HR: 0800h
AN: B31B-0210    [Abstracts]
TI: Sensitivity Of MODIS Global Terrestrial Primary Production To The Accuracy Of Meteorological Data Sets
AU: * Zhao, M
EM: zhao@ntsg.umt.edu
AF: Numerical Terradynamic Simulation Group, Department of Ecosystem and Conservation Sciences, University of Montana, Missoula, MT 59812 United States
AU: Nemani, R R
EM: rama.nemani@nasa.gov
AF: Earth Science Division,Ames Research Center, Moffett Field, Moffett Field, CA 94035 United States
AU: Running, S W
EM: swr@ntsg.umt.edu
AF: Numerical Terradynamic Simulation Group, Department of Ecosystem and Conservation Sciences, University of Montana, Missoula, MT 59812 United States
AB: Abstract: MODIS provides a dramatic improvement in our ability to accurately monitor global terrestrial primary production. The global weekly terrestrial primary production (MOD17) is significant for scientific research and natural resource management. The near real-time MOD17 requires regularly daily gridded assimilation meteorological data set as input, and the accuracy of this meteorological data set shows marked difference in their spatial and temporal distribution. This study compares five global surface meteorological data sets - Data Assimilation Office (DAO) GEOS402, NCEP-NCAR reanalysis 1(NCEP), ECMWF 40 (ERA-40) years reanalysis, Climate Research Unit of University of East Anglia (CRU) and NASA Surface meteorology and Solar Energy (SSE) at vegetated land area - to assess the sensitivity of MODIS global terrestrial primary production to the uncertainties of meteorological inputs. Compared with SSE and CRU, NCEP tends to overestimate surface solar radiation, underestimate temperature and Vapor Pressure Deficit (VPD), and SRA-40 is closer to SSE and CRU, but its radiation tend to lower in tropical region. DAO is closer to ERA-40, which suggests its magnetite may be closer to real value. Global daily observations from WMO weather stations also are used to directly compare with DAO, NCEP and ERA-40. The larger discrepancies among different meteorological data sets occur in low latitude region. Global total GPP and NPP driven by DAO, NCEP and ERA-40 show large differences, and spatially, large difference occurs in tropical region due to higher production, larger vegetated areas and higher uncertainties in meteorological data sets in tropics. MOD17 driven by DAO, NCEP and ECMWF also are compared with it driven by observed surface meteorological data set with solar radiation from over 300 weather stations across USA, respectively. The results demonstrate the need for improving the accuracy of global surface meteorological data set, especially for tropical regions, to better understand and predict global carbon cycle under global climate change.
UR: http://www.ntsg.umt.edu
DE: 1615 Biogeochemical processes (4805)
DE: 1640 Remote sensing
DE: 0315 Biosphere/atmosphere interactions
DE: 0400 Biogeosciences
SC: Biogeosciences [B]
MN: 2004 AGU Fall Meeting