Mapping and predicting groundwater accumulations using remote sensing and aeromagnetic data: a case study from Bahariya Oasis, Western Desert, Egypt

Abstract

The Bahariya Oasis in Egypt’s Western Desert faces critical water security challenges due to intensive groundwater exploitation and limited renewable recharge. To address this challenge, this study integrates a GIS-AHP model for the first time in the Bahariya area and incorporates high-resolution aeromagnetic data to enhance the prediction of groundwater potential zones. Nine groundwater-controlling thematic layers were compiled from multi-source datasets, classified, and weighted using the AHP model: structural geology (lineament density), terrain (slope), hydrology (drainage density and rainfall), hydrogeology (lithology and soil moisture), land-surface conditions (NDVI and LU/LC), and the RTP aeromagnetic signal. The resulting groundwater potential map delineates coherent “very-high” potential corridors concentrated in the southern and southeastern sectors, where high lineament density, low-relief topography, and permeable Quaternary alluvial deposits coincide and correlate with known productive wells, suggesting favorable conditions for enhanced recharge and storage. Potential patterns also show strong spatial correlation with the drainage network and mapped structural trends. Model validation against well occurrences yields an AUC of 93.4%, indicating excellent predictive accuracy. Aeromagnetic interpretation further resolves the subsurface structural architecture and constrains the basement configuration, revealing a marked deepening of the basement toward the south and southeast. This is where a thick sedimentary cover suggests enhanced groundwater storage potential, providing additional support for the groundwater potential zoning results. This integrated workflow offers a robust tool for groundwater prediction and planning, with direct utility for prioritizing drilling locations and supporting sustainable water resources planning in arid environments.