An improved algorithm for VIIRS band-to-band registration characterization with lunar observation

Abstract

Since launch, lunar observations have been performed on a regular basis for the Visible Infrared Imaging Radiometer Suite (VIIRS) instrument aboard the Suomi National Polar-orbiting Partnership (S-NPP) satellite. Unlike its predecessor Moderate Resolution Imaging Spectro-radiometer (MODIS), VIIRS has no on-board calibrator to perform its on-orbit spatial characterization. Therefore, the methodologies for spatial characterization using the Moon that have been developed for MODIS are extended for VIIRS. One of the key spatial parameters is the band-to-band registration (BBR) in both along-scan and along-track directions. While the lunar BBR results can satisfactorily demonstrate that the longterm stability of VIIRS BBR meets the design requirement specification of ±0.1 M-band pixels, seasonal oscillations have been observed in BBR trending, mostly noticeable between the visible/near infrared bands and short/middle wave infrared bands. This paper investigates the cause of this oscillation and proposes a new algorithm to reduce it. It is identified that the oscillation is correlated with the orientations of the lunar images that rotate from event to event. After figuring out an approach to quantify the rotation and compensate its impact, current BBR algorithm is improved and the seasonal oscillation in the BBR results is significantly reduced from up to ±0.05 M-band pixels to less than ±0.01 Mband pixels. With suppressed seasonal oscillation, the actual BBR including its long-term drift can be determined more accurately.