Improving the efficiency of water use in agriculture by modelling the classification of groundwater quality

Abstract

This study investigates the potential of machine learning for classifying groundwater quality in Telangana, India, to optimize water resource utilization in agriculture. The study aims to develop and evaluate a decision tree model capable of accurately predicting groundwater quality based on chemical composition data. The objective is to identify key factors influencing water quality and provide insights for improving water management practices and enhancing agricultural productivity. The study utilizes a dataset of groundwater quality parameters collected over three years (2018-2020) and employs a decision tree algorithm for model development. The results demonstrate the effectiveness of the model, achieving an accuracy of 95.7%. The analysis highlights the significance of sodium content, dissolved salts ratio, total dissolved solids, and total water hardness as key factors influencing groundwater quality. This research underscores the potential of machine learning for enhancing water resource management in agriculture and suggests further exploration of temporal dynamics, predictive modeling, and broader geographic application to further refine and extend the model’s impact.