Multivariate statistical approaches to identify the major factors governing groundwater quality

Abstract

Abstract Multivariate statistical techniques, discriminant analysis, cluster and principal component analysis were applied to the dataset on groundwater quality of Longyan basin of Fujian Province (South China), to extract principal factors controlling the source variations in the hydrochemistry and identify the major factors affecting groundwater quality. The dataset covers ten parameters of monitored wells at five typical locations in the region. The results were evaluated in accordance with the groundwater quality standards suggested by Specification GB/T14848-93, “The Quality Standard of Underground Water.” Cluster analysis results reveal that the groundwater in the study area is classified into two groups (A: 2000–2007 and B: 2008–2011) between the sampling sites, reflecting regular characters of interannual variability. Factor analysis/principal component analysis, applied to the datasets of the two different groups obtained from cluster analysis, resulted in three factors accounting for 85.5% and 100% of the total variance in the water quality datasets, respectively. Three of the ten parameters processed by discriminant analysis obtained a conformation rate of 100% which allowed a reduction in the dimensionality of large dataset, and also it found that most discriminant parameters (total alkalinity, chloride ion, sulfate ion) are responsible for temporal variation of water quality. So this study illustrates the usefulness of multivariate statistical techniques for interpreting complex datasets of water quality, identifying pollution sources/factors for effective groundwater quality management.

Keywords

• Multivariate statistics
• Factor analysis
• Cluster analysis
• Groundwater quality