Assessment of Groundwater Quality in Sri Lanka Using Multivariate Statistical Techniques

Abstract

Spatial variations in groundwater quality of Sri Lanka were evaluated using multivariate statistical analysis techniques. Sampling was conducted in 1262 sampling wells distributed in 25 administrative districts and the respective climatic zones in Sri Lanka. Concentration of 18 water quality parameters including natural mineral ions and heavy metal ions was measured. Sampled wells were clustered into two different clusters based on the attributes of the groundwater quality parameters. Groundwater wells located in dry and intermediate zones, which were rich in mineral ions due to rock–water interactions and lower recharge rates, were clustered into cluster 1. Shallow aquifers in the wet zone and coastal areas and wells located on alluvial aquifers were clustered into cluster 2. Groundwater well clustered under cluster 2 had high average heavy metal concentrations compared to cluster 1. Discriminant analysis (backward stepwise mode) reduced the number of discriminating parameters into ten, but they were a mixture of ions dissolved in groundwater because of anthropogenic activities and natural interactions both. Further, six factors which may influence on groundwater quality of Sri Lanka were identified in the factor analysis. The main factors were natural rock–water interactions, pollution caused by industrialization and waste discharge, agrochemical usage and infiltration of storm water rich in lead into shallow aquifers.