A Sun Glint Correction Method for Low- and Medium-Resolution Visible Imagery Over the Great Lakes

Sun glint presents a considerable challenge in remote sensing imagery, as it interferes with cloud detection and restricts the ability to observe target parameters such as ocean color, temperature, and bathymetry. Sun glint particularly affects the retrieval of water-leaving radiance or reflectance in the direction of the satellite's field of view. Current sun glint correction methods for low- and medium-resolution imagery are limited, especially when winds are weak. This study presents a novel method that uses an anisotropy index to capture the direction and tilt of surface waves based on a bidirectional reflectance distribution function (BRDF). We apply this sun glint correction method to Visible Infrared Imaging Radiometer Suite (VIIRS) observations of the Great Lakes and evaluate its performance compared to the Multispectral Imager (MSI) observations of Sentinel-2 glint removal (GRS) algorithm and in situ measurements from the Aerosol Robotic Network-Ocean Color (AERONET-OC) dataset. Reflectance and normalized water-leaving radiance after atmospheric correction based on our method agree well with GRS and in-situ measurements, respectively, significantly increasing correlations from 0.694 to 0.888 between VIIRS retrievals and in-situ data, while root mean-square error decreases to 0.615. This method can be applied to increase the availability of long-term climate data records for research on lacustrine ecosystems.