Dynamic export coefficient model for evaluating the effects of environmental changes on non-point source pollution

The classic export coefficient model has been questioned due to its fixed coefficient, especially for those large-scale watersheds where great temporal-spatial heterogeneity exists. In this paper, a dynamic export coefficient model (DECM) was proposed for simulating non-point source (NPS) pollution by incorporating the impacts of factors on export coefficients. The relationships between rainfall, slope, soil, land use, other factors and export coefficients were constructed at relatively smaller catchment based on the information of mechanistic-based model, while these dynamic export coefficients were then extended to the large ungauged basins. This new model was tested in the Three Gorges Reservoir Region (TGRR), China. The results indicated the new method improved the accuracy of large-scale NPS prediction as well as reducing the computation burden. The rainfall temporal variability was identified as the major factor influencing the variability of flow and NPS pollution with the coefficient of variation being 0.1678 and 0.2046, respectively. Using the new method, the Long watershed, the Jialing watershed, the Quxi watershed, the Xiangxi watershed and the main stream in the TGRR were identified as those sensitive regions under the changing environment. The DECM could be extended to other large scale to quantify the NPS pollution, especially data-poor watersheds. [Display omitted] •A new method is proposed to simulate non-point source (NPS) pollution in large-scale catchments.•Heterogeneity of export coefficient are quantified based on rainfall and underlying surface.•Sensitive areas of NPS pollution under the changing environment were identified.•Rainfall variability showed greater impact on NPS pollution than land use change.