Document Type: Original Paper
Department of Statistics, Faculty of Science, Tarbiat Modares University, Tehran, Islamic Republic of Iran
Spatial-temporal modeling of air pollutants, ground-level ozone concentrations in particular, has attracted recent attention because by using spatial-temporal modeling, can analyze, interpolate or predict ozone levels at any location. In this paper we consider daily averages of troposphere ozone over Tehran city. For eliminating the trend of data, a dynamic linear model is used, then some features of correlation structure of de-trended data, such as stationarity, symmetry and separability are considered. Next based on the obtained features, an appropriate model is proposed. This model can be used for future predictions of ozone in Tehran.