Analysis of organic pollutant and coliform in Thamiraparani River with artificial neural networks

Abstract

Thamiraparani River, flowing continuously for 128km, passes through many villages, towns and Tirunelveli Corporation. It is a perennial river and monsoon based catchment, where drainage water and night soil pollutants are mixed and the pollutants from Textile and Paper industries are drained to this region. These are the main sources of pollutants of ThamiraparaniRiver. Since long, Thamiraparani Riveris a main source of water supply to many towns which include Tirunelveli Corporation. Rapid urbanization and industrialization are the significant issues in environmental degradation especially in the quality of both surface and ground water quality. The fundamental water quality factors are, natural contamination indicated by Biological Oxygen Demand (BOD) and pathogens confirmed by coliform. The idea of the study is to predict the influences of BOD and Fecal coliform in Total coliform by using Neural Network. Water quality dataof Thamiraparani River for 12 years were collected from Central Pollution Control Board (CPCB), for six different locations such as Papanasam, Cheranmahadevi, Tirunelveli, Murappanadu, Ambasamuthram and Arumuganeri. In those samples, BOD and Fecal coliform biological parameters are used as independent variables and Total coliform is used as dependent variable. The multilayer perceptron neural network is designed to predict the Total coliform with BOD and fecal coliform as input variables. Study revealed that the association between the independent and dependent variables are good and R2value of the model was 0.846. The investigation recommends that fecal coliform and Total coliform versus BOD can be applied as contamination indicator for further research of water quality,Since BOD levels mostly mirror the level of natural contamination related tofecal sources in this waterway. The proposed ANN model uses BOD and fecal coliform to predict the future Total coliform values, which assists in easy predictability of water contaminants.