Entropy-Based Integration of Geological and Geoelectrical Parameters for Groundwater Vulnerability Assessment in Oke-Aro, Akure, Nigeria

Background and Objective: Groundwater resources in basement complex terrains are susceptible to contamination due to the heterogeneous nature of subsurface layers. Oke-Aro community in Akure, Southwestern Nigeria, faces increasing groundwater usage without adequate vulnerability assessment. This study aimed to evaluate the aquifer vulnerability in the area to determine groundwater potability by integrating geologic and geo-electrical parameters using an entropy-based data-driven model. Materials and Methods: A total of 30 Vertical Electrical Soundings (VES) were conducted using the Schlumberger array with AB/2 spacing between 65 and 100 m. Additionally, 30 static water level (SWL) measurements were taken from hand-dug wells to validate the model. The VES data revealed 3-4 geoelectric layers correlating to topsoil, weathered layer, partially fractured basement, and fresh basement. Seven conditioning factors-lithology (L), slope (S), longitudinal resistivity (LR), overburden thickness (OLT), transverse resistance (TR), transverse resistivity (ρt), and coefficient of anisotropy (COA)-were derived and weighted using the entropy method to produce a groundwater vulnerability map (GVM). Model validation was conducted by overlaying SWL data onto the GVM. Results: The GVM classified the study area into four vulnerability zones: Very low, low, medium, high, and very high. The validation using SWL data yielded a model accuracy of 56.25%. Areas with high vulnerability corresponded with shallow water levels, indicating potential susceptibility to surface contamination. Conclusion: The study demonstrates the applicability of entropy-based models in groundwater vulnerability assessment within complex geologic settings. Despite moderate validation accuracy, the approach offers a valuable framework for preliminary vulnerability mapping. Future studies should incorporate additional hydrochemical and land use data for improved accuracy.