Contrasting Water-Partitioning Responses to Land-Use Change and Climate Perturbations in a Forested Mountain Catchment

Forested mountain catchments play a critical role in regulating regional water resources, yet the effects of land-use change and precipitation–temperature perturbations on internal water partitioning and its seasonal and spatial variability remain insufficiently understood. In this study, we integrated remotely sensed land-cover information and MODIS-derived normalized difference vegetation index (NDVI) with a calibrated and validated Soil and Water Assessment Tool (SWAT) model to examine hydrological sensitivities and non-additive responses under three land-use configurations and controlled climate perturbations in the upper Pi River Basin, China. Land-cover transitions affected only 2.7% of the study area and produced relatively small hydrological differences among land-use configurations, whereas a 20% precipitation increase enhanced soil-profile percolation (PERC) and surface runoff (SURQ) by 28.62% and 49.81%, respectively; 2 °C warming increased actual evapotranspiration (ET) by 3.58% but reduced SURQ by 5.39%. Under the combined scenario of 20% higher precipitation and 2 °C warming, ET and SURQ increased by 7.15% and 42.06%, respectively, with a non-additive SURQ deviation of −2.36 percentage points; ET sensitivity to precipitation peaked in August, coinciding with the NDVI maximum, whereas sensitivity to warming peaked in May, and soil water storage (SW) responses to warming differed in direction among subbasins. These findings reveal contrasting influences of precipitation supply and temperature-related evaporative demand on hydrological pathways, with component-specific non-additive effects and distinct seasonal and spatial responses. These patterns support the use of remotely sensed vegetation information and distributed hydrological modeling to identify climate-sensitive periods and evaluate spatial differences in water availability in forested mountain catchments.