Ground Water Quality and Multivariate Statistical Methods

Abstract

In this paper, an attempt was made to study the physico-chemical properties of ground water of the Kozhikode district, Kerala, India, by applying multivariate statistical methods on samples collected from various parts of the study area. Combining principal component analysis and multiple linear regression (MLR), we developed a regression model for predicting total dissolved solids (TDS) in terms of calcium, magnesium, nitrate, sodium, chloride, potassium, bicarbonate and sulfate. This study revealed that statistically, calcium is the most significant component of TDS in the study area. The relevance of the regression model with respect to experimental data was further evaluated by applying structural equation modeling (SEM).