Reconstructing terrestrial water storage and assessing climate–land cover interactions in the Hulun Lake Basin using multi-source data fusion

Study Region

Hulun Lake Basin (HLB), a large transboundary watershed spanning 218,000 km² in northern China and Mongolia

Study Focus

Understanding long-term terrestrial water storage (TWS) dynamics in data-scarce, climate-sensitive regions is critical for sustainable watershed management. This study reconstructs a continuous TWS and groundwater storage (GWS) time series from 1960 to 2023 using a fusion model integrating Random Forest and Kalman Filter algorithms. By leveraging multi-source datasets, the study aims to analyse the nonlinear hydrological response to climate change and land cover shifts in the arid and semi-arid HLB.

New hydrological insights for the region

The reconstructed TWS revealed four distinct hydrological phases driven by climate variability: cold-dry (1972–1981), cold-wet (1981–1990), warm-dry (1990–2007), and warm-wet (2007–2021). A net water deficit was observed due to precipitation consistently falling below evapotranspiration, accompanied by a warming trend of 0.04 °C/year—exceeding the global mean. Groundwater was identified as the dominant driver of TWS changes, with precipitation showing the highest climatic elasticity. Between 1990 and 2005, TWS sharply declined (∼36 % drop in precipitation) despite stable land cover. From 2005–2020, notable TWS gains were concentrated in barren lands within the Ulson River Basin, while limited recovery in the Kherlen River Basin reflected spatial heterogeneity in land cover influence. These findings highlight the asymmetric hydrological responses across sub-basins and offer a valuable dataset for future water resource planning under changing climate conditions.