Statistical Analysis of Data-set of Groundwater Locations of a 4 km2 VES Data-field at the Southern Phase II Development, Gidan Kwano Campus, Minna, Nigeria

Abstract

A limited-extent test of the validity of using a statistical tool in conjunction with acquired ground electrical resistivity data to build a protocol that could ease the efforts and cost that is put into acquiring labourious vertical electrical sounding (VES) data-field in a geologically-similar locality has been completed at the 4 km2 southern Phase II Development of the Gidan Kwano Campus, Minna, Nigeria. The arbitrarily-fixed very high correlation threshold chosen for that study ensured that the statistical tool under consideration was given a negative recommendation even though the fundamental mathematics of this method was sound and the prevailing geology was encouraging. The aim of this study is to expand the scope of that previous study so as to further examine the results of the simple regression analysis (SRA) method as applied to the corpus of the 4 km2 VES data-field for a more agreeable or flexible value of correlation threshold: the key objective of this study is achieved by setting the correlation threshold at the median boundary point of 50%. For each groundwater- prospect location, a table of acquired dependent variables (that is, resistivity values) for particular values of the independent variables (that is, AB/2, the current-electrode spacing) was drawn up; such a table is the "x and y table" where x corresponds to AB/2 and y corresponds to the appropriate column of resistivity values. Furthermore, the necessary statistical parameters associated with the SRA method were computed for each table. Subsequently, tables of correlations for the 57 definite groundwater prospect locations down to the 100 m depth-mark and down to the 40 m depth-mark were computed. Based on the "fair boundary point" threshold correlation of 50% used for this study, against the "tough boundary point" of 75% that was set for the previous study, a 42/57 or 73.684% positive correlation for y and y' of groundwater prospect locations down to 100 m for the 57 definite groundwater prospect locations means that the reliability of the simple regression analysis route is sufficiently excellent to be trusted. Also, based on this 50% threshold correlation, a 42/57" or 73.684% positive correlation for y and y' of groundwater prospect locations down to 40 m for the 57 definite groundwater prospect locations means that the simple regression analysis route can be used as a cost-cutting routine whereby maximum depths of survey of intervening prospect locations should be limited to just this 40 m. and then other values downward would be appropriately predicted before inputting into any purpose-specific interpretation software.