Using Satellite and Fully Coupled Regional Hydrologic, Ecological, … Potser_2009_NY.pdf ·...

1
Using Satellite and Fully Coupled Regional Hydrologic, Ecological, and Atmospheric Models to Study Complex Coastal Environmental Processes NASA Award Number NNX07AL79G 2 nd Year Progress 1 Zong-Liang Yang 1 Guo-Yue Niu 1 Seungbum Hong 1 David Maidment 1 Cedric David 2 Paul Montagna 2 Hae-Cheol Kim 2 Sandra Arismendez 3 James McClelland 4 Hongjie Xie 4 BeiBei Yu 1. University of Texas at Austin 2. Texas A&M University – Corpus Christi 3. University of Texas Marine Science Institute 4. University of Texas at San Antonio OBJECTIVES Improving our understanding of how linked upland and estuarine eco- system responds to combined changes in the hydrological and nutrient cycles that result from variation in climate and land use /land cover (LULC). Integrating research expertise from a diversity of fields that includes climate modeling, remote sensing analyses, biogeochemical cycling in watersheds, surface hydrology and estuary ecology. RESEARCH QUESTIONS (1) What is the relationship between global climate forcing and seasonal- to-interannual climate variability and extreme storm events over the Gulf Coast region? (2) What are the spatial patterns in LULC as defined by satellite data in the Gulf Coast region? (3) How does riverine nutrient export to Gulf Coast estuaries vary with LULC patterns and hydrologic conditions? (4) What is the relationship between the frequency of extreme events in the hydrologic and nutrient cycles and the mean productivity and the resiliency of productivity in Gulf Coast estuaries? (5) Can we use the answers to the questions above to predict the response of Gulf Coast estuaries to future climate perturbations? Schematic representations of the flow of data among the five components of this project. It consists of bias-corrected dynamical downscaling model, hydrological model, river routing model, water quality model, and estuary ecosystem model. Region12 Watershed area in Gulf Coast region STUDY REGIONS IMPROVED NEXRAD PRECIPITATION NEW DEVELOPMENT IN NOAH LSM To provide more improved and accurate precipitation as input to the Noah Land Surface Model (LSM), NEXRAD Stage IV data have been examined and processed for the Gulf of Mexico Coastal region. Spatial downscaling using parsi- monious physically-based mult- ivariate-regression algorithm, which includes rain pixel clustering and optimizing (Guan et al., 2007). Data accuracy examination using bias adjustment (BA), simple kriging with varying local means (SKlm), kriging with external drift (KED), and regression kriging (RK) methods. These methods have been evaluated by percentage bias (PBIAS), mean absolute error (MAE), coefficient of determina- tion (R 2 ), and Nash-Sutcliffe efficiency (NSE). No Correction BA SKlm KED RK 0 10 20 30 40 50 Absolute PBIAS No Correction BA SKlm KED RK 0.4 0.6 0.8 1 MAE No Correction BA SKlm KED RK 0.55 0.6 0.65 0.7 0.75 0.8 R 2 No Correction BA SKlm KED RK 0.65 0.7 0.75 0.8 0.85 0.9 NSE Spatial precipitation estimated by different methods (b: BA, c: SKlm, d: KED, e: RK) compared with the original NEXRAD map (a) at the same hour 8 th on April 24, 2004. Box-and-whisker plot of the PBIAS, MAE, R 2 and NSE values of the 50 rain gauge for four different methods. Results show that the average performance of SKlm is similar to or better than the other three methods. 1. Application of a separated canopy layer that distinguishes the canopy temperature from the ground temperature. 2. A modified two-stream radiation transfer scheme for the 3-D canopy structure effects on radiation transfer. 3. A Ball-Berry type stomatal resistance scheme. 4. A short-term dynamic vegetation model which accounts for photosynthesis, allocation of the assimilated carbon to various carbon pool such as leaf, stem, and wood, respiration of each carbon pool. 5. Application of a simple TOPMODEL-based runoff scheme and groundwater model (Niu et al., 2005; Niu et al., 2007). 6. A physically-based 3-layer snow model to accommodate snowpack internal processes. 7. A more permeable frozen soil to separate a grid cell into a permeable fraction and an impermeable fraction. Differences of annual runoff climatology, a) the Noah baseline model minus Global Runoff Data Center (GRDC), b) modified Noah (EXP6) minus the Noah baseline model, and c) modified Noah minus GRDC. Application to Guadalupe and San Antonio river basins : precipitation (upper), evapotranspiration (middle), runoff (lower). Equipped with multiple parameterization options RIVER ROUTING MODEL NUTRIENT EXPORT STUDY A river network model called RAPID (Routing Application for Parallel computation of Discharge) has been used to produce stream flow estimation from runoff simulations by the Noah LSM. Coupling of RAPID and the Noah LSM is facilitated by a flux coupler which is designed to convert the gridded runoff into lateral inflow for a river network extracted from the NHDPlus dataset. Evaluation of modeled stream flow at five locations in the Guadalupe and San Antonio River Basins (GS: Guadalupe River, CC: Coleto Creek, SAR: San Antonio River). Runoff simulation 900m grid of catchment COUPLING Runoff is spatially averaged for catchments surrounding NHDPlus flow lines. 0.01 0.1 1 10 100 concentration (µM) 0 100 200 300 400 500 600 0.01 0.1 1 10 100 0 10 20 30 40 50 60 Runoff (mm/day) 0.01 0.1 1 10 100 0 4 8 12 16 20 0.01 0.1 1 10 100 0 4 8 12 16 20 Aransas River Mission River Nitrate Nitrate Ammonium Ammonium 0.01 0.1 1 10 100 Concentration (µM) 0 100 200 300 400 500 600 San Antonio 0.01 0.1 1 10 100 0 100 200 300 400 500 600 Guadalupe Nitrate Nitrate 0.01 0.1 1 10 100 0 4 8 12 16 20 Ammonium Runoff (mm/day) 0.01 0.1 1 10 100 0 4 8 12 16 20 Ammonium Runoff (mm/day) 0.01 0.1 1 10 100 DON concentration (µM) 0 10 20 30 40 50 60 Mission River Aransas River Dissolved organic nitrogen (DON) concentrations increase as runoff increases, maintaining values between ~25 and 55 micromoles per liter during high flow Nitrate and ammonium concentrations dilute as runoff increases, maintaining values below 5 micro- moles per liter during high flow. The dilution effect is more evident for nitrate than ammonium. River water sampling in the Guadalupe, San Antonio, Mission, and Aransas rivers during storm events as well as base flow conditions is increasing our understanding of how nutrient concentrations and export vary with flow. Water quality data from independent research and monitoring efforts in Copano Bay have been compiled to support the estuarine ecosystem modeling component. The dataset for Copano Bay is exceptionally rich because it has been the focus of a NOAA sponsored study on watershed export events during the past two years and because the Mission Aransas National Estuarine Research Reserve (MANERR) maintains two monitoring stations in the bay. ESTUARY ECOSYSTEM MODELING Quarterly sampling has been accomplished in July 2008, October 2008, January 2009, and April 2009. Differing basin characteristics, such as size, population, land use, permitted discharges, precipitation and flow, can lead to changes in nutrient loadings, and model simulations revealed different ecosystem responses (e.g., timing, duration and magnitude of dissolved inorganic nitrogen, phytoplankton and zooplank- ton) to different scenarios of nutrient loading. Relationship between the first principal components of water quality (Water Quality PC1) and land use/land cover (LCLU PC1) data for HUC and year (1992 and 2001) in the Guadalupe (GRB) and San Antonio (SARB) River Basins. Temporal dynamics of prognostic variables under three different scenarios of DIN loading: a. no daily DIN loads, b. daily DIN loads only from Lower San Antonio River, and c. daily DIN loads only from Lower Guadalupe River. All three assumed a condition that only 50% of detrital matters is remineralized, and the remaining 50% sinks to bottom. a b c FUTURE WORK Developing soil and water assessment tool (SWAT) for water quality estimation. Simulating more accurate runoff for Region12 through parameterization of vegetation properties with utilization of remote sensing data such as LULC and NDVI. Developing more realistic ecosystem loadings- based model and expanding work to other river basins along Texas coast. Investigating LULC changes in Texas to asses their effects on land surface and estuary ecosystem processes. References Guan, H., H. Xie, and J. L. Wilson (2007) A Physically-Based Multivariate-Regression Approach for Downscaling NEXRAD Precipitation in Mountainous Terrain. American Geophysical Union, Fall Meeting 2007, abstract #H23C-29. Niu, G-Y., Z-L. Yang, R. E. Dickinson, L. E. Gulden, and H. Su (2007) Development of a simple groundwater model for use in climate models and evaluation with GRACE data, J. Geophys. Res., 112, D07103, doi:10.1029/2006JD007522. Niu, G.-Y., Z.-L. Yang, R. E. Dickinson, and L. E. Gulden (2005) A simple TOPMODEL-based runoff parameterization (SIMTOP) for use in global climate models. J. Geophys. Res., 110, D21106, doi:10.1029/2005JD006111. Published and in-preparation papers supported by this grant Arismendez, S., H. Kim, J. Brenner, P. Montagna (2009) Application of watershed analyses and ecosystem modeling to investigate land-water nutrient coupling processes in the Guadalupe Estuary, Texas. Ecological Informatics. Submitted. David, C. H., and F. Habets (2009) Routing Application for Parallel computatIon of Discharge (RAPID). in preparation. Niu, G-Y., Z-L. Yang, K. E. Mitchell, F. Chen, M. B. Ek, M. Barlage, L. Longuevergne, M. Tewari (2009) The community Noah Land Surface Model with Multi-physics options. J. Geophys. Res. Submitted. Wang, X., H. Xie, H. Sharif, and J. Zeitler (2008) Validating NEXRDA MPE and Stage III precipitation products for uniform rainfall on the Upper Guadalupe River Basin of the Texas Hill Country. Journal of Hydrology, Vol. 348(1-2): 73-86, doi: 10.1016/j.jhydrol.2007.09.057 Weissling, B. and H. Xie (2008) Proximal watershed validation of a remote sensing-based streamflow estimation model. Journal of Applied Remote Sensing. Vol. 2, 023549, DOI:10.1117/1.3043664. San Antonio & Guadalupe River Basin Guadalupe Estuary Copano Bay Estuary

Transcript of Using Satellite and Fully Coupled Regional Hydrologic, Ecological, … Potser_2009_NY.pdf ·...

Page 1: Using Satellite and Fully Coupled Regional Hydrologic, Ecological, … Potser_2009_NY.pdf · 2009-07-13 · Using Satellite and Fully Coupled Regional Hydrologic, Ecological, and

Using Satellite and Fully Coupled Regional Hydrologic, Ecological, and Atmospheric Models to Study Complex Coastal Environmental Processes

NASA Award Number

NNX07AL79G2nd Year Progress

1Zong-Liang Yang 1Guo-Yue Niu 1Seungbum Hong 1David Maidment 1Cedric David 2Paul Montagna 2Hae-Cheol Kim 2Sandra Arismendez 3James McClelland 4Hongjie Xie 4BeiBei Yu

1. University of Texas at Austin2. Texas A&M University – Corpus Christi3. University of Texas Marine Science Institute4. University of Texas at San Antonio

OBJEC TIVESImproving our understanding of howlinked upland and estuarine eco-system responds to combinedchanges in the hydrological andnutrient cycles that result fromvariation in climate and land use/land cover (LULC).Integrating research expertise froma diversity of fields that includesclimate modeling, remote sensinganalyses, biogeochemical cycling inwatersheds, surface hydrology andestuary ecology.

RESEARCH QUESTIONS(1) What is the relationship betweenglobal climate forcing and seasonal-to-interannual climate variability andextreme storm events over the GulfCoast region?

(2) What are the spatial patterns inLULC as defined by satellite data inthe Gulf Coast region?

(3) How does riverine nutrient exportto Gulf Coast estuaries vary withLULC patterns and hydrologicconditions?

(4) What is the relationship betweenthe frequency of extreme events inthe hydrologic and nutrient cyclesand the mean productivity and theresiliency of productivity in GulfCoast estuaries?

(5) Can we use the answers to thequestions above to predict theresponse of Gulf Coast estuaries tofuture climate perturbations?

Schematic representations of the flow of data among the five components of this project. It consists of bias-corrected dynamical downscaling model, hydrological model, river routing model, water quality model, and estuary ecosystem model.

Region12 Watershed area in Gulf Coast region

STUDY REGIONS

IMPROVED NEXRAD PRECIPITATION

NEW DEVELOPMENT IN NOAH LSM

To provide more improved andaccurate precipitation as input tothe Noah Land Surface Model(LSM), NEXRAD Stage IV data havebeen examined and processed forthe Gulf of Mexico Coastal region.

Spatial downscaling using parsi-monious physically-based mult-ivariate-regression algorithm,which includes rain pixel clusteringand optimizing (Guan et al., 2007).

Data accuracy examination usingbias adjustment (BA), simplekriging with varying local means(SKlm), kriging with external drift(KED), and regression kriging (RK)methods. These methods havebeen evaluated by percentage bias(PBIAS), mean absolute error(MAE), coefficient of determina-tion (R2), and Nash-Sutcliffeefficiency (NSE).

No CorrectionBA SKlm KED RK0

10

20

30

40

50

Abs

olut

e P

BIA

S

No CorrectionBA SKlm KED RK

0.4

0.6

0.8

1

MA

E

No CorrectionBA SKlm KED RK0.55

0.6

0.65

0.7

0.75

0.8

R2

No CorrectionBA SKlm KED RK0.65

0.7

0.75

0.8

0.85

0.9

NS

E

Spatial precipitation estimated by different methods (b: BA, c: SKlm, d: KED, e: RK) compared with the original NEXRAD map (a) at the same hour 8th on April 24, 2004.

Box-and-whisker plot of the PBIAS, MAE, R2 and NSE values of the 50 rain gauge for four different methods. Results show that the average performance of SKlm is similar to or better than the other three methods.

1. Application of a separatedcanopy layer that distinguishes thecanopy temperature from theground temperature.

2. A modified two-streamradiation transfer scheme for the3-D canopy structure effects onradiation transfer.

3. A Ball-Berry type stomatalresistance scheme.

4. A short-term dynamicvegetation model which accountsfor photosynthesis, allocation ofthe assimilated carbon to variouscarbon pool such as leaf, stem,and wood, respiration of eachcarbon pool.

5. Application of a simpleTOPMODEL-based runoff schemeand groundwater model (Niu etal., 2005; Niu et al., 2007).

6. A physically-based 3-layer snowmodel to accommodate snowpackinternal processes.

7. A more permeable frozen soil toseparate a grid cell into apermeable fraction and animpermeable fraction.

Differences of annual runoff climatology, a) the Noah baseline model minus Global Runoff Data Center (GRDC), b) modified Noah (EXP6) minus the Noah baseline model, and c) modified Noah minus GRDC.

Application to Guadalupe and San Antonio river basins : precipitation (upper), evapotranspiration (middle), runoff (lower).

Equipped with multiple parameterization options

RIVER ROUTING MODEL

NUTRIENT EXPORT STUDY

A river network model called RAPID (RoutingApplication for Parallel computation of Discharge)has been used to produce stream flow estimationfrom runoff simulations by the Noah LSM.

Coupling of RAPID and the Noah LSM is facilitatedby a flux coupler which is designed to convert thegridded runoff into lateral inflow for a river networkextracted from the NHDPlus dataset.

Evaluation of modeled stream flow at five locations in the Guadalupe and San Antonio River Basins (GS: Guadalupe River, CC: Coleto Creek, SAR: San Antonio River).

Runoff simulation 900m grid of catchment

COUPLING

Runoff is spatially averaged for catchments surrounding NHDPlus flow lines.

0.01 0.1 1 10 100

conc

entra

tion

(µM

)

0

100

200

300

400

500

600

0.01 0.1 1 10 1000

10

20

30

40

50

60

Runoff (mm/day)0.01 0.1 1 10 100

0

4

8

12

16

20

0.01 0.1 1 10 1000

4

8

12

16

20

Aransas River Mission River

Nitrate Nitrate

Ammonium Ammonium

0.01 0.1 1 10 100

Con

cent

ratio

n (µ

M)

0

100

200

300

400

500

600

San Antonio

0.01 0.1 1 10 1000

100

200

300

400

500

600

Guadalupe

Nitrate Nitrate

0.01 0.1 1 10 1000

4

8

12

16

20 Ammonium

Runoff (mm/day)

0.01 0.1 1 10 1000

4

8

12

16

20 Ammonium

Runoff (mm/day)0.01 0.1 1 10 100

DO

N c

once

ntra

tion

(µM

)

0

10

20

30

40

50

60Mission RiverAransas River

Dissolved organic nitrogen (DON) concentrations increase as runoff increases, maintaining values between ~25 and 55 micromoles per liter during high flow

Nitrate and ammonium concentrations dilute as runoff increases, maintaining values below 5 micro-moles per liter during high flow. The dilution effect is more evident for nitrate than ammonium.

River water sampling in the Guadalupe, SanAntonio, Mission, and Aransas rivers during stormevents as well as base flow conditions is increasingour understanding of how nutrient concentrationsand export vary with flow.

Water quality data from independent research andmonitoring efforts in Copano Bay have beencompiled to support the estuarine ecosystemmodeling component. The dataset for Copano Bayis exceptionally rich because it has been the focusof a NOAA sponsored study on watershed exportevents during the past two years and because theMission Aransas National Estuarine ResearchReserve (MANERR) maintains two monitoringstations in the bay.

ESTUARY ECOSYSTEM MODELING

Quarterly sampling has been accomplished inJuly 2008, October 2008, January 2009, andApril 2009.

Differing basin characteristics, such as size,population, land use, permitted discharges,precipitation and flow, can lead to changes innutrient loadings, and model simulationsrevealed different ecosystem responses (e.g.,timing, duration and magnitude of dissolvedinorganic nitrogen, phytoplankton and zooplank-ton) to different scenarios of nutrient loading.

Relationship between the first principal components of water quality (Water Quality PC1) and land use/land cover (LCLU PC1) data for HUC and year (1992 and 2001) in the Guadalupe (GRB) and San Antonio (SARB) River Basins.

Temporal dynamics of prognostic variables under three different scenarios of DIN loading: a. no daily DIN loads, b. daily DIN loads only from Lower San Antonio River, and c. daily DIN loads only from Lower Guadalupe River. All three assumed a condition that only 50% of detrital matters is remineralized, and the remaining 50% sinks to bottom.

a

b

c

FUTURE WORKDeveloping soil and water assessment tool (SWAT)for water quality estimation.Simulating more accurate runoff for Region12through parameterization of vegetation propertieswith utilization of remote sensing data such as LULCand NDVI.Developing more realistic ecosystem loadings-based model and expanding work to other riverbasins along Texas coast.Investigating LULC changes in Texas to asses theireffects on land surface and estuary ecosystemprocesses.

ReferencesGuan, H., H. Xie, and J. L. Wilson (2007) A Physically-Based Multivariate-Regression Approach for

Downscaling NEXRAD Precipitation in Mountainous Terrain. American Geophysical Union, Fall Meeting 2007, abstract #H23C-29.

Niu, G-Y., Z-L. Yang, R. E. Dickinson, L. E. Gulden, and H. Su (2007) Development of a simple groundwater model for use in climate models and evaluation with GRACE data, J. Geophys. Res., 112, D07103, doi:10.1029/2006JD007522.

Niu, G.-Y., Z.-L. Yang, R. E. Dickinson, and L. E. Gulden (2005) A simple TOPMODEL-based runoff parameterization (SIMTOP) for use in global climate models. J. Geophys. Res., 110, D21106, doi:10.1029/2005JD006111.

Published and in-preparation papers supported by this grantArismendez, S., H. Kim, J. Brenner, P. Montagna (2009) Application of watershed analyses and

ecosystem modeling to investigate land-water nutrient coupling processes in the Guadalupe Estuary, Texas. Ecological Informatics. Submitted.

David, C. H., and F. Habets (2009) Routing Application for Parallel computatIon of Discharge (RAPID). in preparation.

Niu, G-Y., Z-L. Yang, K. E. Mitchell, F. Chen, M. B. Ek, M. Barlage, L. Longuevergne, M. Tewari(2009) The community Noah Land Surface Model with Multi-physics options. J. Geophys. Res.Submitted.

Wang, X., H. Xie, H. Sharif, and J. Zeitler (2008) Validating NEXRDA MPE and Stage III precipitation products for uniform rainfall on the Upper Guadalupe River Basin of the Texas Hill Country. Journal of Hydrology, Vol. 348(1-2): 73-86, doi: 10.1016/j.jhydrol.2007.09.057

Weissling, B. and H. Xie (2008) Proximal watershed validation of a remote sensing-based streamflowestimation model. Journal of Applied Remote Sensing. Vol. 2, 023549, DOI:10.1117/1.3043664.

San Antonio & Guadalupe River Basin

Guadalupe EstuaryCopano Bay Estuary