Suomi NPP VIIRS Detector Dependent Relative Spectral Response Variation Study using Line-by-Line Radiative Transfer Model Calculations

Modern focal plane array based satellite radiometers have many detectors with slightly different relative spectral response (RSR). Effect of RSR differences on imaginary artifacts, as well as geophysical retrieval uncertainties has not been well studied. Some previous studies used MODTRAN atmospheric radiative transfer model (RTM) for detector-level radiance simulations. However, it is limited by the spectral resolution of the model relative to the narrow spectral bandwidth of the detectors. This study evaluates detector level RSR using LBLRTM, which is recognized as the model of best choice with the highest spectral resolution achievable. In this study, we use the LBLRTM at spectral resolution of 0.01 cm-1 for VIIRS longwave infrared bands (M15 and M16) under different atmospheric conditions. We found that the atmospheric effects are significant for the RSR variations between detectors. Besides the difference between detector-level and band averaged RSR, the atmospheric condition is also a possible factor in causing the sea surface temperature (SST) striping issue as reported by the SST Environmental Data Record (EDR) team. A better understanding of this effect will also allow us to better estimate and reduce uncertainties in geophysical retrievals.