The use of multivariate statistical analysis in the assessment of groundwater hydrochemistry in some parts of southwestern Nigeria

Abstract

Multivariate statistical methods including factor analysis (FA), cluster analysis (CA), and correlation analysis have been used to evaluate the spatial variations and the interpretation of a complex water quality data set of some parts of Oyo State in southwestern Nigeria. Thirty water samples were collected from different stations, and 16 parameters were determined. Correlation analysis shows that the relationship between the parameters with high character of ion was higher than that of the parameters with low character of ion and that the variation in relationship shows the complexity in groundwater quality and the effect of the interactions between rock and water. Regression analysis was used for the prediction of values of one variable using the knowledge of other variables, for which more data are available. FA shows five distinct factors, which explained 84.3% of the total variance in water quality data set. The five factors are anthropogenic, ion exchange, weathering/leaching, anthropogenic, and nitrogen, which explained 28, 23, 14.2, 10.0, and 6.9% of the total variance, respectively. Hierarchical cluster analysis grouped the parameters into three major clusters. This study shows the uses of multivariate statistical methods for the interpretation of complex data sets in the analysis of spatial variations in water quality. This would therefore enhance planning for future studies.