CERES-derived Surface radiative flux consistent with A-train observations

Top-of-atmosphere radiative fluxes derived from CERES instruments on Aqua and Terra provide observations to understand earth radiation budget and its variability. Using cloud and aerosol properties derived from MODIS combined with atmospheric thermodynamical properties from reanalysis, surface radiative fluxes are also estimated in the CERES project. Observations from the A-train constellation, such as CALIPSO, CloudSat, and AIRS, are helping to improve the estimate of surface radiative fluxes. For example, the ability of CALIPSO and CloudSat to detect clouds and cloud base height better than MODIS contributes to improve the surface downward longwave flux. The global annual mean cloud fraction derived from CALIPSO and CloudSat is larger them the cloud fraction derived from MODIS by 0.11 (0.04 excluding clouds with optical thickness less than 0.3). The global annual mean cloud base height derived from CALIPSO and CloudSat is lower than that derived from MODIS by an empirical relationship by 1.6 km. As a consequence, the global annual mean surface downward longwave flux increases by 3.4 Wm-2 compared to the value computed with MODIS only. The regional difference is even larger. The increase of the surface downward longwave irradiance in the Arctic exceeds 10 Wm-2 (~4%) in winter because during polar night CALIPSO and CloudSat detect more clouds compared to the cloud amount derived by the CERES cloud algorithm. With these better cloud properties derived from CALIPSO and CloudSat, and temperature and humidity profiles derived from AIRS, the systematic error and uncertainty of inputs used in surface radiative flux computations in CERES processes can be quantified. The systematic error and uncertainty estimates combined with the CERES TOA EBAF product are used to derive improved surface radiative fluxes that are consistent with A-train observations. In this presentation, we will explain how the improved gridded monthly mean surface radiative fluxes (CERES surface EBAF product) are obtained. We will also present an analysis using improved surface radiative fluxes, such as how surface net radiative flux spatially and temporally correlates with TOA net flux and how surface net and atmospheric shortwave and longwave fluxes changes with the ENSO index.