Modeling Drivers and GLOF Hazards of Glacial Lakes in the Central Asian Mountains Using Geodetector and XGBoost
摘要
Under ongoing global warming, the factors driving glacial lake number and area variations and assessing outburst risks are essential for regional water resource management and hazard mitigation. Few studies have focused on glacial lake variations based on continuous long-term time series and broad regional scales, and systematic investigations of outburst risks for glacial lakes in the Central Asian Mountains remain inadequate. Leveraging the Google Earth Engine (GEE), we determined glacial lake dynamics from 2000 to 2023, performed driving factor analysis with Geodetector, and implemented Extreme Gradient Boosting (XGBoost) for outburst risk evaluation. Among the three subregions analyzed, the Tianshan Mountains experienced the most significant glacial lake expansion, with 363 newly formed lakes and an average annual increase in area of 1.78 km2 yr− 1. Across the entire Central Asian Mountains, elevation, air temperature, and glacier-related factors were identified as the primary controls on glacial lake number and area changes. The accelerated expansion of glacial lakes in the Tianshan Mountains, driven by recent temperature increases and intensified glacier retreat, has further elevated the region’s potential outburst risk. This study provides new insights into the driving factors and outburst susceptibility of glacial lakes in the Central Asian Mountains.
Graphical Abstract