The Development of Mathematical Model in the Applied Sciences with Reference to Environmental Quantification of Coastal Water

Abstract

It is unanimous fact that in Marine ecosystem, pollutants can influence organisms directly or by changing the physical and chemical attributes of seawater. When various pollutants such as radioactive substances, heavy metals, petroleum and chemical compounds enter the sea by waste discharging, dumping, river discharge and stream flow, atmospheric precipitations and port activities. The impact of port and non-port pollutants on marine organisms expressed in acute toxicity, deformation, muta-genesis, effect on growth, physiological behaviors and so on. All marine organisms can influence the physical and chemical characteristics of seawater, bio-transformation and degradation of pollutants, specifically, the microbiological degradation which play a key role in marine self-purification. The Karachi Coastal area along the harbor is highly polluted due to oil spills, organic waste from the Karachi fish harbor, industrial as well as municipal waste water via Lyari River and small tanneries and cargo handling at Karachi port. Recently, various types of artificial neural network (ANN) have been successfully applied in hydrological fields. In this study, we apply Non-linear AutoRegressive eXogenous Neural Network (NARX-NN) to predict the concentration of heavy metals in sea surface water of the Karachi coastal area along Karachi harbor. This method provides significant insight about the comparative study of two different training function of NARX-NN namely, Levenberg-Marquardt (LM) and Scale Conjugate Gradient (SCG). These models will be useful for policy maker and administrative bodies for the improvement of implementing policies to surmount future environmental problems of the Karachi coastal area along the harbor.