Hydrological series data are non-stationary and nonlinear. However, certain data-driven forecasting methods assume that streamflow series are stable, which contradicts reality and causes the simulated value to deviate from the observed one. Ensemble empirical mode decomposition (EEMD) was employed in this study to decompose runoff series into several stationary components and a trend. The long short-term memory (LSTM) model was used to build the prediction model for each sub-series. The model input set contained the historical flow series of the simulation station, its upstream hydrological station, and the historical meteorological element series. The final input of the LSTM model was selected by the MI method. To verify the effect of EEMD, this study used the Radial Basis Function (RBF) model to predict the sub-series, which was decomposed by EEMD. In addition, to study the simulation characteristics of the EEMD-LSTM model for different months of runoff, the GM(group by month)-EEMD-LSTM was set up for comparison. The key difference between the GM-EEMD-LSTM model and the EEMD-LSTM model is that the GM model must divide the runoff sequence on a monthly basis, followed by decomposition with EEMD and prediction with the LSTM model. The prediction results of the sub-series obtained by the LSTM and RBF exhibited better statistical performance than those of the original series, especially for the EEMD-LSTM. The overall GM-EEMD-LSTM model performance in low-water months was superior to that of the EEMD-LSTM model, but the simulation effect in the flood season was slightly lower than that of the EEMD-LSTM model. The simulation results of both models are significantly improved compared to those of the LSTM model.

Many developing countries and regions are currently facing serious water environmental problems, especially the lack of monitoring systems for medium- to small-sized watersheds. The load duration curve (LDC) is an effective method to identify polluted waterbodies and clarify the point sources or non-point sources of pollutants. However, it is a large challenge to establish the LDC in small river basins due to the lack of available observed runoff data. In addition, the LDC cannot yet spatially trace the specific sources of the pollutants. To overcome the limitations of LDC, this study develops a LDC based on a distributed hydrological model of the Soil and Water Assessment Tool (SWAT). First, the SWAT model is used to generate the runoff data. Then, for the control and management of over-loaded polluted water, the spatial distribution and transportation of original sources of point and non-point pollutants are ascertained with the aid of the SWAT model. The development procedures of LDC proposed in this study are applied to the Jian-jiang River basin, a tributary of the Yangtze River, in Duyun city of Guizhou province. The results indicate the effectiveness of the method, which is applicable for water environmental management in data-scarce river basins.