|
Tropical deep convective clouds (DCC) are proven to be an excellent Earth invariant target for post-launch radiometric assessment of satellite imagers. The success of the DCC technique (DCCT) relies on a large ensemble of identified DCC pixels that are collectively analyzed as a stable reflectance reference. The near-Lambertian reflectance of DCC, as well as their high signal-to-noise ratio, location near the tropopause above most of the atmospheric water vapor and aerosol, and availability across the globe make them an ideal target for radiometric scaling of geostationary (GEO) and low-earth orbiting (LEO) satellites. The DCCT has been successfully applied to calibrate reflective solar bands (with wavelengths < 1 µm) of numerous GEO and LEO imagers. The DCC reflectivity in VIS-NIR is mainly a function of cloud optical depth, and the DCCT provides a stable monthly statistical mode that can be tracked over time for monitoring the radiometric stability of the sensor. However, at shortwave infrared (SWIR) wavelengths, the DCC reflectance is affected by both cloud particle size and cloud optical depth. The DCCT for SWIR bands is found to be most sensitive to the BRDF and brightness temperature, resulting in large seasonal cycles of DCC reflectivity that make the implementation of the DCCT more challenging. The key to improving the DCCT at SWIR wavelengths is proper characterization of the DCC reflectance as a function of viewing and solar angular conditions. This paper presents channel-specific seasonal BRDF models for SWIR bands based on five years of SNPP-VIIRS DCC measurements. The seasonal BRDFs are effective in reducing the temporal variability of the DCC time series by up to 65% when applied to both Aqua-MODIS and NPP-VIIRS SWIR bands. The use of seasonal BRDF models for radiometric stability assessment and absolute inter-calibration of the MODIS and SNPPVIIRS SWIR bands is discussed in the paper. In addition, the modification of the baseline DCCT for daily monitoring of the radiometric stability of the GEO imager L1B radiances is also illustrated. The DCCT is capable of detecting daily GOeS-16 L1B radiance anomalies with a magnitude greater than ±3% for bands 2 and 3, and ±4% shift for band 1 with 3σ significance.
|
