Forecasting of Heavy Metal Contamination in Coastal Sea Surface Waters of the Karachi Harbour Area by Neural Network Approach
2019
Muhammad Ayaz and Nasir-Uddin Khan
The major and overriding factors affecting water quality and the aquatic ecosystems in the coastal areas are sewage and nutrient inputs from municipal and industrial wastewater, depletion of seaside contrivances, risks of public health as well as loss of biodiversity. The coastal area of the Karachi harbour is most heavily polluted due to these reasons. In this study, we proposed the artificial neural network (ANN) models to monitor and control the sea surface water quality of the Karachi coastal area along the harbour. Recently, various types of ANN have been successfully applied in hydrological fields. In this study, Nonlinear Auto Regressive eXogenous Neural Network (NARX-NN) shall be applied to predict the concentration of heavy metals in coastal sea surface water of the Karachi harbour area. This method provides significant insight into the comparative study of two different training functions of NARX-NN, namely, Levenberg-Marquardt (LM) and Scale Conjugate Gradient (SCG). The physical parameters like sea surface temperature (SST), salinity, tides and pH are taken as an input and the chemical parameters chromium, copper, lead, nickel and zinc are taken as output individually for all six locations. The performance of the model was evaluated by statistical criteria that include a correlation coefficient (r) and mean square error (MSE). The prediction results indicated that the LM training function is superior to SCG training function. Hope this study is helpful for local authorities and policy makers to develop a new infrastructure and install a water treatment plant to reduce the water pollution of the harbour area.
Afficher plus [+] Moins [-]Mots clés AGROVOC
Informations bibliographiques
Cette notice bibliographique a été fournie par Technoscience Publications
Découvrez la collection de ce fournisseur de données dans AGRIS
