Spatial-temporal source apportionment of nitrogen and phosphorus in a high-flow variable river

Mihe River Basin, China. The composite interaction of natural and anthropogenic factors has increased the complexity of hydrological variation and pollutant transformation in rivers, making it challenging to effectively analyze the spatial and temporal distribution characteristics of pollution sources. We combined two empirical models to quantitatively assess the spatial-temporal distribution of pollution sources in a high-flow variable river such as the Mihe River, providing a scientific basis and methodological support for watershed management. The findings show that: (1) The average nitrogen and phosphorus loading of the river during 2015–2021 were 13.08 × 103 and 38.03 t·a-1, respectively, exhibiting significant fluctuations with flow; (2) a load apportionment model (LAM) revealed that nonpoint sources were primary pollution sources for most of the period, and average contributions of nitrogen and phosphorus loading accounted for 81.7 and 99.6%, respectively; (3) an export coefficient model (ECM) revealed that nutrient loading was concentrated in the middle and lower reaches, which was highly consistent with the distribution of cropland and urban land. (4) The results of the two methods were consistent. However, LAM and ECM underestimate and overestimate total nutrient loading, respectively. Combining these two methods facilitates the accurate determine of spatial and temporal distribution of pollution sources in a high-flow variable river.