Multi-time scale estimation of water retention in a humid basin in southern China using SWAT model

Water conservation, a critical ecosystem service, is primarily quantified through water retention (WR), which plays a pivotal role in sustainable socio-economic development and water resources management. However, the absence of multi-temporal modeling of land use and climate change impacts on eco-hydrological processes limits the accurate estimation of WR, particularly in humid regions. This study employed the Soil and Water Assessment Tool (SWAT) model coupled with the water balance principle to estimate WR in the source area of the Xin'an River (SXAR) in China from 2009 to 2017. The multi-temporal variations of WR and its response to climate and land use changes were analyzed through scenario-based hydrological simulations. Results indicated that annual WR ranged from 256.4 mm to 412.7 mm, monthly WR varied between 0 mm and 67.6 mm, and peak daily WR coincided with extreme rainfall events. Precipitation and evapotranspiration were identified as the primary factors influencing WR variability at daily and monthly scales. Spatially, higher WR values were observed in the northeastern SXAR, reflecting the influences of land use patterns and topography. Notably, agricultural land exhibited negative WR during summer months due to crop water storage demands. Overall, climate change exerted more immediate effects on WR at shorter timescales, whereas land use change produced longer-term impacts. This study offers valuable theoretical insights into WR mechanisms of response to environmental changes and provides practical guidance for water resources planning and management in humid and sub-humid regions.