Validating CERES diurnally averaged fluxes with GERB

Dr. Slingo was a member of the international Geostationary Earth Radiation Budget (GERB) science team and has utilized GERB fluxes to study the diurnal cycle of TOA fluxes, climate variability and has compared them to evaluate model-derived fluxes. The Clouds and the Earth's Radiant Energy System (CERES) and GERB science team members have worked closely together to provide the radiation community climate quality fluxes. The GERB diurnal flux products provide 15-minute resolution TOA broadband fluxes over the Meteosat geostationary (GEO) domain. CERES only observes fluxes during Terra (10:30 AM local equator crossing time) and Aqua (1:30PM) satellite overpass events. In order for CERES observed fluxes to be useful they must be diurnally averaged into daily means. GERB is the only source of high temporal resolution observed TOA fluxes and is used by the CERES project to validate its temporal averaging algorithms. CERES utilizes two temporal averaging approaches. The first, assumes constant meteorology between CERES observations and is contained on the SSF1deg level-3 product. The second, incorporates GEO derived broadband fluxes to properly account for the regional diurnal cycle of fluxes. After the GEO fluxes are carefully normalized against CERES measurements and combined with the CERES measurements in the SYN1deg product. The monthly regional SW flux difference between the approaches can be as large as 25 Wm-2 over maritime stratus and land afternoon convective regions. However, globally the flux difference is small.

Incorporating GEO imager data is not without risk and could introduce artifacts into the GEO enhanced product. Both CERES diurnally averaged product fluxes are validated against GERB fluxes at monthly, daily and 3-hourly temporal resolutions. The GERB comparisons show that the regional diurnal cycle is captured in the GEO enhanced product. In addition, the GERB fluxes allow all steps of the temporal averaging algorithm to be evaluated in order to prioritize algorithm improvements.