Comparison of satellite-derived TOA shortwave clear-sky fluxes to estimates from GCM simulations constrained by satellite observations of land surface characteristics



Clear-sky, upwelling shortwave flux at the top of the atmosphere equation image, simulated using the atmospheric and land model components of the Community Climate System Model 3 (CCSM3), is compared to corresponding observational estimates from the Clouds and Earth's Radiant Energy System (CERES) sensor. Improvements resulting from the use of land surface albedo derived from Moderate Resolution Imaging Spectroradiometer (MODIS) to constrain the simulations are also examined. Compared to CERES observations, CCSM3 overestimates global, annual averaged equation image over both land and oceans. However, regionally, CCSM3 overestimates equation image over some land and ocean areas while underestimating it over other sites. CCSM3 underestimates equation image over the Saharan and Arabian Deserts and substantial differences exist between CERES observations and CCSM3 over agricultural areas. Over selected sites, after using ground-based observations to remove systematic biases that exist in CCSM computation of equation image, it is found that use of MODIS albedo improves the simulation of equation image. Inability of coarse resolution CCSM3 simulation to resolve spatial heterogeneity of snowfall over high altitude sites such as the Tibetan Plateau causes overestimation of equation image in these areas. Discrepancies also exist in the simulation of equation image over ocean areas as CCSM3 does not account for the effect of wind speed on ocean surface albedo. This study shows that the radiative energy budget at the TOA is improved through the use of MODIS albedo in Global Climate Models. Copyright © 2010 Royal Meteorological Society