HR: 0800h
AN: B31B-0211    [Abstracts]
TI: Prognostic/Diagnostic Analysis of Ecosystem Processes: Comparison of MODIS and Biome-BGC Based Land Surface Products Over United States From 2000 to 2003
AU: * Ichii, K
EM: kazu@ntsg.umt.edu
AF: San Jose State University, One Washinton Square, San Jose, CA 95192 United States
AU: Votava, P
EM: pvotava@mail.arc.nasa.gov
AF: California Stete University, Moterey Bay, 100 Campus Center, Seaside, CA 93955 United States
AU: Votava, P
EM: pvotava@mail.arc.nasa.gov
AF: NASA Ames Reserch Center, Moffet, Field, CA 94035 United States
AU: Michaelis, A
EM: amichaelis@mail.arc.nasa.gov
AF: University of Montana, 32 Campus Drive, Missoula, MT 59812 United States
AU: Michaelis, A
EM: amichaelis@mail.arc.nasa.gov
AF: NASA Ames Reserch Center, Moffet, Field, CA 94035 United States
AU: Nemani, R
EM: rama.nemani@nasa.gov
AF: NASA Ames Reserch Center, Moffet, Field, CA 94035 United States
AB: Terrestrial ecosystem models can simulate terrestrial carbon, water, energy cycles, and have a potential to forecast future status of them. However, we need to conduct validation and comparison studies with observed data for improving model capabilities and performances. Four-years of MODIS products, such as snow cover, LAI, and GPP provide the opportunity for validating the terrestrial ecosystem models. In this study, we used Biome-BGC terrestrial ecosystem model and MODIS based land surface products to evaluate the model performance. We analyzed spatial patterns and interannual variations over continental United States from model-based and satellite-based products. We focused on GPP, LAI and snow cover over US from 2000 to 2003. Spatial patterns (4 years average) of GPP, LAI, and Snow Cover (number of days covered by snow in a year) are consistent between MODIS and Biome-BGC. Although magnitudes are different (Biome-BGC underestimates LAI, and overestimates GPP), spatial patterns of GPP and LAI are consistent in forest ecosystems. However, Biome-BGC estimates very low GPP and LAI in grassland ecosystems, and we need to improve them. Snow cover shows very similar spatial pattern with a slight overestimation by Biome-BGC. Seasonal and interannual variations in GPP show very similar patterns except for grassland ecosystems. Both methodologies show relatively low GPP in 2002, which corresponds with low precipitation year. Seasonal and interannual variations in Biome-BGC based LAI show smaller amplitudes, however, both of them show similar year to year variations including negative anomaly in 2002. Snow cover from Biome-BGC and MODIS show very good agreement in terms of both seasonal and interannual variations. Overall, prognostic and diagnostic analyses show consistent results, suggesting capabilities of future prediction.
UR: http://ecocast.arc.nasa.gov
DE: 1615 Biogeochemical processes (4805)
DE: 1640 Remote sensing
SC: Biogeosciences [B]
MN: 2004 AGU Fall Meeting