Evolution characteristics and driving factors of potential non-point source pollution risks in a watershed affected by land use changes.

Land use types, land development and utilization intensity within watersheds have changed based on intensifying human activities and climate change, thereby inducing spatiotemporal variations in non-point source pollution (NPS), significantly impacting soil and water quality. This study performed a case study on an ecological environment functional zone at the northern foot of Qinling Mountains, an area strongly affected by human activities and land use changes. It employed an improved potential non-point pollution index (PNPI) model to analyze potential non-point source pollution (PNPS) and associated risk evolution characteristics in watershed over the past 30 years. The results indicate that from 1990 to 2020, the dominant land use categories were forest and arable land, making up 95 % of the entire watershed area. Notably, urban residential land presented the most significant expansion rates and nearly doubled in area between 1990 and 2020, whereas shrubland, grassland, and unused land showed a decreasing trend. With the application of the quantile classification method, PNPS risk values were divided into five categories: very low, low, moderate, high, and very high. A polarized trend in risk was observed, with increases in areas influenced by human activities and rapid expansion of very high-risk regions. Concurrently, the pollution risk in the upstream water source area decreased. In recent years, accelerated urbanization has been the main driver causing expansion of high PNPS risk regions. This study explores the spatial and temporal evolution of PNPS risk in the Heihe Basin by using an improved PNPI model. The improved model is more accurate in calculations and provides a better understanding of the distribution of PNPS, which is an important reference for watershed management and water resource governance.