Using Coincident Satellites to Automate Atmospheric Correction for Nanosatellite Imagery

Nanosatellites represent new platforms that can be exploited to improve spatial and temporal coverage of coastal ocean color imagery. Challenges remain, however. The nanosatellites have a very limited bandset, which is a particular issue for atmospheric correction. This study uses coincident satellite imagery (covering the same geographical location and time as the nanosatellite data) from the Visible Infrared Imaging Radiometer Suite (VIIRS), which is equipped with two near infrared bands, to build an algorithm to atmospherically correct nanosatellite imagery at each of the nanosatellite’s visible wavelengths. These nanosatellites have a single infrared band, although two such bands are typically required to automatically select an appropriate aerosol model during the atmospheric correction processing stage, prior to estimating water-leaving radiance (Lw), which is the basis for the generation of remote sensing reflectance and other inherent and apparent optical property products. Multiple linear regression is used to determine the relationships between the top-of-atmosphere radiance (Lt), Rayleigh scattering (Lr), and normalized Lw (nLw) from VIIRS and applies those relationships to nanosatellite imagery to derive nLw when only Lt and Lr are available. This study assesses the accuracy for using coincident satellite imagery to aid in the atmospheric correction of nanosatellite imagery, as well as addressing any shortcomings.