Evaluation of Water Quality Parameters Using Multivariate Statistical Analysis (Case Study: Atrak River)

One of the issues considered in the statistical analysis of water quality variables is investigating the correlation between them. Canonical Correlation Analysis method expresses the relation between two sets of parameters by finding some linear combination of the first set of variables that have the most correlation with the linear combination of the second one. In this study, to supply a significant amount of water needed for Khorasan Razavi, Khorasan Shomali and Golestan provinces, the relation between physical and chemical variables was evaluated to provide an appropriate management plan. Five physical and six chemical variables were measured at 38 stations. Results showed that there is an effective correlation between the physical and chemical parameters and from five canonical varieties, only the first two categories are statistically valid according to P-value number. In general, results confirmed by the cluster analysis method indicated the effectiveness of linear combinations for EC and TDS. Chemical parameters are mostly caused by human activities and physical parameters caused by both human and natural sources. It can be concluded that the sources of both categories of parameters are human activities. Thus, It could be recommended to focus more on human activities in future water quality management plans and also control the physical parameters.