Assessment of GOCI-II satellite remote sensing products in Lake Taihu

Introduction The Geostationary Ocean Color Imager-II (GOCI-II), launched on February 19, 2020, offers increased observation times throughout the day and higher spatial resolution compared to its predecessor, the Geostationary Ocean Color Imager (GOCI), launched in 2010. To ensure the reliability of GOCI-II data for practical applications, the accuracy of remote sensing products needs to be validated. This study uses in situ data from Lake Taihu for validation. Methods We assessed the accuracy of GOCI-II remote sensing products, including remote sensing reflectance derived using two atmospheric correction algorithms: ultraviolet (UV) and near-infrared (NIR). The study also evaluated the accuracy of derived parameters, such as chlorophyll-a (Chl-a) concentration, total suspended matter (TSM) concentration, and phytoplankton absorption coefficient (a ph ), based on these atmospheric correction algorithms. In situ measurements from Lake Taihu were used as ground truth data for validation. Results Our results revealed that the UV atmospheric correction algorithm provided higher accuracy in Lake Taihu compared to the NIR algorithm. The average absolute percentage deviations (APDs) for remote sensing reflectance across different bands were: 25.17% (412 nm), 29.69% (443 nm), 22.27% (490 nm), 19.38% (555 nm), 36.83% (660 nm), and 33.0% (680 nm). Compared to NIR-derived products, the UV algorithm showed improved accuracy for Chl-a concentration, TSM concentration, and a ph , with reductions in APD values by 16.92%, 3.32%, and 10.91%, respectively. When applying UV correction, the 412 nm band performed better than the 380 nm band, likely due to a lower signal-to-noise ratio at 380 nm and smaller extrapolation errors at 412 nm. Discussion While the NIR algorithm is suitable for open ocean waters, the UV algorithm demonstrated higher accuracy in turbid environments such as Lake Taihu. Therefore, a combined UV-NIR atmospheric correction algorithm may be more effective for handling various types of water environments. Additionally, further research is needed to develop more suitable retrieval algorithms for Chl-a concentration and aph in eutrophic waters to improve accuracy.