Evaluation of groundwater potential using ANN-based mountain gazelle optimization: A framework to achieve SDGs in East El Oweinat, Egypt

Abstract

Study regionA pilot case study in East El Oweinat (PCSEO), Egypt. Study focusAn artificial neural network (ANN)-based mountain gazelle optimization (MGO) model was applied to map groundwater potential zones (GWPZs). For this purpose, ten layers affecting groundwater occurrence were prepared and normalized against the groundwater drawdown (DD) map. All data was divided into 70:30 for training and testing. After that, sensitivity analysis was adopted to verify the relative importance (RI) of layers. The accuracy of GWPZs was checked using the receiver operating characteristic (ROC) curve and other statistical indicators. The model was finally applied to propose a sustainable strategy for groundwater exploration by implementing the integrated MODFLOW-USG and MGO framework. New hydrological insights for the regionOver 40% of the PCSEO revealed high to very high groundwater potential degrees and was situated mostly on the southwestern side. Sensitivity analysis revealed that GWPZs were significantly affected by groundwater table (GWT), well density (WD), and land use (LU). The results also indicated that the ANN-based MGO model performed well with an area under curve (AUC) of ∼ 90% compared to other conventional models. Additionally, the MODFLOW-USG-based MGO model gave a spatial distribution of optimal discharge and well-depth zones. This finding could significantly match SDGs relevant to ending poverty, affordable groundwater, and life on land.