Water Pollution Monitoring and Decision Support System

Water is a source of life and is regarded as the most essential element of natural resources. Freshwater management is one of the most important tasks since health risks may arise from the consumption of water contaminated with infectious agents and toxic chemicals. Water quality is determined by assessing three classes of parameters which are physical, chemical, and biological. In this study, these parameters are taken into account to predict the water quality and spread of water pollution. Various sample datasets from Kaggle and Goa State Pollution Control Board (GSPCB) are used and analyzed to determine the parameters which can be considered to predict water pollution. The proposed model consists of a system that uses technologies such as Internet of Things, Machine learning, Embedded Systems. The data collected from these sensors is sent to a web-based micro-controller for further processing. The processed data is stored in the database through a cloud platform and various machine learning algorithms are applied to the stored data for the prediction of water pollution. On comparison, k-Nearest Neighbors (kNN) showed better results with the least error. Further on, 1-Dimensional Saint-Venant equations are used to predict the dispersion of pollutants from the polluting source.