Mapping seasonal snow melting in Karakoram using SAR and topographic data

Abstract

Mapping seasonal snow melting is crucial for assessing its impacts on water resources, natural hazards, and regional climate in Karakoram. However, complex terrain in the high-mountain region poses great challenges to remote-sensing-based wet snow mapping methods. In this study, we developed a novel framework that incorporates synthetic aperture radar (SAR) and topographic data for robust and accurate mapping of wet snow over Karakoram. Our method adopts the Gaussian mixture model (GMM) to adaptively determine a wet snow index (WSI) and a computed topographic snow index (TSI) considering the impact of terrain on wet snow distribution to improve the accuracy of mapping. We validated the mapping results against Sentinel-2 snow cover maps, which demonstrated significantly improved accuracy using the proposed method. Applied across three major water basins in Karakoram, our method generated large-scale wet snow maps and provided valuable insights into the temporal dynamics of regional snow melting extent and duration. This study offers a practical and robust method for snow melting monitoring over challenging terrains. It can contribute to a significant step forward in better managing water resources under climate change in vulnerable regions.