Developing an optimized groundwater exploitation prediction model based on the Harris hawk optimization algorithm for conjunctive use of surface water and groundwater resources.

Due to the lack of surface water and groundwater resources, especially in the agricultural parts, the simultaneous and sustainable use of water resources to supply water demands is essential. In this study, a conjunctive use optimization model is developed to minimize the shortage of water demand. This model is implemented for the Mahabad study area in northwestern Iran to improve the conditions of surface water and groundwater resources and the reclamation of Urmia Lake. For this purpose, the current research is accomplished in three parts. At first, the Mahabad aquifer is numerically simulated to investigate the aquifer conditions. In the second part, the optimized model of conjunctive use obtained by the Harris hawk optimization (HHO) algorithm is investigated over a 20-year period in the study area. In the last part of this research, seven scenarios are developed to predict the optimized groundwater exploitation (OGE) using the results of HHO, meteorological data, and some input information on the dam reservoir. Then, the OGE values are predicted using the artificial neural network (ANN) and ANN-HHO machine learning models for the scenarios. The results showed that the scenario that includes all input variables and the ANN-HHO model outperformed other models. Furthermore, the HHO algorithm provides suitable allocation of the surface water and groundwater resources in optimized conjunctive use and also improves the performance of ANN in predicting the OGE values. The findings of this study also show that groundwater resources can be more applied to supply water demand, and in contrast, surface water resources can be used for supplying downstream environmental demands and reclamation of Urmia Lake.