Water quality analysis of Godavari river basin using multivariate analysis techniques

Multivariate statistical techniques, including cluster analysis, principal component analysis factor analysis and discriminant analysis, have been used to evaluate spatial variations and to interpret a large and complex water quality data set collected from the Godavari river basin. The data sets, containing 7 parameters, were generated during the 3-years (2007-2009) at 78 different sites along the river and its tributaries. Water quality indices based on four parameters (pH, DO, BOD and FC) calculated for all the sites were found to be medium to good, good to excellent and bad using modified NSF index. Three significant groups (cleaner, slightly and moderately polluted sites) were detected by CA method, and three latent factors were identified by PCA method. The results of DA revealed that only two parameters (i.e. pH and BOD) were necessary for analysis in spatial variation. 83.3% of the original sites were correctly.classified using discriminant function developed from the analysis.