<p>Correct and sustainable water management represents a major challenge for humanity. Developing friendly and effective tools are pivotal steps to assess water quality and quantify the environmental impact of different compounds affecting surface waters over space and time. However, existing assessment frameworks often lack spatial integration with decision-support tools and fail to capture the dynamic variability of water quality in regions affected by rapid industrialization and climatic fluctuations. Here, we propose a novel approach to characterize and interpret surface (S) water quality (WQ) using a Geographical Information System (GIS) environment, integrated with optimized decision-making techniques, specifically tailored for the Mahanadi River and its distributaries in the Paradip area of Odisha state. This study’s primary goal is to use cutting-edge methods to map and assess surface water potential zones. In this assessment, Weighted arithmetic (WA), entropy (E) -Water Quality Index (WQI), Fuzzy-Synthetic Evaluation (FSE), GIS, and the Machine Learning Ensemble Ranking (ER) modeling technique were used to assess the water quality. All these methods help to identify the most critical factor in controlling SWQ for potable water. For this study, a three-year dataset (2020–2023) comprising 12 water quality parameters, collected from 9 observation sites during the monsoon season, was utilized. SWQ assessment using the WAWQI offers a comprehensive and standardized evaluation, with index values ranging from 40.36 to 176.13 and an average of 90.16. At 33.33% of sampling sites, elevated BOD and TH levels indicate pollution from sewage, industrial discharge, and runoff, posing serious health risks, degrading soil quality, and disrupting aquifer recharge and long-term water sustainability. Considering the E-WQI scores, the drinking water was differentiated into excellent, good, average, poor, and extremely poor categories for human consumption. In addition, the measured E-WQI values were between 67 and 266, with a typical average of 145.89. Regarding all the samples of surface water, 22.22% were framed into good class and remaining, 77.78% were unfit for drinking purposes. More than 100% of the water samples for the FSE-WQI assessment fell into the poor (22.22%), very poor (44.44%), and unsuitable categories (33.33%), proving that the surface water is unfit for human consumption. Additionally, none of the river water samples have excellent or good water quality (FSE-WQI &lt; 100). Therefore, both indices classified the river water is deteriorated at 7 sites (FSE-WQI) and 3 sites (E-WQI), with the inclusion of BOD, TDS, F^-, and PO₄^3-, that determined the river’s water to be unfit for human consumption throughout. Utilizing the water quality classification, in light of Ensemble Ranking (ER) method, the observed range is found to be 0.50 to 5.61. The lowest ranking value of 0.56 was recorded at H-(2), which suggests exceptionally outstanding water quality. The water quality values at H-(6), (8), (9) and (7) were 5.61, 4.87, 3.56 and 3.50 respectively, which indicates poor water quality. However, this apparent stability in the context of an increasing anthropogenic pressure suggests that environmental protection policies have at least mitigated further deteriorations. The study’s findings can serve as baseline data for SWQ development initiatives, and remedial measures are also recommended to improve surface water quality. They also give water researchers and local governing bodies baseline data to create integrated, sustainable water management plans tailored to the unique requirements of the area.</p> Graphical Abstract <p></p>