Modeling of aquifer potentiality using GIS-based knowledge-driven technique: a case study of hard rock geological setting, southwestern Nigeria

The spatial prediction of how aquifers store and transmit sufficient groundwater in subtle hydrogeological conditions typical of crystalline basement complex is quite challenging. This study explored diverse methodologies including remote sensing, geophysical and knowledge-driven data mining model for possible solution. The acquired geophysical, remote sensed, ancillary, and climate data were processed applying geophysical and geospatial software to determine both subsurface geophysical parameters (SGPs) and surface hydrological parameters (SHPs). Five aquifer potentiality conditioning factors (APCFs) based on the results of the interpreted SGPs and SHPs were derived. The produced APCFs maps were assigned suitable weights using the standard Saaty’s scale in the context of analytical hierarchy process (AHP) data mining technique. An aquifer potentiality prediction index (APPI) data mining model was developed for integrating the APCFs maps to compute the APPI values in the range of 1.24–4.37 for the study area. Based on the estimated APPI results, the aquifer potentiality prediction zone (APPZ) map of the area was produced in GIS environment. The map revealed that about 76% of the areal extent account for the very low, low, and medium predicted aquifer potential classes and 24% of the area covers the moderate-high to high aquifer potential classes. The prediction accuracy of the produced APPZ map was established via statistical analysis of Borehole yield rate and Geoelectrical Parameters in-situ Modeling validations results. The results of the study established the new approach of modeling SGPs and SHPs to inform decision-making process for locating appropriate positions of new productive wells in the study area.