Event-based nonpoint source pollution prediction in a scarce data catchment

Quantifying the rainfall-runoff-pollutant (R-R-P) process is key to regulating non-point source (NPS) pollution; however, the impacts of scarce measured data on R-R-P simulations have not yet been reported. In this study, we conducted a comprehensive study of scarce data that addressed both rainfall-runoff and runoff-pollutant processes, whereby the impacts of data scarcity on two commonly used methods, including Unit Hydrograph (UH) and Loads Estimator (LOADEST), were quantified. A case study was performed in a typical small catchment of the Three Gorges Reservoir Region (TGRR) of China. Based on our results, the classification of rainfall patterns should be carried out first when analyzing modeling results. Compared to data based on a missing rate and a missing location, key information generates more impacts on the simulated flow and NPS loads. When the scarcity rate exceeds a certain threshold (20% in this study), measured data scarcity level has clear impacts on the model's accuracy. As the model of total nitrogen (TN) always performs better under different data scarcity conditions, researchers are encouraged to pay more attention to continuous the monitoring of total phosphorus (TP) for better NPS-TP predictions. The results of this study serve as baseline information for hydrologic forecasting and for the further control of NPS pollutants. (C) 2017 Elsevier B.V. All rights reserved.