Monitoring and assessment of Deepor Beel water quality using multivariate statistical tools

Abstract

AbstractThe aim of this study was application of multivariate statistical techniques – e.g., hierarchical cluster analysis (HCA), principal component analysis (PCA) and discriminant analysis (DA) – to analyse significant sources affecting water quality in Deepor Beel. Laboratory analyses for 20 water quality parameters were carried out on samples collected from 23 monitoring stations. HCA was used on the raw data, categorising the 23 sampling locations into three clusters, i.e., sites of relatively high (HP), moderate (MP) and low pollution (LP), based on water quality similarities at the sampling locations. The HCA results were then used to carry out PCA, yielding different principal components (PCs) and providing information about the respective sites' pollution factors/sources. The PCA for HP sites resulted in the identification of six PCs accounting for more than 84% of the total cumulative variance. Similarly, the PCA for LP and MP sites resulted in two and five PCs, respectively, each accounting for 100% of total cumulative variance. Finally, the raw dataset was subjected to DA. Four parameters, i.e., BOD5, COD, TSS and SO42− were shown to account for large spatial variations in the wetland's water quality and exert the most influence.