Advanced statistical analyses on source water quality of Macau

Abstract

Drinking water quality is directly related to the health of the residents living in Macau. This study assesses the water quality of two reservoirs in Macau, i.e., Main Storage Reservoir (MMR) and Seac Pai Van Reservoir (MSR). The water quality data were evaluated via a developed Water Quality Index (WQI) approach coupled with two multivariate techniques of Principle Component Analysis (PCA) and Pearson’s R Correlation. Furthermore, classification models including Logistic Regression, Decision Tree, and Random Forest were adopted for water quality analyses in the study. The WQI scores obtained during the period from January 2016 to April 2019 are determined to be mostly above 76, which is indicative of a desirable water quality. WQI with PCA and Pearson’s r correlation can provide a holistic view of water quality during the studied period. As for the performance of classification models, these three models can predict the class of algal population based upon the given water quality parameters, while Random Forest results outperform the other two. In addition, the Decision Tree and Random Forest methods are able to identify the significant water quality parameters through predictive studies. The study carried out an overview of Macau water quality and provides a deeper insight on the interaction between algal population and water quality in MMR and MSR. The benefits of this study include improvement of water management efficiency and protection of water sources to the public.