Aquifer evaluation in parts of north-central Nigeria from geo-electrical derived parameters

In order to reduce the level of risk associated with borehole drilling, it is important to have detailed knowledge about the aquifer distribution. In a view to generating groundwater potential model of Lokoja and its environs, the detailed subsurface characterization was carried out using a GIS-based multi-criteria decision analysis approach. One hundred and twenty-four vertical electrical sounding (VES) data points were covered within the study area using the Schlumberger array of electrical resistivity surveys. Hydrogeological investigation of one hundred and twenty-four existing boreholes within the vicinity of the sounded VES points was carried out by measuring in situ parameters of each borehole such as borehole depth, elevation, static water level and borehole yield. A Multi-Criteria Decision Analysis-Analytical Hierarchy Process (MCDA)-AHP-based was carried out by estimating the probabilistic ratings for the classes of parameters used for modeling groundwater potential. Four groundwater potential influencing factors, namely coefficient of anisotropy, transverse resistance, aquifer resistivity and aquifer thickness, were classified and rated. The output of the multi-criteria decision analysis was processed in the GIS environment to produce a groundwater potential index map. The obtained model was validated by comparing it with in situ borehole yield data to determine the accuracy of the proposed model. The groundwater potential map generated classified the study area into low, medium and high yield zones. Areas with medium potential zones dominate the largest part of the with 66% area coverage, and the dominance of these zones was visible in the northern and western part of the study area. Areas with high groundwater potential exist toward the southern and eastern sections of the study area. This area was observed to be underlined with sandstone, siltstone and migmatite. The validation exercise carried out on the proposed model reveals a 70% prediction accuracy.