Abstract

Investigation the ability of artificial intelligent based predicting tool inmodeling the effects of atmospheric parameters on air pollution

Author(s): A.HedayatiMoghaddam1, D.Jafari

Prediction of the amount of air pollution on the basis of atmospheric parameters is necessary. Complex relation between atmospheric parameters and the amount of air pollution has been evaluated. So, such a complex relation makes it difficult to simulate the amount of air pollution through mathematical models. Simulation of the amount of air pollution was carried out by Artificial Neural Networks (ANNs). The effects of atmospheric parameters (temperature, pressure, precipitation, and wind speed) as well as the day of week on the amount of air quality index (AQI) were simulated by ANN. Atmospheric parameters were generated using meteorological data collected during 90 days from 23th September to 21th December of 2014 in city Tehran. AQI was measured in terms of 4 types of air pollutants (carbon monoxide, sulfide dioxide, particulate pollutants, and nitrogen dioxide).