The CERES calibration strategy of the geostationary visible channels for CERES cloud and flux products

Abstract

The Clouds and Earth's Radiant Energy System (CERES) project has greatly improved the understanding of the role of clouds and energy cycles in global climate studies. CERES flux and cloud properties rely on not only CERES broadband fluxes and MODIS cloud properties but also on operational geostationary (GOES, METEOSAT, MTSAT) derived fluxes and clouds, which are acquired between CERES measurements such to properly account for the diurnal cycle. The high quality of the CERES products is dependent on a consistent radiometric calibration of the un-calibrated geostationary visible sensors and MODIS. To achieve this consistency, the calibration of a reference sensor must be transferred to the other instruments. Historically, Terra-MODIS and Aqua-MODIS, both of which employ solar diffusers, have been regarded as having a well-calibrated visible channel (650 nm). Recent analysis has revealed the Aqua-MODIS instrument to be more stable than the MODIS instrument onboard the Terra satellite. For this reason, Aqua-MODIS has been chosen as the reference sensor whereas Terra-MODIS adjustments can be used to put the instrument on the same radiometric scale as Aqua-MODIS. The ray-matching technique is used to transfer the calibration of the well-calibrated MODIS instrument to the un-calibrated GEO sensors. Additionally, empirically derived models for pseudo-invariant test sites and deep convective clouds (DCC) have been developed and applied for monitoring and validating the GEO calibration. Multiple pseudo-invariant test site and DCC absolute calibration methodologies are compared. Latest results show that GOES-13 response has drifted 5-6 percent in its first 15 months of operation. The Aqua/Terra-MODIS crosscalibration trends are in agreement with calibration trends obtained from pseudo-invariant test sites and DCC. These results are in preparation for CERES Edition4 products, which will include updated geostationary calibration coefficients and cloud retrieval improvements.