Assessing groundwater potential in the western High Atlas (Morocco): An integrated approach based on remote sensing, GIS techniques, frequency ratio, and Shannon entropy

Abstract

This study aims to map groundwater potential zones in the mountainous and fractured region of Amizmiz, which is characterized by a semi-arid climate. The combined effects of climate change, increasing aridity, and geological complexity place growing pressure on groundwater resources, making their sustainable management a critical challenge. To address this issue, a multi-criteria methodology was developed by integrating fifteen geological, topographic, climatic, and hydrological factors to optimize drilling site selection and enhance the chances of locating productive aquifers. Two statistical modeling approaches, Shannon’s Entropy (SE) and Frequency Ratio (FR), were applied. A spatial database was constructed using remote sensing and Geographic Information System (GIS) techniques, incorporating data from 150 water points along with relevant groundwater-controlling factors. The dataset was divided into 70 % for model training and 30 % for validation. Two groundwater potential maps were produced and classified into five potential levels. Results indicate that zones identified as having high potential correspond closely with most productive water points. The SE model (Rate Success = 0.86) showed smoother spatial continuity, while the FR model (Rate Success = 0.88) effectively highlighted several favorable zones, albeit with sharper spatial transitions. To reinforce validation, an apparent electrical resistivity map was used, revealing a significant concordance between conductive zones and areas of high groundwater potential: 85 % for the FR model and 89 % for the SE model. These findings confirm the reliability of the generated maps for guiding future drilling. Future studies should integrate hydro-dynamic parameters and explore machine learning techniques to further enhance prediction accuracy.