Simulation of non-point source pollution response to extreme rainfall and agricultural activities using the LSPC model: the case of the Zhegao River Basin, a tributary of Lake Chaohu, China

The intensification of precipitation extremes, owing to global climate change, has significantly altered hydrological processes and pollutant export characteristics in watersheds, exacerbating non-point source (NPS) pollution in agricultural watersheds. This study aimed to clarify the dynamic response mechanisms of NPS pollution in agricultural watersheds to precipitation extremes, focusing on the Zhegao River Basin. Using the LSPC model, a dynamic NPS pollution response model was constructed for the watershed and seven extreme rainfall scenarios (S0–S7) were simulated. The results showed that the timing of extreme rainfall events significantly regulates hydrological intensity and pollution responses in the watershed. Notably, a 1.38-fold increase in rainfall during the non-flood season (S3) resulted in a 1.52-fold increase in runoff, and heavy rainfall events during the flood season (S2) increased total nitrogen (TN) loads by approximately five fold relative to the baseline value, implying differences in rainfall impacts across temporal scales. Further, the temporal coupling of fertilisation and heavy rainfall considerably altered pollutant fluxes. Specifically, heavy rainfall before fertilisation during the flood season (S4) increased NH₃-N load by 350%, while heavy rainfall after fertilisation during the flood season (S5) increased TN loads by 233.3%, reflecting synergistic effects between pollutant sources and rainfall runoff. Therefore, extreme rainfall coupled with agricultural activities exerts a nonlinear amplification effect on watershed NPS pollution, and overall, a multi-level “source control—process interception—systematic governance” adaptive management strategy taking into account the impacts of rainfall and fertilisation is required for the case study watershed.