EUVIMEDEUVIMED entdecken
Crossref

Prediction of WQI Using an Artificial Neural Network Model for Drinking Water Management in the Therlam Region, Vizianagaram District, India

Rupakumari Gedala, Srinivasa Rao. G.V.R.

Environment and Ecology · 2026-06 · 44 (2) · 385-400

DOI: 10.60151/envec/ehdf3490 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Environment and Ecology, 2026-06. Verfasst von Rupakumari Gedala und 1 weiteren Autor:innen.

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Groundwater quality monitoring in hard-rock aquifer systems is essential for sustainable resource management in rural regions dependent on groundwater for drinking purposes. The present study evaluates groundwater quality using the weighted arithmetic Water Quality Index (WQI) method and develops an Artificial Neural Network (ANN)-based framework for rapid WQI estimation in the Therlam region of Vizianagaram, Andhra Pradesh. A total of 828 groundwater samples were collected from 23 monitoring stations during November 2018–October 2021 and analyzed for fifteen physicochemical parameters following American Public Health Association (2017) standard procedures. WQI classification indicated that 30.43% of samples belong to the good to excellent category, 47.83% to poor, 17.39% to very poor, and 4.35% to unsuitable category, indicating moderate hydrochemical variability within the study area. To estimate WQI efficiently from routinely monitored hydrochemical parameters, a feed-forward ANN model with a 15–5–1 architecture trained using the Levenberg–Marquardt algorithm was developed. Approximately 70%, 15%, and 15% of the dataset were used for training, validation, and testing, respectively. Multicollinearity among input variables was evaluated using Variance Inflation Factor (VIF) analysis prior to model implementation. Model performance was further compared with Multiple Linear Regression (MLR) and Random Forest (RF) approaches. The ANN model achieved satisfactory predictive performance (R² = 0.9658; RMSE = 0.4964, MAE = 0.3624), outperforming the baseline models. The developed ANN framework is intended as a rapid computational tool for WQI estimation rather than an independent forecasting model. The study highlights the applicability of machine-learning-assisted groundwater quality assessment in hard-rock aquifer regions while emphasizing the need for inclusion of hydro-meteorological and anthropogenic variables in future predictive studies. However, the developed ANN framework represents a site-specific surrogate estimation model and should not be interpreted as an independent groundwater-quality forecasting system.

Zitieren

APA

Gedala, R., & G.V.R., S. R. (2026). Prediction of WQI Using an Artificial Neural Network Model for Drinking Water Management in the Therlam Region, Vizianagaram District, India. Environment and Ecology, 44(2), 385-400. https://doi.org/10.60151/envec/ehdf3490

Vancouver

Gedala R, G.V.R. SR. Prediction of WQI Using an Artificial Neural Network Model for Drinking Water Management in the Therlam Region, Vizianagaram District, India. Environment and Ecology. 2026;44(2):385-400. doi:10.60151/envec/ehdf3490

BibTeX
@article{Gedala2026,
  author = {Gedala, Rupakumari and G.V.R., Srinivasa Rao.},
  title = {Prediction of WQI Using an Artificial Neural Network Model for Drinking Water Management in the Therlam Region, Vizianagaram District, India},
  journal = {Environment and Ecology},
  year = {2026},
  volume = {44},
  number = {2},
  pages = {385--400},
  doi = {10.60151/envec/ehdf3490}
}
Kostenlos starten →