A comprehensive analysis, hydrogeochemical characterization and processes controlling surface water quality: entropy-based WQI, geospatial assessment, PIS, NPI, and multivariate approaches in Mahanadi basin, Odisha (India)
Year: 2025
Type: article
Abstract: • Addresses the critical challenge of surface water prediction, particularly in tropical monsoon regions. • The best combination of input parameters affecting the study area is selected using the PIS approach. • Employs Inverse Distance Weighted (IDW) method to estimate water quality data from 11 observed locations. • Utilizes Entropy and Multivariate networks for enhanced prediction accuracy. • The Bland-Altman approach is recommended based on the EWQI and PWQI, achieving more than 90% accuracy. Evaluating the water quality is one of the most important steps in creating surface water resources for drinking water consumption in the Mahanadi River Basin, Odisha. Surface water possesses a number of water quality problems, such as turbidity, dissolved solids, and coliform contamination, all of which are harmful to human health. The use of a Water Quality Index (WQI) is regarded as a successful technique for assessing the quality of water. In the present study, a total of 13 surface water samples for a duration of 2019–2024, were analysed in post-monsoon (POM) season for eleven monitoring locations. This study’s goal was to assess the performance of several indexing techniques such as Geographical Information System (GIS), Pollution index of Surface water (PIS), and Nitrate Pollution Index (NPI). In addition to the fixed-weight WQI, to predict the water quality, the Entropy (E) weighted-WQI, and Multivariate Statistical Models like PCA (Principal Component Analysis) factor based WQI assessments were used. The findings showed that HCO 3 − was the main anion and that magnesium (Mg 2+ ) dominated the cations. In almost 100 % of the surface water samples, the concentration of turbidity, TDS, and coliform, respectively, displays above the WHO-established acceptable limit. Depending on the PIS model, about 9.09 %, 9.09 %, 36.36 %, 18.18 %, and 27.27 % of examined water samples were categorized as insignificant, low, moderate, high, and very high polluted water quality. The EWQI values indicated that 72.73 % of samples exhibited poor or extremely polluted zone. It showed notable 18.18 % of samples, rated as medium for drinking. The entropy score revealed that three samples (27.27 %) were classified as potable for human consumption. Almost 3 samples (Z-1, 3, and 4) are unanimously defined as clean or light zone, across the NPI rating. Across 5 water quality locations corresponds to significant/very significant pollution, with a nominal agreement of 45.45 %. A few tests have extremely concerned results, necessitating quick action to ensure safe drinking water. The PCA analysis extracted four factors for the studied datasets within eigen values > 1, representing 87.602 % of total variance. The PCA loadings were categorized as strong, moderate, and weak, respectively, accordingly to the loading values of > 0.75, 0.75–0.5, and 0.50–0.30. Observation depicts that variable in PC1 namely Cl − (0.75), SO 4 2− (0.87), NO 3 − (0.78), turbidity (0.89), and coliform (0.85) were the most critical parameters affecting surface water chemistry through anthropogenic and natural occurrences. Based on the findings, the Bland-Altman plot utilized to confirm and reaffirm the validity of the current approach for surface water characterisation by comparing it with PWQI. The results showed that the error chance is very small. The north-eastern and south-eastern regions were found to be severely contaminated by agricultural and human-induced activities, according to spatial analysis. The presence of substantial levels of turbidity, magnesium, and coliform in surface water in the majority of the samples, together with their dependency, may lead to future pollution through the dissolution of parameters. At relevant places of the studied region, its water resource is under threat from urbanization, agriculture, industrialization, and population growth. Moreover, the resultant WQI model produces better forecast accuracy with fewer input parameters and makes a substantial contribution to the sustainable management of surface water resources in arid regions.
Source: Water-Energy Nexus
Author: Abhijeet Das
Institution: C.V. Raman Global University
Language: English
Subfield: Water Science and Technology
Field: Environmental Science
Domain: Physical Sciences
Open Access status: gold
Estimated APC paid: $900
OpenAlex ID: W4413915957