Developing a novel tool for assessing the groundwater incorporating water quality index and machine learning approach

Abstract

Groundwater plays a pivotal role as a global source of drinking water. To meet sustainable development goals, it is crucial to consistently monitor and manage groundwater quality. Despite its significance, there are currently no specific tools available for assessing trace/heavy metal contamination in groundwater. Addressing this gap, our research introduces an innovative approach: the Groundwater Quality Index (GWQI) model, developed and tested in the Savar sub-district of Bangladesh. The GWQI model integrates ten water quality indicators, including six heavy metals, collected from 38 sampling sites in the study area. To enhance the precision of water quality assessment, the study employed six established machine learning (ML) techniques, evaluating the model's performance based on factors such as uncertainty, sensitivity, and reliability. A major advancement of this study is the incorporation of heavy metals into the framework of the water quality index model. To the best of the authors knowledge, this research marks the first initiative to develop a GWQI model encompassing heavy/trace elements. Findings from the GWQI assessment revealed that groundwater quality in the study area ranged from 'good' to 'fair,' indicating that most water quality indicators met the standard limits set by the Bangladesh government and the World Health Organization. In predicting GWQI scores, artificial neural networks (ANN) outperformed the other ML models. Performance metrics, including root mean square error (RMSE), mean square error (MSE), mean absolute error (MAE) for training (RMSE = 0.361; MSE = 0.131; MAE = 0.262), testing (RMSE = 0.001; MSE = 0.00; MAE = 0.001), and prediction evaluation statistics (PBIAS = 0.000), demonstrated the superior effectiveness of ANN. Moreover, the GWQI exhibited high sensitivity (R2 = 1.0) and low uncertainty (less than 2%) in rating water quality. These results affirm the reliability of developed novel model for groundwater quality monitoring and management, especially regarding heavy metals.