A dynamic scaling algorithm for the optimized digital display of VIIRS Day/Night Band imagery

Abstract

The VIIRS Day/Night Band (DNB) is a visible and near-infrared sensor that is sensitive to a broad range of light intensities ranging from daylight down to airglow at night. The on-board calibration of the DNB allows for the quantification of radiance values over the full range of the instrument’s sensitivity, unlike the heritage Operational Linescan System (OLS). For scenes that span the day/night terminator, observed DNB radiance values may vary by up to eight orders of magnitude. Consequently, it is impractical to display the full range of radiance values in a single digital image. In this work, an algorithm is presented that scales the observed radiance values between expected maximum and minimum values that are a function of solar and lunar zenith angles as well as the fraction of the lunar disc that is illuminated by the Sun. This dynamic scaling algorithm preserves scene contrast over the full range of solar and lunar illumination conditions, similar to the Near Constant Contrast (NCC) imagery product. Unlike the NCC algorithm, however, the ‘erf-dynamic scaling’ algorithm (so-called for its likeness to the Gauss error function) presented here requires no ancillary information outside of what is included in the DNB data distributed according to the Joint Polar Satellite System (JPSS) program file standards. Results indicate that this algorithm has improved performance over simple methods for displaying DNB imagery and, in some instances, may exceed the performance of the NCC product itself. This algorithm is expected to replace many ad hoc methods of displaying DNB imagery and may serve as a substitute for operational users that do not have access to the NCC product.