Intra- and inter-annual spatiotemporal variations and climatic driving factors of surface water area in the Irtysh River Basin during 1985–2022

Climate change and human activities have significantly altered the dynamics of surface water area (SWA) in the Irtysh River Basin (IRB). While inter-annual trends in SWA can be detected using Landsat imagery, the characteristics of seasonal SWA changes under long-term scenarios remain uncertain due to reduced data availability caused by cloud cover. In this study, we propose a time-disaggregated water frequency (TWF) that is more suitable for seasonal surface water analysis and develop a cloud-filling algorithm utilizing a Random Forest approach. The results demonstrate that the TWF effectively represents seasonal surface water distribution and achieves high cloud-filling accuracy. Using this method, we reconstructed monthly cloud-filled SWA series for the IRB from 1985 to 2022 at a spatial resolution of 30 m with high accuracy (>94%). Analysis indicates that the multi-year average SWA of the IRB was 41,003 km², reflecting a decrease of 22%. The peak SWA occurs in spring (May), following the general trend of spring > summer > fall > winter. Surface water loss primarily occurs during summer and fall, particularly in the middle reaches of the Irtysh River Basin (35%). Time-series correlation analysis reveals that snowmelt, precipitation, and temperature are the most significant climatic factors affecting SWA in spring, summer, and fall.