A sustainable approach to evaluating groundwater contamination in the Jharsuguda district, Odisha, India, using a multivariate statistical approach, modelling, and viable remedial measures

Abstract

This study was carried out by collecting groundwater samples from various important locations in the Jharsuguda district, Odisha, India. From physicochemical parameters, it was found that the groundwater in most of the areas is heavily contaminated. From the study of bacteriological analysis, it was observed that only five groundwater samples from the residential colony area are contaminated by fecal coliform. Using a scree plot, PCA extracted the first three components with 69.19 %, 18.39 %, and 7.99 % of variance and retained the maximum variation of the data set. Only eight variables were responsible for 99.9 % of the cumulative percentage of variance based on FA. From the comparative analysis, it was found that the performance of a deep neural network (DNN) is comparatively more consistent and clearer than an artificial neural network (ANN) model in predicting the targeted outputs. From the DNN model, the R² value is 0.91, which shows that there is a good variance in the observed data, but in the ANN model, the R² value is 0.79. The purpose of this study is to identify the major pollutants that cause groundwater contamination, investigate pollution trends, and provide credible forecasting tools to help prevent groundwater contamination and provide sustainable resource management. This work facilitates clear understanding regarding the level of contamination of groundwater and specific pollutants that are maximally responsible for the contamination. It also promotes public wellness and sustainable water management in the study area. To remove fecal coliform, chlorination is suggested to be the most viable process.