Air Pollution Analysis for Kannur City Using Artificial Neural Network

The pattern of economic growth that we are adopting is increasingly associated with environmental pollution. It is clear that in the developing or developed countries increased mechanization, transportation, populations, etc. causes air pollution which is a major and effective environmental problem. Present study focuses on analysis of air pollution in Kannur city of Kerala. In this study, prediction models are developed using non-linear autoregressive network with exogenous inputs (NARX Network) for the air pollutants. Several meteorological data such as ambient air temperature, relative humidity, and wind speed were given as input parameters while concentration of nitrogen dioxide (NO2), sulphur dioxide (SO2), respirable suspended particulate matter (RSPM) and suspended particulate matter (SPM) were considered as the output variable in this study. Also the four past values of output were fed back to the input. The performance of the developed model was assessed through a measure of minimum Mean Square Error (MSE). Result obtained shows that pollutant concentration at all two locations were found to be within the limit of NAAQS, India. Developed ANN model was found to be better model for future air pollution prediction with minimum MSE and it gives higher R2.