An Integrated Assessment of Spatial Variability Mapping of Surface Water Properties and its Impacts on Drinking Water in Mahanadi River Basin, Odisha, India

Greater precision in models to aid in the prediction of drinking suitability of surface water is a novel concept that will benefit future studies in the target region, and the methods can be adapted to other locations. However, agricultural activities are the primary source of surface water pollution as a result of rising fertilizer and pesticide use. The Mahanadi River region in Odisha, India, is one such area where the use of fertilizers is causing the surface water quality (WQ) to deteriorate. The demand for precision in prediction models for assessing the suitability of surface water for drinking and irrigation is a growing focus that promises to enhance future studies and applications in various regimes. Hence, this study intends to identify and assess the sources of WQ contamination in the area using an entropy-based approach, self-organizing map (SOM) with cluster analysis, and weighted linear combination (WLC). The results were presented by employing overlay analysis using the inverse distance weighted (IDW) interpolation technique, due to their high accuracy in spatial analysis and ability to integrate various data types, which ensures comprehensive and reliable results. Most often, the IDW tool offer advanced visualization capabilities and user-friendly interfaces, making it ideal for creating detailed maps of surface water quality and drinking suitability. Twenty-one water samples were collected from 19 different locations to analyze the physicochemical parameters for a duration of 7 years (2016–2023). In the proposed area, the prominent cation trend was Fe2+ > B+, where Fe2+ is the primary cation. The predominant anion was arranged in the sequence Cl− > SO > NO > F−, where chloride is the dominant anion. As per entropy classification, the entropy weightage values were computed to be in the range of 0.15 to 4, and the percentage of sites where surface water was not a good source of drinking were estimated to be 47.36%. In addition, IDW detected nine severely polluted zones within the study region with good accuracy. It is noticeable that the water showed higher values of TC, TKN, EC, TDS, Cl−, and SO as compared to other indicators. Here, the water quality is poor due to runoff from nearby landfill dumps, fertilizer waste, or contamination from animal or human waste. In addition, to further ascertain and combine the known sources of ions found in surface water, SOM was performed for the hydrochemical data. The results exhibit three groups based on low, moderate, and high pollution zones. In this research field, comparing the ion concentrations that have been measured, five parameters each belong to Clusters I and II, while Cluster III contains about 10 parameters. Finally, three groups, suggested by SOM, point towards river bank erosion, water mineralization, and untreated sewage. It is clear from the WLC values that river-water quality ranges from low suitability to unsuitable class with the former varying from 0.5 to 154. Relatively lower values were recorded at sampling points P-(8), (9), and (19) in the river water, ascribed to the significant amount of human activity (anthropogenic), including sewage intake, trash dumping, and agricultural runoff. However, the analysis showed that in order to safeguard surface water sources from the dangers of contamination, nine areas require some level of surface water treatment prior to use. The outcomes of this study will hold significant worth for future research as it will reduce the duration and expenses of analysis through the utilization of the algorithms discussed here for predicting the suitability of surface water for drinking. The current study can be applied both locally and globally using similar methods.