We present an empirical model of interaction among different contaminating impurities in the atmosphere. This problem was solved based on the correlation analysis of long-term series of data on the concentration of aerosol, carbon monoxide, nitric oxide, nitrogen peroxide, and so on, including meteorological quantities. Three groups of impurities have been revealed depending on their interrelation character. It has been found that besides meteorological conditions the atmospheric turbulence plays also important role in the interaction of chemical impurities with the aerosol.