Climate Change Impacts on Water Resources in Odisha, India

Strategic Insights -

Research conducted by the Department of Civil Engineering has demonstrated the significant effects of climate change on water resources in the Brahmani Basin in Odisha, India. The study highlights the need for sustainable management of surface water and the application of machine learning models to improve water quality forecast accuracy. The research used a dataset collected from 2019 to 2025 and employed a multi-criteria decision-making methodology to prioritize water resource management sites. The findings indicate that the water quality in the region is threatened by anthropogenic stressors, including pollution from domestic and industrial waste. The study recommends the use of entropy-based classification models and support vector machines to predict water quality and prioritize management strategies.

Key Takeaways:

  • The Brahmani Basin in Odisha, India, is experiencing a growing threat to its natural water supply due to climate change and urbanization.
  • The study used a dataset collected from 2019 to 2025, comprising 15 physicochemical parameters, to analyze the hydrogeochemical trend in the region.
  • The water quality in the region is characterized by high concentrations of cations (Ca2+, Mg2+, Na2+, and K+) and anions (SO42-, Cl-, NO3-, and F-).
  • The Entropy (E)-based classification model proposed by the study achieved an accuracy of 85.46% in predicting water quality.
  • The Technique of Order of Preference by Similarity to Ideal Solution (TOPSIS) approach was used to identify the best site for water resource management, with sites N-(1), (2), (7), and (3) being identified as polluted.
  • The study highlights the importance of machine learning approaches in improving water quality forecast accuracy, with the support vector machine (SVM) model achieving impressive accuracy.
  • The research recommends the use of entropy, TOPSIS, SMOTE, and SVM modeling methodologies in drinking forecasting with surface water appropriateness.
  • The outcomes of this study will be valuable for upcoming researchers and professionals in decision-making and management.

Statistics:

  • 42.85% of the water samples fell under the good category, while 57.14% belonged to the poor category of water quality.
  • The accuracy assessment using GIS was 85.46%.
  • The TopSIS approach identified sites N-(1), (2), (7), and (3) as the polluted ones.
  • The SMOTE-SVM model achieved an accuracy of 85.46% in predicting water quality.
  • The support vector machine (SVM) model achieved impressive accuracy, with values ranging from RMSE = 0.08-2.75, R-2 = 0.91-1.0.

Sources:

  • Water Quality Assessment and Geospatial Techniques for the Delineation of Surface Water Potential Zones: a Data-driven Approach Using Machine Learning Models. Desalination and Water Treatment, 2025; 324.
  • NewsRx. New Climate Change Findings from Department of Civil Engineering Described (Water Quality Assessment and Geospatial Techniques for the Delineation of Surface Water Potential Zones: a Data-driven Approach Using Machine Learning Models). Global Warming Focus. October 20, 2025; p 415.