Ammonium nitrogen (NH₄⁺-N), which naturally arises from the decomposition of organic substances through ammonification, has a tremendous influence on local water quality. Therefore, it is vital for water quality protection to assess the amount, sources, and streamflow transport of NH₄⁺-N. SPAtially Referenced Regressions on Watershed attributes (SPARROW), which is a hybrid empirical and mechanistic modeling technique based on a regression approach, can be used to conduct studies of different spatial scales on nutrient streamflow transport. In this paper, the load and delivery of NH₄⁺-N in Poyang Lake Basin (PLB) and Haihe River Basin (HRB) were estimated using SPARROW. In PLB, NH₄⁺-N load streamflow transport originating from point sources and farmland accounted for 41.83% and 32.84%, respectively. In HRB, NH₄⁺-N load streamflow transport originating from residential land and farmland accounted for 40.16% and 36.75%, respectively. Hence, the following measures should be taken: In PLB, it is important to enhance the management of the point sources, such as municipal and industrial wastewater. In HRB, feasible measures include controlling the domestic pollution and reducing the usage of chemical fertilizers. In addition, increasing the vegetation coverage of both basins may be beneficial to their nutrient management. The SPARROW models built for PLB and HRB can serve as references for future uses for different basins with various conditions, extending this model’s scope and adaptability.

Accurate streamflow forecasting is vital for sustainable water resource management but remains challenging due to pronounced spatiotemporal variability. This study evaluates two process-based models, the SWAT (comprehensive) and the GWLF (parsimonious), and a data-driven random forest (RF) model for monthly streamflow simulations in two contrasting Chinese basins: the humid southern basin (SSB) and the semi-arid northern basin (SRB). Using four statistical metrics (NSE, R², MAE, RMSE), we assess model accuracy, robustness in capturing extremes, and sensitivity to hydrological characteristics and data availability. The results reveal consistently superior performance in the SSB across all models, with SWAT demonstrating the highest overall accuracy—especially for peak flows—due to its physically based structure. The GWLF provides acceptable simulations with minimal data requirements, offering a practical alternative in data-limited regions, like the SRB. RF performs well in the SSB under zero-lag conditions but requires hydrologically informed lag structures in the SRB. However, it consistently underestimates high flows due to its lack of physical constraints. The findings underscore that model selection must, therefore, be guided not only by predictive performance but also by the underlying hydrological context, data availability, and the need for physical realism in decision-making.

Accurate streamflow forecasting is vital for sustainable water resource management but remains challenging due to pronounced spatiotemporal variability. This study evaluates two process-based models SWAT (comprehensive) and GWLF (parsimonious)—and a data-driven Random Forest (RF) model for monthly streamflow simulation in two contrasting Chinese basins: the humid southern basin (SSB) and the semi-arid northern basin (SRB). Using four statistical metrics (NSE, R2, MAE, RMSE), we assess model accuracy, robustness in capturing extremes, and sensitivity to hydrological characteristics and data availability. Results reveal consistently superior performance in the SSB across all models, with SWAT demonstrating the highest overall accuracy—especially for peak flows—due to its physically based structure. GWLF provides acceptable simulations with minimal data requirements, offering a practical alternative in data-limited regions like the SRB. RF performs well in the SSB under zero-lag conditions but requires hydrologically informed lag structures in the SRB. However, it consistently underestimates high flows due to its lack of physical constraints. The findings underscore that model selection must therefore be guided not only by predictive performance, but by the underlying hydrological context, data availability, and the need for physical realism in decision-making.