Water Quality Status of the Narmada River Across Its Basin: Predicting Water Quality Using Artificial Neural Network

Rivers play a critical role in the functioning of any ecosystem. Narmada is a major perennial river of India, which drains through the Deccan plateau and is used for bathing, drinking, irrigation, and industrial purposes. Its utilization had a detrimental effect on its water quality and is a cause of concern. However, data constraints could make it difficult to determine the water quality and take appropriate water quality management measures. This study is performed to evaluate the Artificial Neural Network (ANN) model for predicting Narmada river water quality. The study involved secondary data collection from different river water quality monitoring stations across the basin followed with the data analysis of data. The study used four different types of parameters namely physical, chemical, biological and toxic compounds. The study performed three stages of analysis. In the first stage, the analysis was performed to determine water quality parameters that are not meeting the recommended water quality across different water quality stations. The second stage of the study involved cluster analysis of the parameters to identify the correlated parameters. The third stage of the study involved the use of ANN to predict the water quality status based on other data. The study can predict the Chemical Oxygen Demand (COD) category i.e., high COD level and low COD level with 85.71% accuracy. The study showcases the ability of the ANN to perform the prediction modeling for Narmada River.