Groundwater Potential and Managed Aquifer Recharge Suitability Assessment using GIS-AHP and Machine Learning in the upstream part of Awash River, Ethiopia

Managed aquifer recharge (MAR) is a key strategy for enhancing groundwater storage, reducing overexploitation, and improving climate resilience in semiarid regions. However, its effectiveness depends on accurate identification of hydrogeologically suitable recharge zones. This study developed an integrated Geographic Information System–Analytical Hierarchy Process (GIS–AHP) and ensemble machine learning (ML) framework to assess groundwater potential and MAR suitability in the upstream part of the Awash River in Ethiopia. Eleven thematic layers were weighted using AHP to derive a Groundwater Potential Index (GPI). Geophysical parameters from vertical electrical soundings, transmissivity from pumping tests, and borehole yield data were used for the calibration and independent validation. Random Forest (RF), Support Vector Machine (SVM), and Extreme Gradient Boosting (XGBoost) models were trained using geophysical predictors and GPI, and combined into an ensemble MAR probability model. The AHP-derived GPI exhibited a good predictive performance against the observed borehole yield (receiver operating characteristic area under the curve, ROC–AUC = 0.731; precision–recall area under the curve, PR–AUC = 0.79). The ensemble ML model showed excellent agreement with the observed MAR conditions (ROC, AUC = 0.913; PR, AUC = 0.946). High to very high MAR suitability zones cover approximately 39% of the basin and are predominantly associated with fractured volcanic formations, moderate slopes, and high lineament density, whereas low-suitability areas correspond to steep terrain and urbanized regions. Integrating expert-based multi-criteria decision analysis with data-driven ML substantially improved MAR site identification under complex volcanic hydrogeological conditions. The proposed framework provides a robust and transferable decision support tool for basin-scale MAR planning.