Prediction of Water Quality Index Using Artificial Neural Network Modeling in the Jami Region, Vizianagaram District, Andhra Pradesh, India

Groundwater is the primary source of drinking water in rural regions of Andhra Pradesh, India, and its quality is increasingly influenced by both natural processes and anthropogenic activities. This study evaluates and predicts groundwater quality in selected villages of the Jami region, Vizianagaram District, using the Water Quality Index (WQI) integrated with an Artificial Neural Network (ANN) model. Groundwater samples were collected from twenty-four locations over a three-year period (2018–2021) and analyzed for major physico-chemical parameters following APHA standard methods. WQI was calculated using the weighted arithmetic method, revealing that 62.5% of samples fall under good water quality, 29.2% under poor quality, and 8.3% are unsuitable for drinking. To overcome the limitations of conventional index-based assessment, a feed-forward back-propagation ANN model was developed using the Levenberg–Marquardt algorithm in MATLAB. The ANN-predicted WQI values showed excellent agreement with observed values (R² = 0.9990; RMSE = 0.7940; MAE = 0.5098). The findings demonstrate that ANN-based modeling is a robust and efficient tool for groundwater quality prediction and can support sustainable groundwater management in data-scarce regions.