Modeling and Evaluating of Non-Point Source Pollution in a Semi-Arid Watershed: Implications for Watershed Management

Watershedmodels are cost-effective and powerful tools for evaluating and controlling non-point source (NPS) pollution. Generalized Watershed Loading Functions (GWLF) model, a relative parsimonious model, was applied to simulate watershed streamflow and nutrient loads in Liu River watershed, northern China. The model results were used to evaluate watershed NPS load, and the implications for NPS pollution mitigation measures were investigated. The temporal-spatial distribution, the critical source areas (CSAs) at sub-watershed level, and the best management practices (BMPs) efficiencies at CSAs were investigated by GWLF model. Due to the contrasting hydrological conditions, the NPS loads of wet year and flood season were much more than those of dry year and non-flood season. Four sub-watersheds were identified as CSAs, and the results showed that CSAs had high level of load intensities due to the high percentage of agricultural land and residential land. Through evaluating the efficiency and cost of seven BMPs, conservation tillage and vegetative buffers were selected as the measures to implement in the CSAs, and they can reduce 14.2% and 12.5% of total nitrogen and total phosphorus NPS loads in Liu river watershed. These findings can provide valuable information for implementing more effective and reasonable management measures to control NPS pollution in Liu River watershed and indicate the utility of GWLF with modest data requirements for such application.