This study presents the usefulness of the water quality index (WQI) based on Fuzzy (F)-analytic hierarchy process (AHP), multi-criteria decision-making technique, namely, weighted sum approach (WSA) and machine learning models such as Borda scoring algorithm (BSA) for its evaluation and were further applied to the datasets on water quality (WQ) of the Mahanadi River (Odisha), generated during 5 years (2018–2023) of monitoring at 19 different sites for 20 parameters. The results render two parameters, namely coliform and TKN, exceeding the WHO standards. The results revealed that 52.63% of surface water samples are excellent in terms of drinking WQ, 26.32% of the samples are categorized under medium, and rest 21.05% are grouped under poor/very poor/unsuitable in terms of the F-AHP WQI. According to the results of WSA, 10 samples (52.63%) are low polluted zones, 6 samples (31.58%) are medium-polluted zones, and around 15.79% (3 samples) are highly polluted. The graphic representations obtained by BSA underline that the calculated value ranged between 15 and 256, stating in a zone of good to poor WQ. The best WQ was observed in T-(1), (5), (14), (15), (16), (17), and (18) because there were no changes in land use.

  • The quality of the surface water is characterized by the assessment of physiochemical parameters in the water.

  • AI methods, such as BSA and WSA, have been illustrated and results have been pointed out.

  • WQI is used to identify the water suitable for consumption.

  • By recognizing the probable source of water contamination, policymakers can take the appropriate action.

This content is only available as a PDF.
This is an Open Access article distributed under the terms of the Creative Commons Attribution Licence (CC BY 4.0), which permits copying, adaptation and redistribution, provided the original work is properly cited (http://creativecommons.org/licenses/by/4.0/).