Sensitivity of GOCI-II Rrs products by NO2 absorption correction

Atmospheric nitrogen dioxide (NO₂) absorbs solar radiation, particularly in the blue and green wavelengths that are essential for ocean color algorithms. This absorption challenges achieving high accuracy in satellite-derived ocean color products, such as chlorophyll-a concentration, total suspended material, and remote sensing reflectance. Since NO₂ absorption is influenced by various observation conditions—including the angles between the sun, Earth, and satellite- and the quantity and vertical distribution of NO₂—a precise correction model is crucial. However, the current GOCI-II atmospheric correction algorithm accounts for the absorption effects of water vapor and ozone, while ignoring the impact of NO₂ absorption. The GOCI-II observation area is one of the regions with the highest NO₂ concentrations, along with Europe and the United States, and it is the only region showing an increasing trend in NO₂ amount. Particularly in coastal areas, high NO₂ concentrations are observed due to industrial activities, marine transportation, and agricultural practices. Therefore, neglecting the NO₂ absorption effect in the atmospheric correction algorithm could become a potential source of error in GOCI-II ocean color products. In this study, we analyzed the impact of NO₂ absorption correction on primary ocean color products (remote sensing reflectance, colored dissolved organic matter, and chlorophyll-a concentration) by comparing the values before and after applying the correction. The GEMS data were used as input for NO₂ concentration. GEMS is a hyperspectral sensor onboard the same satellite as GOCI-II (Geo-Kompsat-2B). Unlike polar-orbiting sensors, GEMS provides hourly observations and offers real-time NO₂ concentration data with improved spatial resolution compared to atmospheric model data. This capability makes GEMS highly effective in estimating the spatiotemporal variability of NO₂ distribution and correcting absorption effects, while also reducing uncertainties caused by climatological assumptions. After NO₂ absorption correction, the Rrs at 412 nm showed a significant difference of 7% across the entire ocean in GOCI-II slot 7. However, CHL and CDOM exhibited smaller changes of 3.82% and 5.18%, respectively. In contrast, in coastal pixels with high NO₂ concentrations, the differences in CHL and CDOM before and after NO₂ absorption correction increased significantly to 36.97% and 28.43%, respectively. Therefore, NO₂ absorption correction is essential to improve the accuracy of ocean color products in coastal regions.

* This research was supported by the National Research Foundation of Korea (NRF) grant funded by the Ministry of Science and ICT of Korea (MSIT) (RS-2024-00356738).