Assessment of water quality observation stations using cluster analysis and ordinal logistic regression technique

Abstract

Variations in the surface water quality parameters of the observation stations have been investigated using cluster analysis and the ordinal logistic regression technique. Ordinal logistic regression analysis results showed that nitrite, ammonium and total phosphorus were identified as the parameters that displayed essential variations. These results were supported by the cluster analysis. The cluster analysis of the observation stations identified two clusters. It was concluded that domestic, agricultural and industrial discharges strongly affected the northwest part of the Tahtali dam reservoir. Finally, it was determined that the variations in the water quality parameters could reliably be estimated by these techniques.