Sci Rep. 2025 Sep 30;15(1):34101. doi: 10.1038/s41598-025-19539-w.
ABSTRACT
The prediction and management of water quality are critical to ensure Sustainable water resources, particularly in regions like Odisha, where rivers face increasing pollution from industrialization, agriculture, and urban expansion. The Brahmani River, located in the Odisha State, is the 2nd largest watershed in the province, by which its water quality is affected by natural and anthropogenic changes. In this research, water samples were gathered throughout the monsoon season for four years (2020-2024) from previously selected 7 sampling stations. Geographical Information System (GIS) techniques were used to find out the distribution of surface water quality on land use pattern. A particular focus is given to hybrid models that integrate multiple approaches to improve predictive accuracy and robustness. Therefore, the study was undertaken by incorporating Weighted Arithmetic (WA) Water Quality Index (WQI), Synthetic Pollution Index (SPI), Nemerow Pollution Index (NPI), Overall Index of Pollution (OIP), multivariate statistical method, namely Factor Analysis (FA) or Principal Component Analysis (PCA), and Multi-Criteria Decision-Making (MCDM) approaches like Evaluation based on Distance from Average Solution (EDAS). The goal of this investigation is to evaluate the water’s purity and whether it is Suitable for consumption. Fifteen physicochemical parameters were tested from 7 observation stations. Referring to the present research, the obtained order of anionic abundance was SO42- > Cl– > NO3– >F– > PO43–. However, the order of cationic abundance was Ca2+ > Mg2+ > Na+ > K+. The calculated WA-WQI values ranged between 49 and 72. Toxic heavy metals, nutrients, and microorganisms were the major pollutants influencing water quality, as stated by WA-WQI. In addition, the data was interpreted using pollution indices such as SPI (0.31-0.68), NPI (6-29.91), and OIP (0.45-4.40). By results, it is concluded that mainly 4 sites are unsuitable for drinking and irrigation purposes, due to long-term use of waste water, anthropogenic activities, over-extraction of Surface water and changes in land use pattern. Using the multivariate technique, the PCA method was useful to identify two latent pollution sources, that correctly assign 89% of the total variance in the dataset. During the first component, the major loadings on parameters: TDS, EC, alkalinity, Na+, Ca2+, Mg2+, K+, F–, Cl–, NO3–, and SO42–. It indicates that locations were primarily Harmed by oxygen-consuming organic and Hazardous contamination. Furthermore, the EDAS score fluctuated between 0.01 and 0.97. The results revealed that Y-(1) mentioned high polluted water, followed by Y-(2) and Y-(7). This signifies the existence of dissolving biological material; nitrogen was the major pollutant, originating primarily from anthropogenic local contamination. Later on, the outcomes of water quality parameters on the different indexing methods were evaluated, and the obtained outcomes indicate that the highest mean effective weight value belongs to the TDS, EC, Cl–, SO42- and PO43–, respectively. Notably, effective control of point source pollution and upper river ecological restoration should be done to improve the water quality and protect the reservoir. This research identifies key research gaps and proposes future directions for developing transparent, adaptive, and accurate models. The findings can also guide researchers and policymakers towards the development of smart water quality management systems that enhance decision-making and ecological sustainability.
PMID:41028073 | DOI:10.1038/s41598-025-19539-w
