Water Quality Monitoring and Assessment for Efficient Water Resource Management through Internet of Things and Machine Learning Approaches for Agricultural Irrigation

Water quality monitoring and assessment play crucial roles in efficient water resource management, particularly in the context of agricultural rrigation. Leveraging Internet of Things (IoT) devices equipped with various sensors simplifies this process. In this study, we propose a comprehensive framework integrating IoT technology and Machine Learning (ML) techniques for water quality monitoring and assessment in agri- cultural settings. Our framework consists of four main modules: sensing, coordination, data processing, and decision-making. To gather essential water quality data, we deploy an array of sensors along the Rohri Canal and Gajrawah Canal in Nawabshah City, measuring parameters such as temperature, pH, turbidity, and Total Dissolved Solids (TDS). We then utilize ML algorithms to assess the Water Quality Index (WQI) and Water Quality Class (WQC). Preprocessing steps including data cleansing, Z-score normalization, correlation analysis, and data segmentation are implemented within the ML-enhanced framework. Regression models are employed for WQI prediction, while classification models are used for WQC prediction. The accuracy and efficacy of these models are evaluated using various metrics such as boxplots, violin plots, con- fusion matrices, and precision-recall metrics. Our findings indicate that the water quality in the Rohri Canal is generally superior to that in the Gajrawah Canal, which exhibits higher pollution levels. However, both canals remain suitable for agricultural irrigation, farming, and fishing.