Incorporating receiving waters responses into the framework of spatial optimization of LID-BMPs in plain river network region

Surrogate modelling-based multi-objective optimization for best management practices of nonpoint source pollution

The integrated application of hydrological models and best management practices (BMPs) serves as a pivotal decision-making tool for managing nonpoint source (NPS) pollution in watersheds. Optimizing and selecting BMP options are critical challenges in managing NPS pollution, as these processes are typically computationally expensive and involve mixed discrete-continuous decision variables. Our study integrated a novel method, the multi-objective adaptive surrogate modeling-based optimization for constrained hybrid problems (MO-ASMOCH), with the distributed Soil and Water Assessment Tool (SWAT) model to efficiently optimize the deployment of BMPs in the Four Lakes watershed of China. We compared the optimization results with those obtained using the traditional non-dominated sorting genetic algorithm (NSGA-II) method. Our results demonstrate that MO-ASMOCH significantly outperforms NSGA-II in computational efficiency, achieving comparable Pareto-optimal solutions with just 1,150 model evaluations compared to NSGA-II's requirement of 10,000 model evaluations. This demonstrates that MO-ASMOCH is a more efficient optimization algorithm for BMP optimization problems with both discrete and continuous decision variables. We selected representative scenarios to calculate in-lake concentrations of total phosphorus (TP) and total nitrogen (TN) pollutant loads. The largest reduction scenario could reduce TN and TP loads by 18.3 % and 20.7 %, respectively, at a cost of 1.54 × 108 Chinese Yuan. Under this scenario, the water quality classification level of TN improves from inferior Class V to Class IV-V, while TP attains Class III throughout the year. The methods of this study could enhance our capability to manage NPS pollution in watersheds effectively and provide targeted decision-making insights for environmental management practices.

Long short-term memory models of water quality in inland water environments

Water quality is substantially influenced by a multitude of dynamic and interrelated variables, including climate conditions, landuse and seasonal changes. Deep learning models have demonstrated predictive power of water quality due to the superior ability to automatically learn complex patterns and relationships from variables. Long short-term memory (LSTM), one of deep learning models for water quality prediction, is a type of recurrent neural network that can account for longer-term traits of time-dependent data. It is the most widely applied network used to predict the time series of water quality variables. First, we reviewed applications of a standalone LSTM and discussed its calculation time, prediction accuracy, and good robustness with process-driven numerical models and the other machine learning. This review was expanded into the LSTM model with data pre-processing techniques, including the Complete Ensemble Empirical Mode Decomposition with Adaptive Noise method and Synchrosqueezed Wavelet Transform. The review then focused on the coupling of LSTM with a convolutional neural network, attention network, and transfer learning. The coupled networks demonstrated their performance over the standalone LSTM model. We also emphasized the influence of the static variables in the model and used the transformation method on the dataset. Outlook and further challenges were addressed. The outlook for research and application of LSTM in hydrology concludes the review.