Key points are not available for this paper at this time.
The Moderate Resolution Imaging Spectroradiometer (MODIS) on board NASA's satellites, Terra and Aqua, dramatically improves our ability to accurately and continuously monitor the terrestrial biosphere. MODIS information is used to estimate global terrestrial primary production weekly and annually in near‐real time at a 1‐km resolution. MODIS terrestrial primary production requires daily gridded assimilation meteorological data as inputs, and the accuracy of the existing meteorological reanalysis data sets show marked differences both spatially and temporally. This study compares surface meteorological data sets from three well‐documented global reanalyses, NASA Data Assimilation Office (DAO), European Centre for Medium‐Range Weather Forecasts (ECMWF) (ERA‐40) and National Centers for Environmental Prediction/National Center for Atmospheric Research (NCEP/NCAR) reanalysis 1, with observed weather station data and other gridded data interpolated from the observations, to evaluate the sensitivity of MODIS global terrestrial gross and net primary production (GPP and NPP) to the uncertainties of meteorological inputs both in the United States and the global vegetated areas. NCEP tends to overestimate surface solar radiation, and underestimate both temperature and vapor pressure deficit (VPD). ECMWF has the highest accuracy but its radiation is lower in tropical regions, and the accuracy of DAO lies between NCEP and ECMWF. Biases in temperature are mainly responsible for large VPD biases in reanalyses. MODIS NPP contains more uncertainties than GPP. Global total MODIS GPP and NPP driven by DAO, ECMWF, and NCEP show notable differences (>20 Pg C/yr) with the highest estimates from NCEP and the lowest from ECMWF. Again, the DAO results lie somewhere between NCEP and ECMWF estimates. Spatially, the larger discrepancies among reanalyses and their derived MODIS GPP and NPP occur in the tropics. These results reveal that the biases in meteorological reanalyses can introduce substantial error into GPP and NPP estimations, and emphasize the need to minimize these biases to improve the quality of MODIS GPP and NPP products.
Zhao et al. (Mon,) studied this question.