Answers are grounded in this paper and cite their sources.
Groundwater is a vital and limited natural resource, particularly in semi-arid and arid regions. This study aims at delineating groundwater potential zones in the Zahrez basin, Northern Algeria, using three machine learning (ML) models: Random Forest, CatBoost, XGBoost along with their ensemble using the Stacking approach. Eight predictive variables were considered in this study, namely; Rainfall, Soil Type, Geology, Normalized Difference Vegetation Index, Elevation, Slope, Drainage Density, Topographic Wetness Index and Topographic Position Index. The Receiver Operating Characteristic model and its Area Under the Curve (AUC) were employed to compare the predictive performance of the selected models. Results demonstrated the superiority of the Stacking ensemble model over CatBoost, Random Forest and XGBoost models, achieving AUC scores of 0.94, 0.92, 0.91 and 0.90%, respectively, demonstrating excellent predictive accuracy across all four models. In addition, their derived weights revealed that Geology and Slope were the most influential factors for groundwater potential zoning. The outcomes of this study highlight the potential of machine learning and their ensemble models in enhancing the delineation of Groundwater Potential Zones. Furthermore, they offer valuable insights to policy-makers and stakeholders regarding sustainable water resources management across the Zahrez Basin, especially in light of the current context of climate change and increasing water demand.
No takes yet. Share an insight, caveat, or question.
Bouhadeb et al. (2026) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: