Water quality index modified by using multivariate analysis: a case study of Pelotas Stream, RS, Brazil

ABSTRACT This research aimed at developing a modified water quality index (WQIMod) by using multivariate analysis, from a comparison with the WQI used by the Environmental Company of São Paulo (Companhia Ambiental do Estado de São Paulo-Cetesb) (IQACETESB), to be applied in the Pelotas Stream, RS, Brazil. Secondary data from Pelotas Stream surface water quality were used, collected in the sampling site GER 48, located in the municipality of Pelotas, where they were monitored and provided by the State Foundation of Environmental Protection Henrique Luís Roessler (Fundação Estadual de Proteção Ambiental Henrique Luís Roessler-FEPAM) between 1997 and 2013. Among the multivariate analyses, the factor analysis/principal component analysis (PCA) was used. The results indicated that the PCA extracted two principal components, which explained 100% of the total data variance in the correlation matrix, while the factorial scores for WQIMod calculation were extracted from the first factor, which explained 70.6% of the total data variance. We concluded that the WQIMod exhibits biochemical oxygen demand (BOD5), turbidity (TH) and total solids (TS) as the most important variables for the index construction, differing from the WQICETESB, which shows the variables dissolved oxygen (DO), thermotolerant coliforms (TC) and hydrogenionic potential (pH) as the most relevant ones, indicating the need of WQICETESB calibration for applying in this specific ecosystem.