Author(s): Jianxiang Xiong; Lingqiao Kong; Zhiyun Ouyang

Linked Author(s):

Keywords: Water retention; Climate change; Land cover pattern; VIC model; Ecosystem; Watershed

Abstract: Forests, wetlands, and grasslands regulate the hydrological cycle by intercepting precipitation, conserving water, and modulating runoff, thereby mitigating extreme hydrological events. This study examines the Xiangjiang River Basin in southern China, a su btropical monsoon region with abundant but unevenly distributed precipitation and frequent floods and droughts. We quantify the hydrological regulation capacity of different ecosystems and their roles in water balance. Using the VIC model and machine learn ing approaches, we analyzed temporal variations of hydrological variables, including precipitation, surface runoff, and vegetation water retention, and characterized their seasonal dynamics. We further quantified how ecosystems partition precipitation into evapotranspiration, runoff, and storage. Grid-scale regression analyses of runoff coefficients against land cover proportions, performed with OLS, Random Forest, and XGBoost, reveal a pronounced nonlinear response of runoff to land-use patterns. These findings highlight the critical influence of ecosystem composition on basin-scale hydrology, and provide policy-relevant guidance for setting scientifically informed ecological restoration targets to enhance ecosystem resilience under future climate scenarios.

DOI: https://doi.org/10.64697/iahr.proc.ise2026.abs.242

Year: 2026