Predictive Modeling of Groundwater Recharge under Climate Change Scenarios in the Northern Area of Saudi Arabia

Abstract

Water scarcity is considered a major problem in dry regions, such as the northern areas of Saudi Arabia and especially the city of Hail. Water resources in this region come mainly from groundwater aquifers, which are currently suffering from high demand and severe climatic conditions. Forecasting water consumption as accurately as possible may contribute to a high level of sustainability of water resources. This study investigated different Machine Learning (ML) algorithms, namely Support Vector Machine (SVM), Random Forest (RF), Linear Regression (LR), and Gradient Boosting (GB), to efficiently predict water consumption in such areas. These models were evaluated using a set of performance measures, including Mean Squared Error (MSE), R-squared (R2), Mean Absolute Error (MAE), Explained Variance Score (EVS), Mean Absolute Percentage Error (MAPE), and Median Absolute Error (MedAE). Two datasets, water consumption and weather data, were collected from different sources to examine the performance of the ML algorithms. The novelty of this study lies in the integration of both weather and water consumption data. After examining the most effective features, the two datasets were merged and the proposed algorithms were applied. The RF algorithm outperformed the other models, indicating its robustness in capturing water usage behavior in dry areas such as Hail City. The results of this study can be used by local authorities in decision-making, water consumption analysis, new project construction, and consumer behavior regarding water usage habits in the region.