|
The NASA Clouds and the Earth’s Radiant Energy System (CERES) Energy Balanced and Filled (EBAF) product provides the scientific community observed TOA SW and LW fluxes for climate monitoring and climate model validation. To provide continuity for cloud retrievals between MODIS and VIIRS, the CERES project inter-calibrates MODIS and VIIRS utilizing coincident ray-matched radiance pairs over all-sky tropical ocean. The Aqua and Terra satellites have started drifting, thereby preventing any coincident tropical MODIS and VIIRS inter-calibration events. Similarly, no simultaneous nadir overpasses (SNOs) will exist between SNPP and NOAA satellites, which will fly in the same 1:30 PM orbit but positioned a half an orbit apart. The CERES project will utilize the Libya-4 and Dome-C invariant targets to radiometrically scale between MODIS and VIIRS reflective solar bands. This study has advanced the Libya-4 and Dome-C characterization by considering all angular conditions, improving clear-sky identification and atmospheric corrections. The BRDF for each VZA and RAA stratified angular bin is approximated using a 2nd order regression with respect to cos(SZA). The clear-sky filtering utilized spatial homogeneity thresholds applied to channels not impacted by atmospheric parameters and additional angular bin specific dynamic filtering. Multiple sources of PW, ozone and aerosol optical depth are considered. The improved atmospheric characterization is evaluated by comparing the trendSE consistency across channels. For Libya-4 the 2.2μm and 0.91μm strong water vapor absorption bands, the trendSE was reduced by ~60% and ~80% by including the PW term. The Libya- 4 trendSE with atmospheric correction was reduced from within 2% to 1% for all channels except MODIS B17. The Dome-C 0.55mm and 0.65μm band trendSE was reduced by between 40% to 60% after accounting for ozone absorption. The Dome-C imager channel trendSE was reduced from within 2% to 1% by including atmospheric corrections. The Dome-C post-solstice ozone and PW daily variations are much smaller than prior to solstice. The Dome-C resulting post-solstice imager channel trendSE was reduced from 1% to 0.85% by including atmospheric corrections and closer to the 0.48μm band trendSE of 0.6%, which was not impacted by the atmosphere. Smaller trendSE can be realized by limiting the large VZA observations over Libya-4 as well as utilizing only post-solstice observations over Dome-C.
|
