Application of geochemical modelling and multiple regression analysis to reassess groundwater evolution in Kaduna Basin, NW Nigeria

Geochemical modelling (GM) and multiple regression analysis (MRA) are applied to reassess the evolution of groundwater to understand and estimate groundwater's quality and chemical changes. These methods provide a comprehensive procedure for evaluating and managing groundwater sources, especially in identifying the pollution sources, predicting upcoming changes and designing strategies for remediation. This review reassesses groundwater's hydrochemical evolution in the Kaduna Basin using GM and MRA. Hydrochemical data were obtained from the literature using a systematic approach and were subjected to GM and MRA. The geochemical modelling results showed that aquifers are saturated with Anhydrite, Chlorite, Chrysotile, Dolomite, Gypsum, Halite, K-Mica, Pyrite, Quartz, Dolomite, and Goethite. The Na2O-Al2O3-SiO2-H2O system indicated that the studied locations fell in Na-montmorillonite and Kalinite fields. However, the K2OAl2O3-SiO2-H2O system showed that groundwater is stable with Montmorillonite, Kaolinite, and K-Feldspar. MRA and Na+/Cl− molar ratio revealed that rock weathering significantly controls groundwater. The Na+/Cl− values are above 1 in 83.33%, 54.84%, and 26.27% sampling locations in the Kudenda-Nassarawa area, Kaduna South, and Kakuri and its Environs, respectively. The overall p-value was less than 0.01 and R-Sq = 63%, suggesting that all elements significantly influence aquifers' hydrochemistry. Results of geochemical modelling, Na+/Cl− molar ratio, and MRA are concurrent as they both show silicate weathering as the factor controlling groundwater hydrochemistry. Thus, geochemical and multiple regression analysis offers a reliable and user-friendly tool for assessing aquifers' hydrochemistry. We hope this review's findings will stimulate other researchers to an analogous procedure in a forthcoming study on hydrochemical analysis, especially those from semi-arid environments.