A COMPREHENSIVE REVIEW ON MACHINE LEARNING ROLE IN GROUNDWATER QUALITY MONITORING

Abstract Groundwater contamination is a worldwide issue that significantly affects ecological services and human health. In the civil, industrial, and agricultural domains, groundwater is an essential resource. A third or so of the world's population gets their drinking water from groundwater. In arid and semi-arid locations with limited surface water and precipitation, groundwater is an especially valuable resource. One of the main factors influencing a country's ability to develop sustainably is ensuring a steady and renewable supply of groundwater for use. Planning, policy formation, and groundwater resource management all depend on the ability to predict groundwater pollution from various chemical components. However, urbanisation, industry, climate change, and agricultural activities pose a major danger to groundwater quality. Hazards to human health, ecological services, and sustainable socioeconomic development include hydrocarbons, hazardous metals, pesticides, trace organic contaminants, nanoparticles, micro plastics, and other newly discovered contaminants. The use of machine learning (ML) techniques to forecast groundwater quality (GWQ) has significantly increased during the last 20 years.This review is the most comprehensive contemporary review on the subject because it reviews all ensemble, semi-supervised, supervised machine learning and unsupervised machine learning models used to predict any parameter related to groundwater quality.