Скачать презентацию Land Data Products For Energy Budget and Water Скачать презентацию Land Data Products For Energy Budget and Water

ca4b02e198f8497cf8b24cfb0566defd.ppt

  • Количество слайдов: 43

Land Data Products For Energy Budget and Water Cycle Trends and Processes • Is Land Data Products For Energy Budget and Water Cycle Trends and Processes • Is there a mostly observational pathway? • What are the roles of “models” and data assimilation? • What did we learn from Initiative II about the land data? – – What’s finished? What’s in good shape, what’s not? What is completely missing? What needs to happen next? Forrest Hall 05/07

International Satellite Land Surface Climatology Project (ISLSCP) Initiative II Forrest Hall 1, 2, Eric International Satellite Land Surface Climatology Project (ISLSCP) Initiative II Forrest Hall 1, 2, Eric Brown de Colstoun 1, 4, David Landis 1, 3, Lahouari Bonoua 1, 4 AND James G. Collatz 1 (PI) 1 NASA/Goddard Space Flight Center, Greenbelt, MD, U. S. A. 2 University of Maryland Baltimore County, Catonsville, MD, University of Maryland College Park, MD, Science Systems and Applications Inc. , Lanham, MD, U. S. A.

Science Working Group D D D Pavel Kabat, DLO Winand Staring Centre (ISLSCP II Science Working Group D D D Pavel Kabat, DLO Winand Staring Centre (ISLSCP II Chair) Paul Try, GEWEX Ichtiaque Rasool, University of Paris VI Randy Koster, NASA/GSFC (Modeling requirements) Carbon: Scott Denning, Colorado State Univ. , Dick Olson, Oak Ridge National Laboratory Vegetation: Sietse Los, Ruth De. Fries, Univ. of Maryland, Alan Strahler, Boston Univ. Radiation/Clouds: Paul Stackhouse (NASA/Langley) Near-surface Meteorology: Alan Betts, Pedro Viterbo (ECMWF), Glenn White (NCEP), Paul Dirmeyer, Center for Ocean-Land-Atmosphere Studies (COLA), . Socioeconomic: Marc Levy, Deborah Balk, Socioeconomic Data ad Applications Center (CIESIN) Snow/Ice: Richard Armstrong, National Snow and Ice Data Center, Univ. of Colorado. Topography/Soils/Runoff: Kris Verdin USGS/EROS Data Center, Balasz Fekete, Univ. of New Hampshire. Precipitation: George Huffman (SSAI) Arnold Gruber (NOAA)

ISLSCP SPECIAL ISSUE JGR ATMOSPHERES, NOV 2006 Forrest Hall 05/07 ISLSCP SPECIAL ISSUE JGR ATMOSPHERES, NOV 2006 Forrest Hall 05/07

 • We have assembled a comprehensive, multi-disciplinary data collection to support global carbon, • We have assembled a comprehensive, multi-disciplinary data collection to support global carbon, water and energy modeling efforts, spanning 1986 -1995. • 52 data sets (272 individual parameters) have been processed to common grids and land/water boundaries (1/4, 1/2 and 1 degree spatial resolution). • All data sets and documentation have been peer-reviewed. • Data collection currently available at http: //islscp 2. sesda. com/ISLSCP 2_1/html_pag es/islscp 2_home. html • A special issue of JGR-Atmospheres reports the community-wide evaluation of the collection: * Hall, F. G. , E. Brown de Colstoun, G. J. Collatz, D. Landis, P. Dirmeyer, A. Betts, G. J. Huffman, L. Bounoua, and B. Meeson (2006), ISLSCP Initiative II global data sets: Surface boundary conditions and atmospheric forcings for land-atmosphere studies, J. Geophys. Res. , 111, D 22 S 01, doi: 10. 1029/2006 JD 007366. * Koster, R. D. , B. M. Fekete, G. J. Huffman, and P. W. Stackhouse Jr. , Revisiting a hydrological analysis framework with International Satellite Land Surface Climatology Project Initiative 2 rainfall, net radiation, and runoff fields. * Betts, A. K. , M. Zhao, P. A. Dirmeyer, and A. C. M. Beljaars (2006), Comparison of ERA 40 and NCEP/DOE near-surface data sets with other ISLSCP-II data sets. *Guo, Z. , and P. A. Dirmeyer (2006), Evaluation of the Second Global Soil Wetness Project soil moisture simulations: 1. Intermodel comparison. *Guo, Z. , P. A. Dirmeyer, Z. -Z. Hu, X. Gao, and M. Zhao (2006), Evaluation of the Second Global Soil Wetness Project soil moisture simulations: 2. Sensitivity to external meteorological forcing. * Bounoua, L. , J. Masek, and Y. M. Tourre (2006), Sensitivity of surface climate to land surface parameters: A case study using the simple biosphere model Si. B 2. * Imhoff, M. L. , and L. Bounoua , Exploring global patterns of net primary production carbon supply and demand using satellite observations and statistical data. * Hall, F. , J. G. Masek, and G. J. Collatz (2006), Evaluation of ISLSCP Initiative II FASIR and GIMMS NDVI products and implications for carbon cycle science. • Brown de Colstoun, E. C. , R. S. De. Fries, and J. R. G. Townshend (2006), Evaluation of ISLSCP Initiative II satellite-based land cover data sets and assessment of progress in land cover data for global modeling, * Aires, F. , and C. Prigent (2006), Toward a new generation of satellite surface products?

ISLCP INITIATIVE II • SCIENCE QUESTIONS • ANALYSIS FRAMEWORK – WATER CYCLE – CARBON ISLCP INITIATIVE II • SCIENCE QUESTIONS • ANALYSIS FRAMEWORK – WATER CYCLE – CARBON CYCLE – INTERACTIONS • DATA SETS • EVALUATION • ISSUES Forrest Hall 05/07

GEWEX SCIENCE QUESTIONS • Phase II (2003 -2012) – Are the Earth's energy budget GEWEX SCIENCE QUESTIONS • Phase II (2003 -2012) – Are the Earth's energy budget and water cycle changing? – How do processes contribute to feedback and causes of natural variability? – Can we predict these changes on up to seasonal to interannual? – What are the impacts of these changes on water resources? Forrest Hall 05/07

GEWEX Phase II Objectives • Produce consistent research quality data sets complete with error GEWEX Phase II Objectives • Produce consistent research quality data sets complete with error descriptions of the Earth's energy budget and water cycle and their variability and trends on interannual to decadal time scales, and for use in climate system analysis and model development and validation • Enhance the understanding of how energy and water cycle processes function and quantify their contribution to climate feedbacks Forrest Hall 05/07

GEWEX ANALYSIS FRAMEWORK Are the Earth's energy budget and water cycle changing? P-E-S=R Got GEWEX ANALYSIS FRAMEWORK Are the Earth's energy budget and water cycle changing? P-E-S=R Got Records? Common Grid, Format, Land-sea Mask, QA’d Peer-Reviewed? Forrest Hall 05/07

SURFACE FLUX “Models” & “Algorithms” (5) g. C = m(An/Cs)F(e)P + go (6) Are SURFACE FLUX “Models” & “Algorithms” (5) g. C = m(An/Cs)F(e)P + go (6) Are these landscape element “models” applicable at the necessary spatial scales?

Carbon Uptake C Evapotranspiration E Absorbed long and short wave dissipated by evaporation and Carbon Uptake C Evapotranspiration E Absorbed long and short wave dissipated by evaporation and aerodynamic transport Stomatal Regulation Links ET and C gs = m(An/Cs)F(e)P + go An = F(Vmax, Fapar, Par, T, VPD) Forrest Hall 05/07

Initiative II Collection Forrest Hall 05/07 Initiative II Collection Forrest Hall 05/07

What Can We Measure? • TOP OF THE ATMOSPHERE RADIANCE • THE REST IS What Can We Measure? • TOP OF THE ATMOSPHERE RADIANCE • THE REST IS INFERRED USING “ALGORITHMS” SURFACE ALBEDO SURFACE TEMPERATURE TS FAPAR FROM NDVI (RELATED TO An) PROFILE WATER VAPOR AND TEMPERATURE AEROSOL OPTICAL DEPTH SHORT WAVE DOWN SOIL MOISTURE? FREEZE/THAW? VEGETATION STRUCTURE (LAI, FAPAR LAND COVER TYPE) Forrest Hall 05/07

Surface Energy Budget Net Radiation Absorbed = Net Short Wave {Sw (1 - )} Surface Energy Budget Net Radiation Absorbed = Net Short Wave {Sw (1 - )} + Net Long Wave {Lw - Lw } Lw Lw Sw Remote Sensing Inputs Atmospheric Aerosols and Clouds, Surface Albedo Net Short Wave {Sw (1 - r)} Atmospheric Water Vapor, Temperature Profiles, Cover Type ( ) Forrest Hall 05/07 Net Long Wave {Lw - Lw }

Fife and Boreas, ISLSCP II Results on LW, PAR etc. • PAR can be Fife and Boreas, ISLSCP II Results on LW, PAR etc. • PAR can be inferred to an accuracy of 8. 2 Wm– 2 • Solar insolation to an accuracy of 21. 6 Wm– 2 • Surface albedo to about 3% absolute, ~15% relative • • Downwelling longwave radiation to about 20 Wm– 2 • Net radiation to roughly 50 Wm– 2 Forrest Hall 05/07

Sensible Heat Radiative Temperature ≠Convective Temperature Forrest Hall 05/07 Sensible Heat Radiative Temperature ≠Convective Temperature Forrest Hall 05/07

Surface Heat and Mass Budget Net Radiation Absorbed = Rn = Latent Heat (LE) Surface Heat and Mass Budget Net Radiation Absorbed = Rn = Latent Heat (LE) + Sensible Heat (H) + Ground Heat Flux (G) LE H Remote Sensing Inputs G Fpar, Cover Type (C 3 or C 4), LAI Latent Heat (LE) Canopy Temperature, Air Temperature, Roughness Length, Sensible Heat (H) Snow Cover, Soil Moisture Content and State Forrest Hall 05/07 Ground Heat Flux (G)

“MODEL” PERFORMANCE HOMOGENOUS LANDSCAPE ELEMENTS LOCAL KNOWLEDGE OF PARAMETERS Forrest Hall 05/07 “MODEL” PERFORMANCE HOMOGENOUS LANDSCAPE ELEMENTS LOCAL KNOWLEDGE OF PARAMETERS Forrest Hall 05/07

Si. B 2 Bats 2 Flux Model Forrest Hall 05/07 Si. B 2 Bats 2 Flux Model Forrest Hall 05/07

Si. B 2 Performance Grasslands (FIFE) COLELLO et al, 1998 AMS Si. B 2 Performance Grasslands (FIFE) COLELLO et al, 1998 AMS

Si. B 2, BATS 2 Brazil Fig. 8. Monthly average latent heat flux calculated Si. B 2, BATS 2 Brazil Fig. 8. Monthly average latent heat flux calculated by BATS 2 using option-1 parameters, BATS and Si. B 2 for the Ji-Parana site over the 26 months for which meteorological forcing data are available. Sen et al, 2000 J. Hydromet Forrest Hall 05/07 Fig. 9. Comparison between hourly average latent heat fluxes calculated by BATS 2 relative to those calculated by (a) Si. B 2 and (b) BATS from the meteorological data available at the Ji. Parana site during 1993. (c) Hourly average latent heat flux on five selected days.

Remote Sensing Inputs Landscape scale Forrest Hall 05/07 Remote Sensing Inputs Landscape scale Forrest Hall 05/07

NDVI, Fapar, LAI • NDVI = (NIR-VIS)/ (NIR+VIS) – SURFACE NIR ACCURACY ± 4% NDVI, Fapar, LAI • NDVI = (NIR-VIS)/ (NIR+VIS) – SURFACE NIR ACCURACY ± 4% ABS (10% RELATIVE) – SURFACE VIS ACCURACY ± 1% ABS (30 PERCENT RELATIVE) • NDVI TO ± 5% (0. 05 OUT OF 1) • GLOBAL NDVI ANOMALY OF 0. 01 RESULTS IN MODELED GLOBAL CARBON FLUX ANOMALY OF 2. 0 Pg/yr • LAI BELOW SATURATION VALUE OF ~3, ± 20% Forrest Hall 05/07

LANDCOVER Figure 4. Per-cell agreement of IGBP-DIScover and. UMD land cover products at several LANDCOVER Figure 4. Per-cell agreement of IGBP-DIScover and. UMD land cover products at several spatial resolutions and using either a dominant or modified dominant aggregation scheme. Results of a comparison including the IGBP-DIScover permanent ice category against the UMD bare category are also shown. The agreement value is 48%. Forrest Hall 05/07

LANDCOVER Figure 2. UMD vegetation continuous fields product at 1/4 spatial resolution. The % LANDCOVER Figure 2. UMD vegetation continuous fields product at 1/4 spatial resolution. The % bare, herbaceous, and woody cover for each cell has been coded as red, green, and blue, respectively, to create this global representation. Other vegetation continuous fields provided in ISLSCP Initiative II include leaf type (needleleaf, broadleaf) and longevity (deciduous, broadleaf) for tree cover. BROWN DE COLSTOUN ET AL. : EVALUATION OF ISLSCP II LAND COVER DATA Forrest Hall 05/07

Landsat Biophys Deciduous Average Crown Green Leaf Area, 1999 Landsat Biophys Difference in Deciduous Landsat Biophys Deciduous Average Crown Green Leaf Area, 1999 Landsat Biophys Difference in Deciduous Average Crown Green Leaf Area, 1999 -1988 Forrest Hall 05/07

Surface Carbon Budget Net Ecosystem Exchange = NEE = Gross Primary Production (GPP) Autotrophic Surface Carbon Budget Net Ecosystem Exchange = NEE = Gross Primary Production (GPP) Autotrophic Respiration (Ra) - Heterotrophic Respiration (Rh) Ra GPP Rh NEE Remote Sensing Inputs LAI, Fpar, Cover Type, Soil Moisture, Temperature Gross Primary Production (GPP) Biomass, Cover Type, Temperature, Autotrophic Respiration (Ra) Soil Moisture, Soil Temperature, Snow. Forrest Hall 05/07 Cover Heterotrophic Respiration (Rh)

SURFACE CARBON BUDGET • NEE = GPP - Rh - Ra – Monteith Model SURFACE CARBON BUDGET • NEE = GPP - Rh - Ra – Monteith Model • GPP = ∫t par fapar dt • Rh= f (temp, precip, GPP) – Coupled NEE Models • Si. B 2, Si. B 3 • BATS 2 – An = F(Vmax, Fapar, Par, T, VPD) – NPP = F(GPP, Ra) – NEE = NPP-Rh • Fapar = a. NDVI+b, a&b are cover type dependent Forrest Hall 05/07

THE AHVRR NDVI DATA RECORD • • Different calibrations for each AVHRR sensor. • THE AHVRR NDVI DATA RECORD • • Different calibrations for each AVHRR sensor. • Aerosol variability from El Chichon 1982, Pinatubo 1991, biomass burning, dust etc. • Sub-pixel cloud contamination • NOAA 11 Equatorial crossing time allowed to drift within each series, steadily increasing solar zenith angle. • NOAA 7 NOAA 9 1982 -1998 SPANNED BY NOAA SATELLITES 7, 9, 11, 14. Max NDVI Compositing: FASIR: middle 9 day interval. GIMMS: average both 15 day intervals. NOAA 14 Forrest Hall 05/07

GIMMS/FASIR Algorithms Side by Side GIMMS: Empirical BRF correction, no atmospheric correction, AVHRR NDVI GIMMS/FASIR Algorithms Side by Side GIMMS: Empirical BRF correction, no atmospheric correction, AVHRR NDVI “adjusted” to match SPOT Vegetation NDVI. FASIR: Vicarious cal. , BRF correction, no tropospheric aerosol correction.

GIMMS-FASIR NDVI REGIONAL AND SEASONAL Forrest Hall 05/07 GIMMS-FASIR NDVI REGIONAL AND SEASONAL Forrest Hall 05/07

COMPARISON TO LANDSAT SURFACE REFLECTANCE Forrest Hall 05/07 COMPARISON TO LANDSAT SURFACE REFLECTANCE Forrest Hall 05/07

GIMMS, FASIR MONTHLY NDVI & Fapar ANOMALY COMPARISON Forrest Hall 05/07 GIMMS, FASIR MONTHLY NDVI & Fapar ANOMALY COMPARISON Forrest Hall 05/07

GIMMS VS FASIR • GIMMS AND FASIR NDVI, FPAR AND ANOMALY RECORDS DIFFER SUBSTANTIALLY GIMMS VS FASIR • GIMMS AND FASIR NDVI, FPAR AND ANOMALY RECORDS DIFFER SUBSTANTIALLY IN 1984, 1986 AND 1994. • BOTH SHOW POSITIVE TRENDS IN THEIR NDVI AND Fapar ANOMALIES. • CAN WE TELL WHICH ONE IS BETTER? Forrest Hall 05/07

Annual Fapar NPP, Rh Anomalies CASA/FASIR CASA/GIMMS Forrest Hall 05/07 Annual Fapar NPP, Rh Anomalies CASA/FASIR CASA/GIMMS Forrest Hall 05/07

CASA, FASIR & GIMMS Global Net Ecosystem Exchange (NEE); GIMMS FASIR “Top-Down” El Chichon CASA, FASIR & GIMMS Global Net Ecosystem Exchange (NEE); GIMMS FASIR “Top-Down” El Chichon El Niño La Niña Pinatubo El Niño Forrest Hall 05/07 El Niño Strong El Niño followed by La Niña

CONCLUSIONS • FASIR-GIMMS NDVI DIFFER – From each other (1984, 1986, 1994) – From CONCLUSIONS • FASIR-GIMMS NDVI DIFFER – From each other (1984, 1986, 1994) – From Landsat surface reflectance – NDVI/Fapar differences not mitigated sufficiently by fapar-NDVI scaling or use of anomalies. • Fapar differences are significant in terms of CASA estimates of NEE. – GLOBAL NDVI ANOMALY OF 0. 01 DRIVES CASA CARBON ANOMALY OF 2. 0 Pg/yr. • CASA NPP driven mainly by NDVI, Fapar variations. Other variations are important. – Fire emissions – Diffuse/Direct PAR Forrest Hall 05/07

Future AVHRR/MODIS/NPP Climate Data Records • • Need seamless join of AVHRR/MODIS/VIIRS data records. Future AVHRR/MODIS/NPP Climate Data Records • • Need seamless join of AVHRR/MODIS/VIIRS data records. – MODIS has significantly different band passes than AVHRR – Vegetation visible and nir reflectance differ between MODIS, AVHRR – Need to seamlessly join these two data records, eventually to VIIRS Need to complete the AVHRR band 1, band 2, NDVI surface reflectance record. – Aerosol correction in band 1, band 2 – Water vapor correction in band 2 – Investigate the BRF correction. • • Need to focus AVHRR, MODIS algorithms on retrieval of biophysical parameters (vegetation structure, optical properties) required by surface exchange models, rather than ambiguously defined land-cover categories. Future sensors must be extremely well calibrated, with bands sufficient to completely deal with atmospheric effects. Forrest Hall 05/07

What Is Missing? • Global Access to Tower Flux Data and Associated Biophysical Parameters What Is Missing? • Global Access to Tower Flux Data and Associated Biophysical Parameters • VEGETATION STRUCTURE – MODIS BIOPHYSICAL RETRIEVALS – FROM LIDAR, RADAR • SOIL MOISTURE FROM MICROWAVE • DIRECT RETRIEVAL OF VEGETATION LIGHT USE EFFICIENCY Forrest Hall 05/07

LIDAR/RADAR FOR VEG 3 D STRUCTURE AND BIOMASS RADAR/In. SAR VOLUME SCATTERING 100 m LIDAR/RADAR FOR VEG 3 D STRUCTURE AND BIOMASS RADAR/In. SAR VOLUME SCATTERING 100 m LIDAR POINT PROFILES 25 m

Photosynthetic Light Use Efficiency Forrest Hall 05/07 Photosynthetic Light Use Efficiency Forrest Hall 05/07

Light Use Efficiency Forward Scatter Bsck Scatter Forrest Hall 05/07 Light Use Efficiency Forward Scatter Bsck Scatter Forrest Hall 05/07

MODIS LIGHT USE EFFICIENCY Forrest Hall 05/07 MODIS LIGHT USE EFFICIENCY Forrest Hall 05/07