Assessment of water quality using principal component analysis: a case study of the Marrecas stream basin in Brazil

Abstract

Monitoring water quality is a fundamental process to ensure proper anthropogenic usage and environmental protection of this resource. This study collected monthly measurements of 9 parameters (pH, Temperature, BOD, Total Solids, Thermotolerant Coliforms, Dissolved Oxygen, Total Nitrogen and Total Phosphorus) in 5 sampling stations along the Marrecas water stream, during a 1-year period. Temporal and seasonal variations were analyzed and interpreted for each element, explaining how specific geographical and anthropogenic factors affected the water body. Principal Component Analysis (PCA) was applied to evaluate each element’s correlation and to reduce the number of parameters, easing the assessment of water quality for each location. Results were followed by the creation of an improved index for the region, which could better estimate the quality of water, only considering 4 of the original parameters. It was also recognized that each water body possesses several subtleties that impact on how its water quality should be measured and indexed into a single value, which validates the case for the creation of regional WQI’s.