Application of GOCI-II Quality Assurance Algorithm and Verification of Improvement of Atmospheric Correction

Abstract

To derive remote-sensing reflectance (Rrs) from the Geostationary Ocean Color Imager-II (GOCI-II), it is vital to employ an atmospheric correction algorithm based on the radiative transfer model. The atmospheric correction is crucial for estimating a variety of ocean color products. In a clear ocean, the atmospheric path radiance can significantly exceed the water-leaving radiance in the blue band. That means Rrs is sensitive to the atmospheric correction process. A Quality Assurance (QA) algorithm is introduced, initially grounded in in-situ Rrs data archived within the SeaBASS database, and adapted to suit the spectral characteristics of GOCI-II. The QA algorithm assigns quality analysis scores to Rrs, ranging from 0 to 1, and categorizes water types into 23 groups. When applied to preliminary GOCI-II data with limited calibration, the algorithm produces lower scores (0.625). For improved GOCI-II atmospheric correction results with additional calibrations, it generates higher scores (0.875). The analysis of water types using the QA algorithm reveals that specific regions of the East Sea, South Sea, and the Northwest Pacific Ocean are primarily characterized as relatively clear waters. In contrast, the coastal areas of the Yellow Sea and the East China Sea are predominantly classified as highly turbid waters. The QA algorithm can eliminate erroneous Rrs data during the atmospheric correction process, leading to an expected increase in the reliability of GOCI-II data. The algorithm will be integrated into the level-2 flag data provided alongside GOCI-II atmospheric correction.