Answers are grounded in this paper and cite their sources.
Surface water resources are essential for drinking water supply, irrigation, industry, and aquatic ecosystems. Continuing pollution necessitates effective monitoring for sustainable management. Water quality indices (WQIs) combine physical, chemical, and biological measurements into a single score, but methods vary across regions and applications. This review examines ten WQI frameworks across five development stages: parameter selection, sub-index generation, weighting, aggregation, and classification. These methodological percentages describe only the ten selected frameworks reviewed here. Three main limitations emerge. First, expert judgment introduces subjectivity into parameter selection and weighting. Expert-based weighting was used in 40% of the reviewed frameworks. Reproducing these weights requires clear reporting of expert panels and weight derivation and normalization procedures. Second, aggregation may mask poor individual parameter values. Additive methods allow substantial compensation, while multiplicative methods reduce, but do not eliminate, compensation and may be strongly influenced by a very low sub-index. None of the ten frameworks used logical aggregation. The minimum operator prevents compensatory masking but remains sensitive to uncertainty in the lowest sub-index. Third, classification differences limit comparisons across regions. Principal component and cluster analysis, fuzzy logic and machine learning, may reduce reliance on expert judgment. However, performance depends on data and evaluation procedures. For example, the 99.9% accuracy reported for extreme gradient boosting (XGBoost) in one classification study should not be generalized. Data-driven approaches may shift, rather than remove, subjectivity. Transparent and adaptable WQIs should support cross-regional comparison while retaining local relevance. These indices should complement, rather than replace, regulatory assessments under the European Union Water Framework Directive and United States Clean Water Act.
No takes yet. Share an insight, caveat, or question.
Gbilly et al. (2026) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: