Study of water quality response to water transfer patterns in the receiving basin and surrogate model

Lei Liu,Xue-yi You
DOI: https://doi.org/10.1007/s11356-021-17469-4
IF: 5.8
2021-11-26
Environmental Science and Pollution Research
Abstract:AbstractThe spatio-temporal response of water quality in the receiving basins to water transfer remains unclear when considering more practical factors. Moreover, a much fast surrogate model is highly required to realize the purpose of rapid prediction of water quality. In this study, a two-dimensional hydrodynamic and water quality model was primarily established by considering the most comprehensive important pollution sources so far to investigate the water quality variation in Baiyangdian Lake, China. The pollution sources include atmospheric deposition, sediment release, village and tourism domestic sewage, livestock breeding, rainfall runoff, aquatic plants, and animals. The water quality at different sample sites was assessed using the water quality index (WQI). Subsequently, a surrogate model was proposed using the nonlinear autoregressive network with exogenous inputs (NARX), and the NARX model can realize fast prediction of the annual time series of water quality in a few seconds. The results showed that there are significant differences in water quality at different sample sites in each season. The highest WQI value occurred in autumn (65.46), and the average values at different sample sites were between 47.54 and 58.22 in this season. The water quality at Nanliuzhuang site and Caiputai site was significantly affected by water transfer patterns. In addition, the correlation coefficient R2 between the predicted results and the simulated ones except that of parameter TP was larger than 0.90. The prediction performance of surrogate model was shown to be very high efficiency and accurate. The proposed models and the results of this approach are guiding significance for lake management.
environmental sciences
What problem does this paper attempt to address?