A comprehensive comparison of GEMS Background Surface Reflectance (BSR) with OMI GLER and TROPOMI DLER at 440 nm

In remote sensing, improving the accuracy of Level 2 (L2) algorithms requires accurate assessment of surface reflectance across the ultraviolet (UV) to visible (VIS) spectrum. However, the interdependence between L2 algorithms and surface reflectance retrieval poses significant challenges, highlighting the need for alternative methods. To address this, many satellite algorithms produce Lambertian Equivalent Reflectance (LER) products as prior surface reflectance information, but these often results in underestimation. Therefore, to address this underestimation problem and produce more realistic surface reflectance, this study introduces a novel background surface reflectance (BSR) retrieval approach using GK-2B/Geostationary Environment Monitoring Spectrometer (GEMS) data by implementing a semi-empirical bidirectional reflectance distribution function (BRDF) model. The BSR method represents one of the first operational applications of BRDF modelling to hyperspectral satellite data in the UV-visible spectrum, addressing the limitations of traditional LER-based approaches in capturing surface anisotropy. In recent years, many studies have announced that the accurate representation of surface reflectance through BRDF effects has become critical in satellite-based aerosol optical depth (AOD) and gas retrieval algorithms such as NO2. While datasets such as OMI geometry-dependent surface Lambertian-equivalent reflectivity (GLER) and TROPOMI  directionally dependent Lambertian-equivalent reflectivity (DLER) have been developed to account for surface anisotropy for operational purposes, comprehensive comparisons of these products remain limited. Therefore, this study systematically compares BSR products with existing operational datasets (OMI GLER, TROPOMI DLER, and other LER databases) to assess their performance under different environmental and observational conditions, emphasising the need to consider BRDF effects.The comparative analysis reveals the specific characteristics, advantages and limitations of each dataset, providing valuable insights for their application in satellite-based retrievals. The results contribute to improving the accuracy of the Geo-Ring project and environmental monitoring initiatives by enhancing our understanding of surface reflectance behavior.