I used multivariate statistical methods, including cluster analysis (CA), discriminant analysis (DA) and principal component analysis (PCA) to evaluate water quality in the Ying River Basin, the largest tributary of Huai River, China. A total of 12 water quality parameters were measured at each of 15 sites from 2008–2010 (540 observations), allowing investigation of temporal and spatial variation and indication of potential pollution sources. Hierarchical CA classified the 15 monitoring sites into three groups, representing heavily, moderately and least polluted sites. Three parameters (temperature, pH and TP) distinguished temporal variation with close to 67.4% correct assignment in the DA, separating summer from winter and spring-fall. In the spatial variation analysis, the DA used eight parameters (temperature, pH, DO, CODMn, CODCr, BOD5, NH4-N, and Hg) and correctly assigned about 85.7% of the sites to spatial clusters. PCA did not result in a significant data reduction in this study, but it did extract and identify significant factors/variables responsible for variation in river water quality at the three groups of sites identified by CA. Sites in Group 1 were mostly correlated with CODCr, NH4-N and volatile phenol, suggesting that they received pollutants mainly from industrial discharge. Group 2 sites correlated most strongly with temperature, pH and DO, which may indicate that these sites were mainly affected by natural processes. Group 3 sites were dominated by CODMn, As and Hg, perhaps indicating influence by both point and non-point pollution sources.
【 预 览 】
附件列表
Files
Size
Format
View
Assessment of Water Quality Using Multivariate Statistical Techniques in the Ying River Basin, China