes gib, 65' to 90°MIDDLE latitudes 30' to 65'Higher middle 45' to 65'Lower middle 30' to 45',SUBTROPICAL latitudes 10" to no

C. For each of the 10 major questions, using the self- rating definitions provided below, pleaseindicate your level of su antive expertise.

D. Please identify those other respondents whomeyou would rank as EXPERT (5) or QUITEAM I LIAR (4) in responding to each of the questions.

E. Guidance for self-ranking expertise:

5) EXPERTYou should consider yourself an4expert if you belong to that small communityof p le who currently study, work on, and dedicate themselves to the subject matter. Typically,you know who else works in this area; you know the US and probably the foreign literature; youattend conferences and seminars on the subject, sometimes reading a paper and sometimes chairingthe sessions; you are most likely to have written up and/or published the results of your work. Ifthe National Science Foundation, National Academy of Sciences, or a similar organization were toconvene a seminar on this subject, you would expect to be invited, or, in your opinion, you shouldbe invited. Other experts in this field may disagree with your views but invariably-respect yourjudgment; comments such as "this is an excellent person on this subject" would be typical wheninquiring about you.

59

CLIMATE QUESTIONNAIRE

INSTRUCTIONS FOR QUESTIONS (Continued)

(4) QUITE FAMILIARYou are quite familiar with the subject matter either if you were anexpert some time ago but feel somewhat rusty now because other assignments have intervened (eventhough, because of therprevious interest, you have kept reasonably abreast of current developmentsin the field); or if yo4re in the process of becoming an expert but still have some way to go toachieve mastery of the subject; or if your concern is with integrating detailed developments in thearea, thus trading breadth of understanding for depth of specialization.

(3) FAMILIARYou are familiar with the subject matter if you know most of the argumentsadvanced for and against some of the controversial issues surrounding this subject, have read asubstantial amount about it, and have formed some opinion about it. However, if someone tried to

Irvin you down and have you explain the subject in more depth, you would soon have to admit thatyour knowledge is inadequate to do so.

(2) CASUALLY ACQUAINTEDYou are casually acquainted with the subject matter if youat least know what the issue is about, have read something on the subject, and/or have heard adebate about it on either a major TV or radio network or an educational channel.

1) UNFAMILIARYou are unfamiliar with the subject matter if the mention of itencounters a veritable blank in your memory or if you have,. heard of the subject, yet are unable tsay anything meaningful about it.

60


GLOBAL TEMPERATURES

Shown below is a historical record of changes in the annual mean temperature during the pastcentury for the latitude band, 0=80°N.

CHANGE ( °C) IN ANNUAL MEAN TEMPERATURE, 0-80 N. )_ATITUDE

0.8

0.6

0.4

0.2

0

0.21860 1880 1900 1920 1940 1960 1980 '20(

YEAR Source: Mitchell, NOAA

On the graph shown above, indicate your estimate of the general future course of the change inmean annual temperature (for 0-80' N. Lat.) to the year 2000 by:

drawing a temperature change path to the year 2000 so that you estimate only 1 chancein 10 that the path could be even lower

drawing a change path to the year 2000 so that you estimate an even cha4e that he pathcould be either lower or higher

drawing a change path to the year 2000 so that you estimate 1 chance in 10 that he pathcould be higher

61

0


I. GLOBAL TEMPERATURES

In the space below, state your line of reasoning for the family of lines you have drawn,referencing if you wish, articles you or other scientists have written that clearly state your positionon this subject.

Using the self-ranking definitions provided in the instructions, please indicate your levelof substantive expertise on this major question.

5 3=2-1Again using the self-ranking gu de, please identify -thole other respondents whom youwould rate as "expert (5)" "quite familiar (4)" in their answer to this particularquestion.

EXPERT (5) QUITE FAMILIAR (4)

62


IL T1MPERATURE

Pleasifill in each block of the matrix below with your estimate of the probability of thechange (° C) iytthe annual temperature by the year 2000, as 6-6mpared with 1970-75, for theregions

PROBABILITY TEMPERATURE CHANGE ( °C)

Cooling No Change" WarmingTotalProbability

More than1.0Cooler

0.5"-lto1.0"'Cooler

0 'to 0.5'.Cooler

0°to 0.5"Warmer

0.5° to1.0°Warmer

More than1.0°Warmer'

No. Hem, polar lartudes

No, Hem. higher m d latitudes

No. Hem. lower, mid latitudes

No. Hem. subtropical latitudes

So. Herd, subtropical latitudes

So. Hem. lower mid latitudes

So. Hem. higher mid latitudes

So. Hem. polar latitudes

1.0

1.0

1.0

1.0

1.0

1.0

1.1

.1.0

If you fudge that there is a significant probability that the temperature change in some latitudinal belt mayexceed 1.0°(either cooler or warmer), please indicate the level of change expected along with the probabilityestimate.

63


II. TEMPERATURE

For the preceding major question, please state the line of reasoning for your response,adding any amplifying remarks as you desire, or referencing articles you or otherscientists iiave written that state your position on this subject. Please use the spaceprovided below or a separate sheet.


5-4-3-271Again using the self-ranking guide, please identify those other respondents whom youwould rate as "expert (5)" or "quite familiar (4)" in their answer to this particularquestion.


64


III. CARBON DIOXIDE, TURBIDITY, AND CLIMATE

Carbon dioxide, atmospheric particles, etc., have different effects on the atmosphere and donot have the same relative importance in their influence. Indicate the relative weight (using

percentages) of each of the factors identified elow in influencing global climate over the next 25years.

Carbon Dioxide

Fluorocarbons and other gases

Smoke

Volcanic dust

Other Particles (aerosols)

Relative Weight

(Percentage)

Sum: 100%

65


III. CARBON DIOXIDE, TURBIDITY, AND CLIMATE

For the preceding major question, please state the line of reasoning for your response,adding any amplifying remarks as you desire, or referencing articles you or otherscientists have written that state your position on this subject. Please use the spaceprovided below or a'separate sheet.


5-4-3-2-1Again using the self-ranking guide, please identify those other respondents whom youwould rate as "expert (5)" or "quite familiar (4)" in their answer to this particularquestion.

EXPERT (5) QUITE FAMILIAR (4)z


IV. PRECIPITATION

1. Under the alter rive assumptions of a temperature change as shown, please fill in eachblock of th,=. following m trix with your estimate of the probability of change in the mean annualprecipitation by he year 2000, as compared with the 1941-70 -normal pattern, for..the- regionsshown.

2. In your judgment, would these probability estimates be equally as valid for growing seasonprecipitation as for annual totals? (yes D; no D) If not, please indicate the-appropriate probabilityestimates for changes in growing season precipitation.

3. In many parts of the earth the annual isotherms-and annual isohyets tend to run parallel toeat other. In North America, however, as Newman and Pickett dpscribe in a 6 December 1974Science article, the effect of the Rocky Mountains range causes the annual isotherms and isohyetsto run at approximately 90° to each other, particularly in the midContinental grasslands region, (Seetheir map_ below.) South America is similar to North America in this respect due to the Andes.Newman and Pickett note that:

= NORTH 20'AMERICA

35 F

40 F

50 F Fr60 F

/0 F20'

30-r

q

In terms of agricultural production,these climatic features give an advantage tothe NawWorld continents. The mean annualisotherms and isohyets, because they are notparallel in the vast grassland climatic areas ofthe Americas, allow a favorable water balanceto be extended over very broad areas in thenorth-south direction. Also, they allow foragricultural production areas to be less

stratified in a north-south direction than theyare in the Eurasian continent, and ensureagainst a slight mean seasonal shift in theprevailing westerly flow; thus, they ['educeclimatic risk=

Would these characteristics of the Western Hemisphere region significantly alter your estimates ofprobability of precipitation change as given in parts 1 and 2 above? If your estimates for North.andStith America are significantly different ftorp those of the broad latitudinal bands, please indicatehe magnitude of these differences in the comments at the end of this question. .

67

LrLI MICU C I Iti.PIIINIMInc-

PROBABILITY OF PRECIPITATION CHANGE

Increase by ".No h-ange" Decrease by Total10% or more"' .(Less than t 10%) 10% or mdre*. Probability

GrowingSeason

*Growing,axon

Assuming temp.increase of0.5°C or more

higher mid-latitudes

s lower mid-tatitudes

subtropicallatitudes

Assuming_ temp.change ofless than'± 0.5°C


- lower mid-.latitudes


Assuming tempdecrease of0.5°C ormore

higher mid:latitudes

lower mid-latitudes

subtropicallatitudeS

1,0.

1_0 -

'1.0 1.0

`If you judge that there is a significant probability that the precipitation change in some la10% (increase or decrease), please indicate the level of precipitation change expected aloestimate; also, if the temperature change assumed exceeds 1,0't, please so indicate.

68

al belt may exceedwith the probability

4'

IV, PRECIPITATION

For the prepeding major question, please stets the line of reasoning for your r spon5e,adding any amplifying .remarks as you detire; or referencing articles you ,Ctr-otherscientists have written 'that state your position on this subject. Please use the sp-aceprovided below or a separate sheet.

S

CLIMATE QIJE STIONN IRE-

Using the self4anking definitions Provided in the instructionc,--ffease indicate your level'of substantive expertise on this major question.

Again using he self-ranking guide, please identify those other respondents whom you,wciuld rate as "expert (5)- or "quite fimiliar (4)" in their answer to this particulir-question.

9

QUITE FAMILIAR (4)

69

i'FIC1PIXATION VARIABILITY

',The report of climate/food con erence in Bellagio, Italy, June 1975 includes thistement:

TE QUEQUESTIONNAIRE

Temperature ,change per se is not the most serious potentialclimatic threat to food p roduction: There is a pblity thatassociated with the temperature changes, there- will be increasedclimatic variability:- Particularly such variability increases thefluctuations of precipitation, it poses threats to agriculturalproduction.

- 4-The climatologists participating in the conference aimed tkat:

1. Climatic variabilityregion by region and from year to yearin particular regionsis 'and will contifice to be great, resulting insubstantial- variability in crop yields in the face of increasing globalfood needs and short supplies;

2. There is some cause to believealthough it is far fromcertainthat climatic variability in the remaining years of thiscentury will be even greater than during the 1940-1970 period.

Under the alternative assumptions o temperature change as shown, please fill in each blockof the matrix below with your probability timate that the standard deviation of the mean annualprecipitation will change by the indicated amount over The next 25 years, as compared with theaverage for the previous 25-year period.

2. In your judgment, would, these-probability estimates be equally as valid for variability ingrow7g season precipitation as for annual totals? (yes 0; no D) If not, please indicate theappropriate probability estimates for changes in variability of growing season precipitation.

3. As noted in part 3 of Question IV above, mountain ranges in North and South Americacause the annual isotherms and isohyets to run at approximately 900 to each other in major regionsof these continents. If because of these characteristics, your estimates of the probability of changein precipitation variability in North and South America are significantly different from those of thebroad latitudinal bands, please indicate the magnitude of these differences- in the comments at the

-end of this question.

71

- CLIMATE QUESTIONNAIRE

PROBABILITY OF CHANGE IN. PRECIPITATION VARIAWCITY

Increaserirys.d. by 5:.

of mor

Annual

...Lessthan Decrease in± 25% change - s.d. by 25% Totalin

.n s.d. or more -

,Probability

GrowingSeason Annu0

Growing Growing GrowingSeason Annual' Season Annual Season

' Assuming temp.increase of0.5° or more-

higher mid-, latitudes 1.0 1.0

I

I- lower mid-latitu es

sUbt opicallatitudes

Assuming temp.change of lessthan I- 0.5° C


lower mid-latitudes


Assuming temp.decreSse of0.5°- C or more

higher mid=latitudes

1.0 1.0

1.0

1.0

1,0 1.0

lower mid_ :latitudes: 1.0 1_0

subtropicallatitudes 1.0 1.0

If you judge that there is a significant probability that the change in standard de ation may ex6eed 25%(increase or decrease), please indicate the level of the change expected along with e probability estimate;also, if the temperature change assumed exceeds 1.0°C, please so indicate.

72-

V. : PRECIPITATION VARIABILITY

gLINIATEQUEST-IONNAR:

the preceding major question, please state the line of reasoning for your response,a ding 'any -amplifyirig remarks as you desire, or referencing articles you or Otherscientists have written that state your position on this subject. Please use the ,spadeproVided below or "a separate sheet.


Again using the self-ranking guide, please identify those other respondents-Wham youwould rate as "expert (5)" or "quite familiar (4)" in their answer to this particularquestion.


73

VI. MID-LATITUDE DROUGHT

At a 1975 climatic change symposium at the:University of North Carolina, Hurd Willett brieflydescribed the drought pattern of the 1930-1970 period as follows: _

The warm decade of the_-thirties witnessed the mosdseveredroughts of the century, in the early to mid-thicities, in many regioniof the 35' -5Cr latitude belt, notably the dust boWl in our westernplains, the Russian droughts that triggered liquidation of the Kulaks)severe drought, in southern Australia, and in other parts of the worldNote that these droughts occurred in marginal -midcontinental asopposed to east coastal regions.

The forties were in general a decade of generous rains in thedrought regions of the thirties, but with a tendency to substantialdeficiency in east coastal regions. .----- The early to° mid-fifties, like the thirties, were a markedly dryperiod- --in___the -marginal interior continental "regions, notably theAmerican iduthwestern plains. Severe drough(was restricted tolatitudes equatorWard of 400. Againthere was a notable tendency toeast coastal wetness.

The sixties were like thkforties, a decade- of generous rainfall inthe marginal interior contirkentaliregions of middle latitudes, butwith some record dry -years in extensive_east coastal regions. Duringthe sixties and early seventies, the deveblopment of severe droughtoccurred in the middle and lower subtropics; notably in so6thernAsia and Africa (Sahelian area).

Willett relates the drought occurrences to a solar-climatic hypothesis of climatic fluctUation,based on an observed relationship between solar and 'climatic cycles, but recognizing thatquanlitative physical explanations are as yet nonexistent.

4. With this statement as background, plus your own knowledge and interpretation of pastevents, please fill in each block of the matrix below with your estimate `cif the probability offrequency of drought occurrence, for global mid-latitude continental areas.

2. Would your estimateses for the United States be significantiy, iffedrent r6m o r

mid-latitude continental area (yes E; no 0) If so, please indicate probabilities sepaately for X.I.and other mid-latitudes.

[- CLIMATE QUOTIONNAIRE

MID-LATITUDE-DROUGHT-

-FREQUENCY-OF DROUGHT*

Timeperiod

"Frequent " -i.e.,similar to early toraid- 1930's and earlyto mid- 1950's

"Average"-i.e.,similar to thefrequency overthe longest periodof record available

"I nf requent"-i.e., similar to1940's and 1960's

.

TotalProbability

.

.

USOther mid-

US-Other mid-latitudes . US

Other mid-latitudes US '

Other mid-latitudes

1977. -to1980

'

1.0

to1990

1.0 1.0

1

to2000

.

1.0 1.0 -r

-*There are almost infinite number of definitions of drbu norfetitwhith m.are copletely satisfactory_ Drought- --.is defined here as ill the December 1973 report to the Ad nistrator of °AA; 3770: influence of Weather andClimate on United States Grain Yields:, Bumper -Cfo roughts: A: c ion

,:of temperaturer and

precipitation over a period of several months leading to a reduction in yield of the ajor crops to a level less than90% of the yield expected with temperature/precipitation near the long-term a values." It should berecognized, however, that yields usually are quoted on a harvested acre basis. In drought years there tends to be a

much larger abandonment of crops as well as some deviation in expected yield.

76

VL MID-LATITUDE DRIbupHT

CLIMAT QUESTIONNAIRE

For- the preceding major question, please state t e Fine of reasoning for your response,adding any amplifying rerrLarks as you desire, or referencing articles you or 'otherscientists have written that state your position op this subject. Please use the spaceProvided below or a separate sheet.


Again using the self-ranking guide, please identify those other respondents whom you.woqld rate as -expert (5)- or '!quite farnilar (4)" in their answdr to this particular,question.

EXPERT. ) QUITE FAMILIAR (4)

77


VI L OUTLOOK FOR 1977 CROP YEAR

The hibh plains of the. United States (that area apprOximately 400 miles wide, centered on101° west longitude from Mexico to Canada) has experienced lower than normal` precipitation andattendant drought conditions approximately every 20 to 22 years, i.e., the 1930's, Mid-1950's.

The chart below shows the 75-year record of summer average temperature and rainfall in thefive major wheat producing states of the United States. The drought period of the 1930's dust bowl rtand the generally favorable conditions afler the late 1950's show clearly.

FIVE "WHEAT BELT" STATES, (Oklahoma, Kansas, Nebraska South Dakota, and Wirth Dakota)

SUMMER RAINFALL

DUST BOWL ERAI

'HIGH YIELD ERA

SUMMER TEMPERATURE

YEAR 1900 1910,t 1920 1930 1940 1950 1960 1970

Source7 D. Gilman, NOAA

Low soil moisture levels in a number of stes in the U wheat and corn belts this past fall andwinter have raised considerable concern over prospects for the 1977 harvest.

1. With thd above information as background plus any supportive evidence or hypotheses youmay have developed as aids in forecasting, please fill in each block of the follow* matrix withyour probability estimate of weather conditions to be expedted in /1977that would result in theyield changes indicated_

(

79

CLIMA)-8 QUESTIONNAIRE

VII. - OUTLOOK FOR 1977 CROP YEAR

Probability of Yield. Change Due to Weather(Re [give to Recent Trend Levels)

DecreaseDecrease' Decrease -or increase Increasemore than 10% to - less than, more than Total15% 15% 1O% 10%- Probability

US winter wheat belt2

US spting wheat belt

US corn belt4

Implies drought`- conditions similar to the 1930's.

1.0

1.0

.0

2Six states Missouri, Nebraska, Kansas, Oklahoma, Texas, and Colorado count for about 50percent of winter wheat pr'aduction and have the most variable yields. Th 1976 total US winterwheat yield Was almost 32 bushels per acre, or about 8% below an estima d trend (1950-1976)value. In the 1970.76 period, the maximum deviations from that 1950-1976 rend line were +11%in 1971 and -12% in 1974.-

'Five states Minnesota, North Dakota, South Dakota, Montana, and Idahoaccount for about 90percent of spring wheat production. The 1976 total US spring wheat yield was 27 bushels per acre,_or about 7,.% below an estimated trend (1950-106) value. In the 1970-76 period, the maximum-deviations from that 1950-1976 trend line were +17%1 1971 and -20% in 1974.

'Nine states Ohio, Indiana, Illinois, Minnesota, Iowa, Missouri, Nebraska, South Dakota, andWisconsinaccount for-about 80 percent of corn production. The 1976 total US corn yield was 87bushels] per acre, or about 5% below an estimated trend (1.950-1976) value..In the 1970-76 period,the maximum deviations from that 1950-76 ,trend I ine_were +15% in 1972`and -20% in 1974.

2. Should a severe drought occur in 1977 -, dmila to the early to mid-4'930's, what is theprobability that it will persist for:

less than 2 years

b) 2 but less than 4 years

(c)\ 4 years or more

80

Probbility

1.0

CLIMATE l QUESTIONNAIRE

VII.' OUTLOOK FOR 1977 CROP YEAR

3. There appears to be a correlation with high plains drought conditions and solar activity, i.e.,the double or 22-year sunspot Ode, .although as Walter On Roberts noted in congressionalvstimony in MA/ 1976, -"There is no plausible explanation of how this [sunspot activity] couldaffect the weather of the high plains.- Roberts also commented that "Most of the world's droughts,howevet, show no recognizable recurrtnce pattern. They appear to occur at random in time and!option, thOugh they often persist ,for.live years or more in a giiren region." Would you expect tofind a correlation between drought conditions and solar activity on a global basis similar to thatwhich appears fa- have been identified for the US high plains? (yes 13; no D) Please amplify you ?..answer if you desire.

4: In view of the dry conditions noted above in parts of the United States and the resultantconcern over prospects for US crops this. year, plus the generally less than optimum level of worldgrain reserves, the prospects for harvests in other major grain producing areas of:the world are alsoof concern.

Based on any evidence available to you on current conditions, or hypotheses you may havedeveloped as aids in forecasting, please fill in the following tables with your probability estimates Ofweather conditions to be expected in 1977 that would result in the total grain yield changes

*indicated for the given countries.

A. USSR: Probability of total grain yield change due to weather (relative to recent trendlevels)

(a) Decrease more than 20%1(b) Decrease 10% to 20%(c) Decrease or increase less than 10%(d) Increase more than 10%

Probability

1

0

implies severe drought conditions similar to those in 1963, 1965, and 1975. For total So?iet graihyields in the 1970-76 period, the maximum deviations from a 1950-1976 trend line were +22% in1973 and -28% 410975.

81

ATE QUESTIONNAIRE

. VII. OUTLOOK FOR 1977 CROP YEAR

.INDIA: -Pr sbability of ,total grain yield change due to weather (relative to recent trendlevels)

(a) Decrease more than 10%'(b) Decrease 5% to 10©%(c) Decrease or increase less thar-%(d) Increase more than 5%

Probability

1.0

'Implies drought conditions similar-to or even more severe than in 1957, 1965, and 1966. For totalIndian grain yields in the 19/0-76 period, the maximum deviations from a 1950 -1976 'trend:linewere +8% in 1972 and about -7% in 1974 and 1976. -

CANADA: Probability of total grain yield change due to weather (relative to recenttrend levels)

(a) Decrease more 'than 25 %'(b) Decrease 10% to 25%(c) Dedrease or increase less than 10%(d) Increase more than 10%

Probability

1.0

Implies drought conditions similar to or even more severe than in 19§1. For total Canadian grainyields in the 1970-76 period, the maximum deviations from a 1950 -1976 trend line were +14% in1970 and -16% in 1974.

D. AUSTRALIA: Probability of total grain yield change due to weather (relative to recent,trend levels)

(a) Decrease more than 25 %'(b) Decrease 10% to 25%(c) Decrease or increase less than 10%(d) Increase more than 10%

Probability

1.0

' Implies drought conditions similar to or even more severe than 1957 and 1972. For totalinAustralian grain yields in the 1970-76 period, the maxi-mum deviations from a 1950-1976 trend

line were +10% in 1974 and -27% in 1972.

82

VII. OUTLOOK FOR 1977 CROP YEARN.;

ARGEItITINA: Probability of otal grain yield %hang due to weather (relative to regetfttrend levels)


(a) Decrease more than 15%'(b) Decrease Id% to 15%(c) Decrease or increase less than 10%

Increase more than 10%--r

1.0

lo Implies 'drought,eonditions similar to of even severe than in 1968. For total oirgenline4-Arainyields in the '1970-76 period, the maximum deviations from a 1950-1976 trend line were +14%

Probapility

in 4974 and -14% in 1972.

CLIMATE QUESTIONNAIREL

*-*- VII./ OUTLOOK FOR 1977 CROP YEAR

For the preceding major question, ease state the Iadding any amplifying remarks as 'you desire, orscientists have written that state your position onprovided below or a separate sheet.

Using thethe self-ranking definitions provided in the iof substantive.exper'tise on this major question.

5-4-3-2-1Again using the self-,ranking guide; please identify those other respondents ham youwould rate as "expert (5) or "guile familiar (4)" in their answer to this particularquestion.

ine of reasoning for your response,-referencing articles you or otherthis subject. Please use -the space.

L

actions, please indicate your level

'EXPERT (5) QUITE FAMILIAR (4)


VIII. ASIAN MONSOONS

PE CENTAGE OF WEATHER STATIONS HAVING LESS THAN 'HALF NORMAL RAINFALL

w

2,0

10

0

1900 1920 1940

Shown above.is a chart by Reid Bryon showing the trends in the percentage of weatherstations in northwest India reporting less than half the norrnal annual rainfall in a given year(overlapping 10:year averages).

1960

Source: Reid Bryson

With this information as background,, plus your iown knowledge and interprptation of pastpatterns in India, please fill in each block of., the following matrix with your estimate of theprobability of frequency of monsoon failures in northwest India.

0'

2. In your judgment, would Charts for (a) otheli" parts of India and (b) other summer monsoonregions in Asia show a similar pattern to that of northwest India (not necessarily using "less than

_half of normal annual rainfall- as a cutoff point)? (yes 0 no 0) If you judge that other regions ofIndia or monsoon. Asia would have a different pattern, pleaseincii5ate the appropriate pr bilitisin the following matrix.

85


VIII. AsjANMONSOONS

FREQUENCY OF MONSOON FAILURE

TimePeriod

'-'

"Frequent" =i.e., similarto 1900-1925 period .

"Average" -i.e., similarto the frequency overthe longest period ofrecord availab e

"Infrequent" -i.e similarto 1930-1960 periOd ,

' ,-,

TotalProbability ----

NWIndia

OtherIndia 'Monsoon

Other

Asia

NWIndia

OtherIndia

Othe rMonsoonAsia

NWIndia

OtherIndia

Other,1MonsoonAsia

fNN!India

OtherIndia

OtherMonsoonAsia'

1977to1980

1.0 1.0:

r-*

1.0

1981to

- 1990

.

. 1°.0 - 1.0

1991',.to - .

2000, -1.0

,1.0 1.0

( 11

VIII. ASIAN MONSOONS

CLIMATE QUESTIONNAIRE -

For the preceding major question, please state the line of reasoning for yoCir response,adding any amplifying remarks as you desire, or referencing articles you or otherscientists have written that state your position on this subject. Please use the spaceprovided below or a separate sheet.

Using the self- ranking definitions provided in the instructions, please indicate your levelof substantive- expertise on this major question.

5-4-3-2-1Again using the self-ranking guide, please identify those other respondents whom youwould rate as "expert (5)" or "quite fa -tiller (4)" in their answer to this particularquestion.


87

IX. _ SAHEL DROUGHT

E0

C0

4

30

1'0

203040

L


945 950 1955Year

1960 1965 1970

The above chart shows percentage deviations of the 5-year running means of annual rainfallfrom the 1931-60 mean for five stations in the SahelGao and Tessa lit in Mali; Atar andNouakchott in Mauritania; and Agadez in Niger. (From Bunting, et. al, in Nature, February 20,1975-, asad on Winstanley's data in Nature September 28, 1973.)

The above chart is consistent with National Academy of Science data shown below for fiveother stations in the Sahel. (From Michael Glantz in his Value of a Reliable Long-Range ClimateForecast for the Sahel, May 1976.)

Previous Extremes of Rainfall *

Runs of years of particularly high rainfall in ZONE 2 (Sahel)IndividualYears Rainfall as % of 1931-1960 Mean1929-1931 109 '106 - 121

1952 -1955 117 116 114 1081957-1962 112 106 .105 107

duns of years, Of particularly low rainfaA in ZONE 2 (Sahel)`Individual"Years1912.19151940-19441947:14491968-1973

itk85808372

4

Rainfall as %- of 1931-1960 Mean55 , 78 9482 76 101

93 . 6498 97 74

102 106

73

69 65

*National Acadey.of'Sciences, Arid Lands of Sub-Saharan Africa: Appendices (Washington, DC,1975), p. 155. Zone 2 rainfall stations were as follows: iamey and Zinder (Niger), Sokoto, Kano,and Maidurgu-ri (Nigeria).

89


IX. SAHEL DROUGHT

With this information as background, plus your knowledge and interpretation of the Saheliandrought patterns, please fill in each block of the matrix below with your estimate of-the probabilityof frequency of droUghts in the Sahel.

FREQUENCY OF DROUGHT

TimePeriod

"Frequent" -i e.similar to 1941950 and 1861973 periods

"Average" -i.e.,similar to thefrequency overthe longestperiod of recordavailable

"Infrequent -i.e.,similar to1950-1965 period

-TotalProbability

1977to1950 .

1.0

1981to'1990 .

1.0

1991to2000

-- 1.0

99


IX, SAHEL DROUGHT

For the preceding major question, please state the line of reasoning for your resporpe,adding any amplifying remarks as you desire, or referencing articles you or 'otherscientists have written that state your position on this subject. Please use the spaceprovided below or a separate sheet.

Using the self=ranking definitions provided in the instructions, please indicate your levelof substantive expertise on this major question.

- 5-4-3-2-1Again using the self-ranking guide, please identify -those other respondents whom Owwould rate as -expert (5)- or "quite, familiar (4) in their answer to this particulaquestion.

EXPERT,(5) QUITE FAMILIAR (4)


X: LENGTH OF GROWING SEASON

Professor Hubert Lamb has stated: "The average growing season in England for the coldestdecade. of theilast 300 years seems to have bee- n almost a month shorter than in 1930-60 . .. Thedifferenges of prevailing temperature in the English lowland districts have meant changes of 10 to20 days(in the average length of the grouting season in different decades since 1870. The shorteningby 9 to 10 days, since the warmest decades (1930's-40's) continues into the 1970's owing to thecold springs and, in the last year .01-, Iwo, colder autumns also." (Appendix A of report ckirRockefeller Foundation-sponsowl conference in Bellagio, Italy, J ne 1975.))

Walter Orr-Roberts has stated that In the USSR growin seasons are now erhaps lo gays to2 weeks shorter than in 1940-50." (House hearings on National Climate Program, May 20, 1976.)

In Wiscansin, however, the abstract of a paper to be presented at an AMS meeting in April1977 reports that:

Analysis of growing season records from stations representativeof each of Wisconsin's nine climatic divisions indicates that thegrowing season became cooler and shorter from 1958 to, themid-1960's. Subsequently, the same records hibit aveneral trendtoward warmer and longer growing sea3ons t rough 1973 in spite ofa continued fall in mean annual hemispheric " temperature and

1Ldeterioratin rowing weather elsewhere.

'Reduction in the length of the growing season, }particularly in the higher middle latitudes,might require certain crops to be groWn farther south than some areas where they are now grownand might require substitution of ;Hier maturing varieties in Some areas which could -rOuce cropyields.

fn your judgment, what is ti-Cprobability of the following changes in the mean length ofthe growing season in the higher middle latitudes during the next 25 years as compared with thepresent: /

a. Significant increase (say by 10 days or rr;ore)b. Change of less than -±10 days,c. Significant decrease (say by 10 days or'more)'

Probability

93

CLIMATE QUESTIONNAIRE9

X. LENGTH/ OF GROWING SEASON4

2. An analysis of the length of the frost-free season u F.1 in Iowa done a numb& of yearsago indicated a standard_ deviation' of about 16 to 17 ays. In your judgment, what is theprobability of a significant change in the inter-annual variabili v in the lengeT of the growing seasonsin the higher middle latitudes during the next 25 years as compared,with the present:

Significa'nt increase (say a 25% increase in the:standard deviation)

Change of less then ±-25% in the standard deviationSignificant decrease (say a 25% decrease in the

standard deviation)

Probability

.0

94


X. GTH OF GROWING SEASON

For tie preceding major question,' please state the line of reasoning for your response,adding any amplifying remarks as you desire, or referencing articles you or otherscientists have written that state your position on this subject. Please use the spaceprovided below or a separate sheet.

Using the self-ranking definitions provided in the instructions, please indicate your levelof substantive expertise -ion this major question.

5-4-3-2j1Again using f self 4nking guide, please identify those other respondents whom youwould rate as "expert (5)- or -"quite -familiar (4)" in their answer. to this 'parxicularquestion.


4

q,

.'

11.1:1,14

,

10$

,4P,

41.

APPENDIX BOUTLOOK FOR 1977 CROP YEAR &11-1EPERSISTENCE OF DROUGHT

During the mo th of April 1977, the climatology panelists prdvidedestimates for t: s expedted crop Oeld change-due to weather in 1977-4or thefollowing U.S. 'crops: winter wheat, spririg wheat, ar d corh, The lu'rfrimary oftheir re§ponses to Question V I I follows:

Table B-1

PROBABILITY erF YIELD CHANGE DUE TO WEATHER,(Relative to recent mind levels)*

U.S. vvint=er wheat beltU.S. sOring wheat beltU.S., corn belt

0,1

0.20.1 '

0.40,20.2

0 0 '2 ci _c _c - -0.40.50.5

0.1

0.1

0.2

tThe 12 October 1977 U.S. Department of Agriculture yield estim tes forthese crops were as. follows (figures in parentheses are the per entagedeviations from an estimated 1950-76 trend value extrapolated.to 197 )

Table B-2

USDA YIELD ESTIMATES

U.S. winter wheatstaring wheat

U.S. corn

-31.5 bu/acre (-9'%)

27.6 bu/acre (-6%)

90'.8 bu /acre (:3%)

The panelists-relative pessimism in the estimates for wheat, particatkrly thevvintef- whett crop, probably reflected the concern over the dry contemnsthaisted in many areas in late March and early April. Above normafipr capitation during the last half of April in much of the Great Plains regionimproved the grdwing conditions markedly.

*See 6uestionnaire for details on recent frond levels. 97,

77 CROP YEAR

In --addition the panelists provided estimates for drought persiStence., . s. .,,-,

Specifically, they were' asked, "Ivf a severe drou'ght similar to the early ormid-1930's'should occur in 1977, how long would it last?" The summary ofresponses fol tows:

Table 8-3 -, .

LENGTH OF DROUGHT IN THE UNITED STAT

Probability

.Persist less thaN2 yearsPersist for 2 but'less than 4 yearsersist for 4 or'more years

An 9Iternate method for presenting the summary esponses_ for droughtpersistence- is by a probability tree. This type of resentation is useful for,determining the probability ofdrought in the n xt year, given_ that the

ought has lasted a number of 'years. In the diagram on page 99, adrougbt is s*n to occur in 1977: Accordingto the panelists; theprobability that the drought ends in 19.78 is 0.6:- The probability that the'drought continues, then, is 1- minus the probability' of no drought; or 0.4.The "drought ends" branch of the tree terminates at the point A.

98

We now continue with the branqh in which &Light periists through 1978:The panelists estimated probability-of 0.3-thattp drought would last for 2or 3 Yea#. in pray to extend the tree, we make a reasonable interpolation:the probability of a drought taking only 2 years is assumed to be 0.2 and theprobability of a 3-year drodiht is takento be 9.1 In order to make theproduct of probabilities cqual -0.2 along the path' leading to an end of the,:drought in 1979 (point B);ve must ascribe a conditional probability of 0.5to t re case in which the drought:perSiStS through 1978 bufends in 1979.This, of course, iriplies a conditional probability of 0.5 that a droughtp rsisting through 1978'also persists u 111979.

By using the same conditional probabitities-0.5 for termin ion and 0.5 forpersistence of a drought that has lasted for at least a yearwe continue the,tree into 1980: Theeproduct of the probabilities along the branches leadingto the termination of drought' in 1980 (point C) is 0.1, in agreement with theinterpOlpion assumption about the. probability Of a drought lasting exaetly 3

Extending the tree one step' ;further 'into 1981 with the same conditionalprobabilities, we get a probability of 0.05 for a 4-year drought (point D) anda probability of 005 for 5 or More years of drought:rThese probabilities areconsisterrt with the panelists' probability of 0.,1 for a droUgii t which persiksfor 4 or more years. At each cApught-eneling,_ branch, the product of theprobabilities from the beginXW of the path to the -end is equal to thepanelists' estimate for a drought of that duration starting in 1977.

DROUGHTriilERSISTENCE IN THE UNITED STATES -

DROUGHT*1977

DROLIgHT

One may note that once a drought has started, there rela:tively highprobability that it will continue into the next year. Howeysr; .the reasoningabove should pot 'be extrapolated beyond the panelists' data Even as drawn,with the interpolated probabilities. for 2- and 3-year droughts and with theinferred conditicinal probabilities used for later years, the probability tree issubject to uncertainty and ambiguity. For instance, the panelists may haveassumed different definitions of drought termination, e.g., -the return to r

normal precipitation or -the return_.to normal water supplies. Moreover, it isre that ,the reference year 1977, taken in the context of double or

22-year sunspot 'cycles, 'affected' the panelists' perceptions of droughtpersistence. Although 1977 was a year of serious localized droughtconditions in the United States fell short of the WidespieaddroUghtexperienced in the-early to mid-1930's.

99

'APPENDIXNUMBER OF RESPONSES.AVERAGE.' EXPERT"- E, RATINGS

Approximately 50 international authorities on climatic change were consicl-ered Jas potential recipients of -the climatequestionnaire, but because of timeand:fealiurte constraints the number had to be reduced. Questionnaires weresent to 28 ,scientists, including -19 from .8 foreign countries. 'Replies werereceived frOm 24; their names and affiliations are listed:inthe acknowledg-ments. The 24 respondents participated in varying -degrees. Three, forexample, restricted their replies to qualitative- cornments. The other 21submitted the requested quanptatiVe data for ai_leas.t1 question; of these, 3answered onbra few questions. Fifteen Orovided: i<r itative responses to atleast 7 of the 10

The table belt5w shows fdr each of. the 10 questions the number ofclimatologists who submitted quantitative date' and the average of theirexpertise on the scale of 1 to 5 described in the questionnaire (AppendixThe number responding quantitatively to -a given question ranged fiorn 12(for Question VIII, Asian monsoons) to 19 (for Question I, globaltemperature, and Question III, atmospheric constituents).

Table C-1

.NUM 13ER OF RESPONSES AND EXPERTISE RATINGS ,

Question SubjeotNumber of .

Reipondents

19

17

19

.14'14-17

14

i2'1715

Global temperptureLatitudinal temperatureAtmospheric constituentsPrccipitapvtiPrecipitation variabilityMid - latitude drotiohtCrop Outlook for 1977Asian monsoonsSahel drought

Length of growing season

AverageExpertise

4.24.24.1

3.94,0

-4.0,3.73.4

j.8 _

3.9

red all the pprts of every question. The 10 quegions crossed aof global; hemispheric, zonal, and regicirrat" problems

ot only to climatic change but also to growing seasons and the.-6r. 1977 crops, arid nctindiviOual could have been expected to

nt in,all categories.

The average expertise ratings,for each of the questio were in the range of3.4 to 4.2 (3-familiar, 4-quite familiar, 5-expert). e firSt,two---questkopsabout temperatures had the highest expertise ratings. Question VII ron Asian

. _

monsoons had the lowest rating, 3.4.

102

,

4

-f.

D

REFffiENCE

In the climate questionnaire (Appendix A), the panelists*eregiVeri the opportunity to:inclureferences which explain or elaborate their response to each question. These references, arranged

'-;-:alphabetically- by questions=, are included hell as additional background material which may be ofvalUe to the reader. No specific references were giten for Questions VIII and X, on Asian Monsoonsand Length of Growing Season, respectively. References of a general nature are included at the endof the list.

I. Aiierage Global Temperature

An ell, J. and Karshover, J. ' Estimate of the Global, mange in. Tropospheric Temperature-getween 1958 and 1973." Monthly Weathei Review 103 (November 975): 1007-101.

'.'Estimate of the Glefol Clfenge in Temperature, Siirface to 100 mb, Between 1958nd 197 Monthly Weather Re View 105 (April 1977): 375-385.

f,:L

Borzenkova, 1.-1:, nnikov, p., .and Stekhnovskiy, D. I. Change in AirTemperature in the Northern Hemisphere during the Period 1881-1975:: - rology andHydrblogyzna. 71July 1976): 27-35. r

Broecker, W. S. Change: AreiTte on the Brink of a Pronounced Global WarmingS nce 189 (August 1975): 460-461'

rr

Bryson, Reid A. and 'Dittberner,_ Gerald J. "A Nan- Equilibrium Model of Hemispheric MeanSurface Temperature.- Journal of the Atmospheric Sciences 33 (November 1976)2094-2106.

Lamb, H. H. - Climate Present, Past and Future. London: Methuen, Vol. 1 (1972) and Vol. 2(1977).

Machta, L., Henson, K., and Keeling, C. D. -Atmospheric Carbon Dioxide and Some 4nterpreta-tidis,- in Proceedings of the Hawaiian Meeting, The Fate of Fossil Fuel carbon Dioxide,January 1976 (To be published).

Manabe, Syukuro 'and Wetherald Richard T. "The Effects of Doublinwthe'CO2 Concentration onthe Climate of a General Circulation Model." Journal of the Atmospheric Sciences 32(January 1975): 3-15.

National' Science Foundation. Interdepartmental Committee for Atmospheric Sciences (ICAS).Report of the Ad Hoc Panel on the Present Interglacial, ICAS 18b-FY75, August 1974. 103

131

REFERENCES

Newell, Reginald E. and Weare, Bryan C. "Factors Governing Tropospheric -Mean Temperature."Science =194 (Deceinber 1976): 14134414..

U.S. Congress. Senate. -CommAte-- on Aeronaulical and Space Sciences; Joseph Smagorinsky"Stratospheric Ozone Reseaith and Effects" in Hearings before the Subcdrnmittee on theUpper Atmosphere. 94th Cong., 2d sess.February 25Marchl, 1976..

Loon, Harry and Williams, Jill. "The Connection Between Trends of Mean Temperature indCirculation at the Surface: Part I. Winter?' Monthly Weather Review 104 (April 1976):365-380.,

"The Connection Between Trends of Mean Temperature and Circulation at theSurface: Part IL Summer." l to nthly Weather eview 104 (August 1976: 1003-1011.

_ 4

. "`he Connection Between Trends of Mean Temperature and Circulation at theurface: Part III. Spring and AutuMn.- Monthly Weather Review 104 (December 1976):

1591-1596.

"CompaAir and With the A

II. Average Latitudinal=

sons of Surface -Changes in t e Northern Hemisphere With the Upperarctic in Winter.",. Monthly W.e :Or Review 305 (May 1977): 636-6471.

em nature

Borzenkova, 1., Vinnikov, K. V., Spirina, L. P., and Stekhnovskjy, D. I: -Chang4 in AirTemperature in the Northern Hemisphere during the Period 1881- 1975." Meteorology and-Hydrology, no. 7 (July 1976): 2 S.

Damon, Paul E. and Kunen, Steven M. "Global Cooling? No. Southern/ Hemisphere WarmingTrends May Indicate Onset of the CO2 'Greenhouse' Effect." Science 193 (August 1976):447-453.

Flohn, H. -"Climate and Energy: A Scenario to a st.Century Climatic Change, Vol. 1,No. 1 -(March 1977): 6-20.

Manabe, Syukuro and Wetherald, Richard T. "The Effects of Doubling the CO2 Concentration onthe Climate- of a General Circulation Model."(January 1975): 3-15.

Journal of the Atrnolpheric Sciences 32

Salinger, ,M. J. " "News Zealand Temperatu es Since 1300 Nature 260 (March 1310-311.

Salinger, ryt. SI and Gunn. J 1'

(July ,lk5);

Tucker; G. B. "Climate:, Is Australia's Changing?" Search 6 (August 1975): 323-328.

I

r

ecent Climatic Warming Around N Zealand." Nature' 256

REFERENCES -

pni atom. World Meteorological Organization. "A clinical Report by the WMO ExecutiveCommittee Panel of Experts on Climatic Change." .W.l140 Bulletin 26 (January 1977): 50-55.

Nan Loon, Harry and Williams, Jill. "The Connection Between Trends of Mean Temperature andCirculationZ65-380.

at 'the Surface: Part l Winter.- Monthly Weather Review 104 (April 1976):

"The Connection Between Trends of Mean Tempetature and Circulation at theSurface: Part II. Summer."_ Monthly Weather Revie 04 (August 1976): 1003-1011.

"The Connection Between Trends of Mean Temperature and Circulation at theSurface: Part III. Spring and Autumn." Monthly Weathe, Review 104 (December 1976)':1591-1596.

. "Comparison of Surface Changes in the No thern Hemisp ere/With the Upper Airand With theArgarctjc in Winter." Monthly Weather R few 105 (May-1977):, 636-647.

--' ' L . e.

III Carbon Di xide and Turbidity

Bryson, Reid A. and Dirt ierner, Gerald J. "A Non-Equilibrium Model of HemispheA ean

Surface Temperature.- Journal of the Atthospheric Sciences -33 (November 1976):2094-21p6.

Budyko, M. I. and Vinnikov, K. Y. "Global Warming." Meteorology and Hydrology, no. 7 (July1976):16-26.

Damon, Paul E. and Kunen, Steven M. "Global Cooling? No. Southern Hemisphere WarmingTrends May. Indicate Onset of the CO2 'Greenhouse' Effect." Science 193 (August 1976):447-453.-

Gentilli, J. (Article relating to Temperature Changes After Volcanic Eruptions) GeologicalMigazine, Vol.,85, No. 3 (1948): 172-175.

rey, W. J. Physics 'f the Air. New York: McGravii Hill, 1940.

IKellogg, W. W., Citakley, J. A., and Grams, G. W. "Effect of Anthropageic Aerosols on the Global

Climate," in:I3roceedin& of WMO/IAMAF Symposium on Long-Term Climatic Fluctuations()Norwich, WMO Document 421. Geneva 1975.'323 -330.

Matthews, W. H., Kellogg, W. W. and Robinson, G. D., s. Man's Impact on the Climate.Cambridge, Mass.: MIT Press, 1971.

National Acade y of Sciences. Panel on Energy and Climate. Energy and climate. Washington,D.C.: 197-

Wang, W. C., Yun Y. L., Luis, A. A., Ma, T., and Hansen, J. 6. "Greenhouse Effects Due toMan er urbations of Trace Gases." -Science 194 (November 1-976): 685-690. 105

fIEFERENCES-

IV. Precipitation Change

Bryson, Reid A. "The Lessons of Climatic History.(Autumn 1975).

Ehviionmen al Conservation Vol. 2, No._

Lamb, H. H. Climate Presint, Past and Future. London: Methuen,(1977).

Leith C..

E : The Natural and Radietively urbed TrOposphere. CIAP Monograph 4, Chapter 3.Report No. DOT-TST 7.5-54, September 975.

Vol. 1 (1972) and Vol. 2

Manabe, Syukuto and Wetherald, Richard T. "The Effects of Doubling the CO2 Concentration onthe- Climate of a General Circulation Mcidel.- Journal of the Atmospheric Sciences 32(January 1975): 3-15.

Tucker, G. B. ate: Is Australia's Changing?' SeaPch 6 (August 1975) :"323 -328.

Wetheraid., Richard T. and Manabe, Syukuro. "The Effects of Changing the Solar Constant on the,Climate of a General Circulation Model." JoUrnal of the Atmospheric Sciences 32 (November1975): 2044-2059.

V. Precipitation Variability

Gentilli, J. "The Clirjrates of Australia and New Zealand." In13. Elsevrer Prets, 1971. pp. 157 -151

Id Survey of Climatology, vol.

Hare, F1 K.' (Chairman). Climate Change, Food Pr6duction, and Interstate Conflict, Report of anference, Bellagio, Italy, June 1975. New York: The Rockefeller Foundation, 1976.

Manabe, Syukuro and Wetherald, Richard T. "The Effects of Doubling the CO2 Concentration onthe Climate, of 4 General ,,Circulation Model." Journal of the Atmospheric Sciences 32(January 1975): 3 -15.

VI. Mid-latitude Drought

Eddy, J. A. "The Maunder Minimum." Science 192 (1976): 1189-1202.

"The Sun Since. the Bronze Age.- In Vol. 2 Physicsfo Solar PlanetaryEnvironments. A.G.U.: 1976. pp. 958-972.

Mar;hell, J. R. "The Annual Volume of United States Precipitation and the Variation_ for thePeriod- 1931-1971.- Journal of Geophysical Research 81 (1976): 4485-4486.

)Mock, Steve and Hibler, W. D. "The 21 Year Oscillation in Eastern North American Temperature."Nature 261 (1976): 484-486.

106

a

VII. Otitio Ok 1977 Crop Year

REFERENCES

Bryon, Reid A. and Starr, Thomas B. "Chandler -Tides inAtmosPreric Sciences 34 (December 1977): 1975-1986.

Eddy, J. A. 'The Maunder Minimum.- Science 192 (106): 1189-1202.

he Atmosphere." Jeurnal o

Environments

, .

Gentilli, J. ."The C ates of Australia and New Zealand." In World Survey of Climatology, Vol.13. , 1971. pp..190-193.

h Sun Since the Brqnze Age." In Vol. 2: Physics of Solar PlanetaryG.U.: 1976. pp. 9b6-972.

b, H. H. Clima Present, Pat and Future. London: Methuen, 1972. Vol. 1 'Chapter

Mavi, H. S. and Newman, J. E. 4 "Crop Production-Potential_ in IndiaA Water Availability BasedAnalysis." Agricultural Meteorology 17 (1976): 387-395.

IX. Sahel Drought

Bryson, R. A. "Drought in Sahelia.- Ecologist 3 (October 1973): 366-371.

Charney, J. G. "Dynamics.of Deserts and Drought in the Sahel." Quarterly Journal of the RoyalMeteorological Society 101 (April 1975): 193-202.

Charney, Jule and Stone, Peter H. "Drought n the Sahara: A Biogeophysical FeedbackMechanism." Science 187 (February (1975): 4-435. See- also-:'E-. A. -Ripley's.commentsand reply by Jule Charney and Peter_ H. Stone in Science 191 (January 1976): 100-102.

Landsberg, H.,. E. "Sahel Drought: Change of Climate or Part of Climate Arch. Met. Geoph.Siokl., Ser. B, 23 (1975): 193-200.

Mason, B. J. "Man's Influence on Wather and Climate." Journal l5f the Royal Society of Arts,Febnpry 1977.

es, M. K., and Pollard, C. K. "Changes in trhe Latitude of the Climatic Zones of the Northern.

Hemisphere.' Nature 252 (19744: 616

Tanaka, M., Weare, B. C., Navato; A. R., and Newell FL E. -ReCent African Rainfall Patterns."Nature 255 (1975): 201-203.

Tyson, P. D., Dyer, T. G..J. and Mametse,,p, N. "Secular Changes in South African Rainfall:to 1972.",.Quatterly Journal the Royal Meteorological Society 101 (1975): 817-

, - '-

United Nations. e. World Meteorologil Organiz n. "Drought.- WMO Special EnvironmentalReport-1Na 5 WMO No. 403. Geneva: 1975.

880

107

REFERENCES

"An Evaluation of Climate and Water Resources for Development of Agricultureiin-the Sudan-Sahelian Zone of W. Africa." WMO Special Environmental Report No. 9, WMONo. 459, Geneva 1976.

Wtnirtanley, Dere -Recent----43ainfa I--Trends in Africa, the Middle East and India." Nature 243(June 1973 : 464465.

"Rainfall Patterns and General Atmospheric Circulation.): 190-194. -46

Nature 245 (September =,

'Clime e Changes and the ,Future of ,the Sahel." In The Politics of a NaturalDisaster: The Case of the Sahel Drought, pp. 189-213. Edited by M. H. Glantz. New York:Praeger, 1976.

General Referencer.

Kellogg, W. W. "Mankind as a Factor-in Clin-We Change." In The Energy Question; pr. 241-258.Edited by E. W. Erickson and C. VVaverman. Toronto: University of Toronto Press, 1974 .

"Climatic Non-Limits to Growth." Proceedings of Conference on AtmosphericQuality and Climatic Change/ pp:76-89. University of North Carolina, Chapel Hill, North Caro-lina: 1975.

"Climate Change and the Influence of Man's Activities on the GlobalEnvironment." In The Changing Global Environment, pp. 13-21.- Edited by .S. F. Singer andD. Reidel. Dordrecht, Holland: 1975.

"Global Influences of Mankind on the Climate."- In Climate Change- Edited_byJ. Gribbin. Cambridge, England: Cambridge University Pre% thcoming.

-Anthropogenic Influences on Climate.- World Meteorological OrganizationTechnical Note. Geneva: forthcoming.

't4'Kellogg, W. W. and Schneider, S. H. -Climate Stabilization: Fbr Better or for Worse?" Science186 (1974): 1163-1172

'P

"Desertification, Human Activities, and Climate.- In Desertification. Edited byM. Glantz. Boulder, Colorado: Westview Press, forthcoming.

eLandsberg, H. E. and Albert, J. M. "Some Aspects of Global Climatic Fluctuations:, tArch.

Geoph. Biokl., Ser. B, 23 (1975): 165-176. , 7

National Academy of Sciences; Panel on Water and Climate. Climate, Climatic Change, and WaterSupply. Washington, D.C.: 1977.

Willett, Hurd C. "Do Recent Climatic Fluctuations Portend an Imminent Ice Age?" GeofisicaInternational 14 (1974): 265-302.

108

RE-FERENCES

"The ,Prediction of Great Salt Lake Levels on the Basis of Recent Solar-ClimaticCycles.- Paper presented at rAternational Conference on Terminal Lakes, Weber StateCollege, Ogden, Utah. May 197

Willett, Hurd C. andProhaska, John T. "Patterns, Possible Causes and Predictive Significance ofRecent Climatic Trends of the Northern Hemisphere.- The Solar Climatic Research Institute,Inc. Cambridge, Mass. Pre-publication Paper: March 1977.

109

ABTRACT tc - Education Resources Information Center · 9. Jointed ruler, early 20th century. 10....

Documents

Transcript of ABTRACT tc - Education Resources Information Center · 9. Jointed ruler, early 20th century. 10....