Real-Time Turbidity Monitoring Using Machine Learning and Environmental Parameter Integration for Scalable Water Quality Management

A real-time turbidity monitoring system was developed to address limitations in traditional methods by integrating advanced optical sensors with a machine learning framework. The system incorporates environmental parameters such as rainfall, flow velocity, and temperature to enhance predictive accuracy. Field evaluations across five diverse sites achieved high performance (R 2 >0.94) and low errors (MAE as low as 1.4 NTU), with real-time processing latency averaging 48 milliseconds. The results demonstrated the system's adaptability to site-specific conditions, effectively capturing turbidity variability driven by rainfall-induced sediment mobilization and flow velocity. These findings highlight the system's potential for deployment in regulatory compliance, flood risk management, and industrial monitoring applications. Future efforts will focus on expanding calibration datasets and addressing sensor drift to enhance scalability and robustness for long-term environmental monitoring.