Multivariate Statistical Analysis for Water Quality Assessment: A Review of Research Published between 2001 and 2020

Abstract

Research on water quality is a fundamental step in supporting the maintenance of environmental and human health. The elements involved in water quality analysis are multidimensional, because numerous characteristics can be measured simultaneously. This multidimensional character encourages researchers to statistically examine the data generated through multivariate statistical analysis (MSA). The objective of this review was to explore the research on water quality through MSA between the years 2001 and 2020, present in the Web of Science (WoS) database. Annual results, WoS subject categories, conventional journals, most cited publications, keywords, water sample types analyzed, country or territory where the study was conducted and most used multivariate statistical analyses were topics covered. The results demonstrate a considerable increase in research using MSA in water quality studies in the last twenty years, especially in developing countries. River, groundwater and lake were the most studied water sample types. In descending order, principal component analysis (PCA), hierarchical cluster analysis (HCA), factor analysis (FA) and discriminant analysis (DA) were the most used techniques. This review presents relevant information for researchers in choosing the most appropriate methods to analyze water quality data.