Modeling water quality parameter bod in johor river and muar river

Abstract

Rivers are one of the important sources of water. Rivers also support a diverse flora and fauna. As Malaysia is a fast becoming an industrial country, many of the rivers have become polluted. Water pollution is very harmful to human, animals and water life. Therefore, water quality models are very important to describe and to predict the observed effects of a change in the river system. Biochemical oxygen demand (BOD) concentration data of Sungai Johor and Sungai Muar in Johor were fitted to two models, namely Streeter-Phelps and nonlinear regression (NLR) models, by using weighted least squares method and Gauss-Newton method respectively. BOD is one of the indicators for river. The data used in the study is BOD concentration data of Sungai Johor and Sungai Muar in Johor state from year 1981 to year 1990 and the data were obtained from Drainage and Irrigation Department Malaysia. The performance of the results was tested by mean squared error (MSE). The NLR model gives the smaller MSE values for Sungai Johor and Sungai Muar with values of 0.662319 and 2.367337 respectively. Conclusively, the results showed that NLR model is a better model than Streeter-Phelps model in estimating the parameters of BOD concentration data of Sungai Johor and Sungai Muar in Johor.