Influence of seasonality and ENOS oscillation on the quality of the waters of a tropical river: multivariate statistical methods

Abstract

The growing demand for freshwater in domestic, industrial, and food production demonstrates the current need to consider drinking water's quantity and quality. The water quality index (WQI) is widely adopted for synthesizing information on freshwater's various physical, chemical, and biological parameters to inform the lay public and guide water quality management actions. Thus, it is indispensable to analyze the parameters used to calculate it, and if necessary, to reassess if the parameters indicate the quality of the water. This study aims to evaluate if the parameters used in the WQI are influenced by climatic seasonality in the Rio Vermelho - MT basin. The data used in the survey provide information on the water quality parameters of four sampling stations throughout the basin between the years 2006 and 2017. A total of 126 individuals and 30 variables were surveyed. Because of the large number of variables considered initially necessary, it was decided to perform multivariate statistical analysis since solving such a question would not be possible using univariate analysis techniques. Multivariate analysis was conducted in the R. Among the multivariate methods, principal component analysis (PCA) and discriminant function analysis were used. The principal component analysis verified that the WQI is influenced by variations of climatic factors, which makes it questionable to characterize water quality only by analyzing the nine parameters determined by the WQI. Therefore, it is necessary to adapt these parameters to regions with well-defined rainy and dry seasons, such as the one studied here.