Air pollution forecasting for Tehran city using Vector Auto Regression

Recently, in many urban areas, air quality has decreased because of human activities such as biomass burning and development of industrialization. There are some harmful pollutants like CO, NO2, and SO3 in the atmosphere that can lead to diseases such as asthma and lung cancer. One of the cities that seriously faces the problem of air pollution is Tehran. So, air pollution prediction for this city is of great importance and may lead to provide proper actions and controlling strategies. In this paper, we are going to use Vector Auto Regression (VAR) model to forecast daily concentrations of air pollutions in Tehran city for one day ahead. Since there are some correlations between air pollutants, it seems to get better results in forecasting if we get use of these correlations. So our approach is to use a VAR model which considers the impact of variables on each other. For choosing the right variables in our model, we will use a causality test. Experimental results demonstrate the high efficiency of the proposed approach in forecasting the concentrations of air pollutions in Tehran city.