Exploring controls on precipitation-runoff dependencies: Implications for non-stationary and spatially heterogeneous analyses

Climate change and complex anthropogenic activities present significant challenges to understanding Precipitation-Runoff Dependencies (PRD). Traditional methods, which often assume stationary and linear conditions, may not fully capture these complex relationships. To address this limitation, we propose an integrated framework that incorporates non-stationary and spatially heterogeneous analyses to identify the controlling mechanisms influencing PRD. This framework was applied to eleven sub-basins within the Illinois River Basin, a region characterized by high spatiotemporal variability and intense anthropogenic activity. The study introduces the novel Controlling Index for Changes in Precipitation-Runoff Dependencies (CC-PRD) and employs a geographical detector model to effectively identify and analyze these controlling factors. Our findings reveal that under non-stationary conditions, baseflow (BF) is the primary driver of PRD across all sub-basins. However, its impact varies by basin type: in urban sub-basins, BF weakens PRD, while in rural sub-basins, BF enhances PRD. Beyond BF, anthropogenic factors, such as impervious surface percentage (ISP) and rural area percentage (RAP), emerge as the key drivers of PRD variation in urban sub-basins, whereas in rural sub-basins, natural factors, particularly potential evapotranspiration (PET), play a dominant role in shaping PRD. Spatially, PET, BF, and RAP emerge as the main determinants of PRD, with their interactions with other factors significantly amplifying their influence, accounting for over 90% of its variability. This comprehensive approach enhances our understanding of the non-stationarity and spatial heterogeneity in PRD, while considering the challenges posed by climate change and human activities to the stationary assumption, thereby supporting strategies for sustainable watershed development.