ASSESSMENT OF WATER POTABILITY AND ENVIRONMENTAL FACTORS USING MULTIVARIATE ANALYSIS

Abstract

Safe drinking water is fundamental to public health and environmental sustainability, yet water quality continues to be threatened by increasing anthropogenic pressures and natural environmental variability. This study assessed water potability and examined the influence of environmental factors using an integrated multivariate statistical framework. A quantitative cross-sectional design was adopted using a dataset comprising 1,500 water samples representing diverse environmental settings and water sources. Physicochemical parameters and environmental variables were analysed using descriptive statistics, Pearson correlation, Principal Component Analysis (PCA), Exploratory Factor Analysis (EFA), Hierarchical Cluster Analysis (HCA), and binary logistic regression. The results revealed that 68.2% of the water samples were classified as non-potable, indicating substantial variation in water quality. PCA extracted five principal components explaining 66.04% of the total variance, while EFA identified five latent factors representing industrial pollution, agricultural influence, mineral composition, topographic conditions, and microbial contamination. Hierarchical clustering grouped the samples into three distinct environmental clusters with different pollution characteristics. Logistic regression identified total dissolved solids, hardness, iron, manganese, and biological oxygen demand as significant predictors of water potability, with the model achieving a classification accuracy of 74.4% and an ROC-AUC of 0.770. The findings demonstrate that multivariate statistical analysis effectively identifies the dominant environmental determinants of water quality and provides a comprehensive framework for water potability assessment. The proposed analytical approach can support evidence-based environmental monitoring, pollution management, and sustainable water resource planning for improved drinking water safety.