The Use of Innovative Methodology for the Characterization of Benzene, Toluene, Ethylbenzene and Xylene Sources in the Belgrade Area

The growth of urban population, economic development, urbanization and transport have a strong impact on environmental pollution. The increase in air pollutant concentrations over the last few decades has been in focus of contemporary science and research mainly for its adverse effects on public health, environment and climate change. In this paper, we are using the innovative integrated methodology for spatio-temporal air pollution modeling, based on receptor-oriented air circulation modeling and artificial intelligence implemented through machine learning methods for detailed characterization of toxic, mutagenic and carcinogenic representatives of volatile organic species – benzene, toluene, ethylbenzene and xylene, in the Belgrade area. Also, the study evaluates the possibilities of spatio-temporal forecast based on the integrated methodology. The results suggest that temperature and wind speed represent the main parameters which govern the spatio-temporal distribution of benzene, while the impact of other factors shows significant variations depending on the characteristics of receptor location.