ASSESSMENT OF THE GLOBAL MONTHLY MEAN SURFACE INSOLATION ESTIMATED FROM SATELLITE MEASUREMENTS USING GLOBAL ENERGY-BALANCE ARCHIVE DATA

Global datasets of surface radiation budget (SRB) have been obtained f rom satellite programs. These satellite-based estimates need validatio n with ground-truth observations. This study validates the estimates o f monthly mean surface insolation contained in two satellite-based SRB datasets with the surface measurements made at worldwide radiation st ations from the Global Energy Balance Archive (GEBA). One dataset was developed from the Earth Radiation Budget Experiment (ERBE) using the algorithm of Li et al. (ERBE/SRB), and the other from the Internationa l Satellite Cloud Climatology Project (ISCCP) using the algorithm of P inker and Laszlo and that of Staylor (GEWEX/SRB). Since the ERBE/SRB d ata contain the surface net solar radiation only, the values of surfac e insolation were derived by making use of the surface albedo data con tained in the GEWEX/SRB product. The resulting surface insolation has a bias error near zero and a root-mean-square error (RMSE) between 8 a nd 28 W m(-2). The RMSE is mainly associated with poor representation of surface observations within a grid cell. When the number of surface observations are sufficient, the random error is estimated to be abou t 5 W m(-2) with present satellite-based estimates. In addition to dem onstrating the strength of the retrieving method, the small random err or demonstrates how well the ERBE derives the monthly mean fluxes at t he top of the atmosphere (TOA). A larger scatter is found for the comp arison of transmissivity than for that of insolation. Month to month c omparison of insolation reveals a weak seasonal trend in bias error wi th an amplitude of about 3 W m(-2). As for the insolation data from th e GEWEX/SRB, larger bias errors of 5-10 W m(-2) are evident with stron ger seasonal trends and almost identical RMSEs.