Updated atmospheric correction scheme for the Geostationary Ocean Color Imager (GOCI)

Abstract

This study describes the updated atmospheric correction algorithm that aims to retrieve the remote-sensing reflectance (Rrs) from the total signal recorded by the Geostationary Ocean Color Imager (GOCI) at the top-of-the-atmosphere (TOA). Here, we first describe a gaseous absorption correction model that is related to the atmospheric transmittance. Then we present a new aerosol correction scheme that is methodologically more straightforward.For the gaseous absorption, we found water vapor absorption significantly effects on 660, 745, and 865 nm band of the GOCI for large zenith angles. The correction model is developed and verified using a radiative transfer simulation (Vermote et al., 2006). For the aerosol reflectance correction, we suggest a more direct aerosol correction scheme compared to previous methods. This scheme is based on the AOT-independent spectral relationships between the aerosol multiple-scattering reflectance at different wavelengths and directly determines the two most appropriate aerosol models and their mixing factor with no residual errors in the NIR. The vicarious gains and the near infrared band correction model for turbid waters are adjusted following the atmospheric correction scheme updates. Notably, the updated calibration gains are within the ranges of onboard calibration uncertainties, 3.8%.To assess the performance of the updated algorithm relative to the errors in the Rrs retrievaere, we first describe a gaseous absorption correction model that is related to the atmospheric transmittance. Then we present a new aerosol correction scheme that is methodologically more straightforward.For the gaseous absorption, we found water vapor absorption significantly effects on 660, 745, and 865 nm band of the GOCI for large zenith angles. The correction model is developed and verified using a radiative transfer simulation (Vermote et al., 2006). For the aerosol reflectance correction, we suggest a more direct aerosol correction scheme compared to previous methods. This scheme is based on the AOT-independent spectral relationships between the aerosol multiple-scattering reflectance at different wavelengths and directly determines the two most appropriate aerosol models and their mixing factor with no residual errors in the NIR. The vicarious gains and the near infrared band correction model for turbid waters are adjusted following the atmospheric correction scheme updates. Notably, the updated calibration gains are within the ranges of onboard calibration uncertainties, 3.8%.To assess the performance of the updated algorithm relative to the errors in the Rrs retrieva