A NEW MULTIPLE-SCATTERING ALGORITHM FOR THE GOCI ATMOSPHERIC CORRECTION

Abstract

An atmospheric correction algorithm is designed to retrieve the remote-sensing reflectance (Rrs) or normalized water-leaving radiance (nLw) from the total signal observed at the top-of-the-atmosphere (TOA) level. We suggest a more straightforward algorithm in terms of aerosol model selection compare to the traditional algorithm using single-scattering reflectance ratio used with the Sea-viewing Wide Field-of-view Sensor (SeaWiFS), the Moderate Resolution Imaging Spectroradiometer (MODIS), the Visible Infrared Imaging Radiometer Suite (VIIRS), the Global Imager (GLI), and the Geostationary Ocean Color Imager (GOCI). The new atmospheric correction scheme estimates the aerosol reflectance in visible bands from candidate aerosol models using multiple-scattering reflectance ratio (multiple scattering epsilon) between two near-infrared (NIR) bands recorded by the satellite observation. We established spectral relationships between multiple-scattering aerosol reflectances at several wavelengths with polynomial functions while an aerosol optical thickness changes. In the process, the two optimum aerosol models are selected by comparing the observed aerosol reflectance ratio in NIR to theoretical reflectance ratio of candidate models. A weighting factor for the two contributed aerosol models can be computed by solving the quadratic equation. Then the aerosol reflectance at the NIR is extrapolated to the visible with the deterorward algorithm in terms of aerosol model selection compare to the traditional algorithm using single-scattering reflectance ratio used with the Sea-viewing Wide Field-of-view Sensor (SeaWiFS), the Moderate Resolution Imaging Spectroradiometer (MODIS), the Visible Infrared Imaging Radiometer Suite (VIIRS), the Global Imager (GLI), and the Geostationary Ocean Color Imager (GOCI). The new atmospheric correction scheme estimates the aerosol reflectance in visible bands from candidate aerosol models using multiple-scattering reflectance ratio (multiple scattering epsilon) between two near-infrared (NIR) bands recorded by the satellite observation. We established spectral relationships between multiple-scattering aerosol reflectances at several wavelengths with polynomial functions while an aerosol optical thickness changes. In the process, the two optimum aerosol models are selected by comparing the observed aerosol reflectance ratio in NIR to theoretical reflectance ratio of candidate models. A weighting factor for the two contributed aerosol models can be computed by solving the quadratic equation. Then the aerosol reflectance at the NIR is extrapolated to the visible with the deter