Multi-Source Remote Sensing Reveals Multi-Timescale Variations of Lakes on the Tibetan Plateau: A Case Study of Three Representative Lakes

Lakes on the Tibetan Plateau are sensitive indicators of regional climate change and hydrological variability. Comprehensive analyses of the area, level, and volume of typical lakes across multiple time scales remain limited. In this study, based on the Google Earth Engine platform, we integrated Landsat TM/ETM+/OLI and Sentinel-2 imagery, nine satellite altimetry datasets (Jason-1, Jason-2, Jason-3, Cryosat-2, Sentinel-3A, Sentinel-3B, ICESat, ICESat-2, and GEDI), and four machine learning methods (Lasso, SVM, Random Forest, and XGBoost) to investigate multi-timescale (annual, seasonal, and monthly) variations in lake area, level, and volume of Qinghai Lake, Nam Co, and Bangong Co, from 2000 to 2022. All three lakes showed overall increases in lake area, level, and volume, but with different magnitudes and temporal patterns. Qinghai Lake showed the most pronounced expansion, Nam Co experienced moderate growth, characterized by temporal fluctuations, whereas Bangong Co showed a relatively small but persistent increase. Seasonal analysis showed that lake area expansion was generally strongest in autumn, whereas monthly lake volume peaked in October for Qinghai Lake and Nam Co and in September for Bangong Co. Climate analyses indicated that Qinghai Lake may be mainly influenced by increased precipitation and runoff as well as reduced evaporation, Nam Co may be mainly influenced by precipitation, runoff, and delayed hydrological responses to glacier meltwater, and Bangong Co was likely associated with runoff and cryospheric water supply, while basin characteristics may have also contributed to its long-term hydrological response. These results highlight regional differences in lake responses to climate variability across the Tibetan Plateau.