A Decision Support Tool for the Dynamic Groundwater Management of Coastal Aquifers Under Uncertainty

In this paper, we described a computationally efficient simulation–optimization (S/O) framework for coastal groundwater management (CGM), based on the combined application of numerical modeling, artificial neural networks, and genetic algorithm. The objective was to analyze the ‘trade-off’ between optimality and risk in deriving CGM strategies and to show that reformulating the problem using a mean–variance bi-criterion objective function can be a valuable tool in minimizing the risk of non-optimality. As a case study for our analysis, we studied the optimal design of an aquifer storage and recovery (ASR) system in the Muscat metropolitan area in Oman. S/O was applied to find the optimal constant daily abstraction and injection rates that maximize the net present value (NPV) of the ASR system. The results show that the choice of the decision variables significantly affects the risk of non-optimality, and reducing this risk comes at the cost of a decrease in the expected NPV.