Development of a health risk-based weighted water quality index using multivariate statistical analysis: a case study from Taichung's Dajia River Basin, Taiwan

Conventional water quality indices (WQIs), such as the Canadian Council of Ministers of the Environment Water Quality Index (CCME-WQI), typically assign equal weights to all parameters - a practice that may obscure the health relevance of pollutants with chronic toxicity. This study proposes a health risk-weighted variant of the CCME-WQI that integrates multivariate statistical analysis and toxicological criteria to enhance public health responsiveness. Using long-term monitoring data from the Dajia River Basin in Taichung, Taiwan (2003-2024), 30 physical and chemical parameters were analyzed using principal component analysis and factor analysis to reduce redundancy and identify key indicators. Ten core parameters were selected based on statistical contribution and health risk thresholds, including hazard quotient (HQ > 1), cancer risk (CR > 10^-4), and IARC classifications. Risk-based weights were assigned accordingly. Seasonal validation showed strong agreement between the optimized and original CCME-WQI models (RMSE = 7.72; <i>p</i> = 0.610), while improving sensitivity to high-risk contaminants such as arsenic, lead, and cadmium, particularly for children. The proposed framework offers a scalable and resource-efficient tool, making it suitable for both centralized and decentralized water quality management contexts, while supporting health-informed monitoring and contributing to Sustainable Development Goal 6.