Natural versus anthropogenic forces affecting spatial variations of Ebinur Lake in Northwest China over the past two decades 2001–2020

The Ebinur Lake Watershed, located in Northwest China, serves as a crucial ecological barrier for the region. Over the past two decades, Ebinur Lake and its surrounding marshes have experienced a significant loss of water area due to excessive human exploitation and encroachment, resulting in severe ecological degradation. This study examines the driving factors behind changes in water area from 2001 to 2020. Using the Google Earth Engine (GEE) platform, the Normalised Difference Water Index (NDWI) was applied to extract lake area, while hydrological data from meteorological stations were integrated with regional economic indicators, including primary industry output and crop sowing areas. An eXtreme Gradient Boosting (XGBoost) model, enhanced with SHapley Additive exPlanations (SHAP) analysis, was employed to quantify the influence of natural and anthropogenic factors on lake area fluctuations. The results indicate a significant decline in Ebinur Lake’s water area, peaking at 946.32 km² in 2003 and subsequently decreasing to 409.33 km² by 2015. Seasonal variation was greatest in May (637.35 km²) and lowest in October (480.74 km²). Changes were driven by both natural and anthropogenic factors, with inflow river run-off having the most substantial natural impact and population density being the dominant human factor. This research provides valuable insights for sustainable development and effective ecological resource management in the Ebinur Lake Watershed. • Extract and validate Ebinur Lake's water area (2001–2020) using NDWI on GEE platform. • Monitor monthly and annual changes in Ebinur Lake's water area. • Construct XGBoost model with SAHP to explain index contributions for Ebinur Lake sustainability.