Možnost použití biomarkerů k identifikaci zdrojů znečištění

Abstract

The presented dissertation thesis deals with the possibilities of the identification and quantification of the energy pollution sources using the proposed classification tool allowing the reduction of the total amount of the organic compounds identified in the samples and extraction only of the most important markers of the selected energy pollution sources in the sites of interest. Based on the selected markers, the influence of the fuel combustion on the air pollution is identified. The theoretical part is focused on the characterization of specific organic compounds – markers, which are characteristic for the particular energy pollution sources, namely the type of the fuel combusted. Specifically, the markers for biomass, coal and plastics burning and markers released from diesel and petrol engine vehicles. The most commonly used models for source apportionment of energy air pollution sources are also described, mentioning their particular advantages and disadvantages. The individual steps of the proposed model are described in the Methods section of the thesis including the justification of its mathematical soundness. The proposed model was applied in two measurement campaigns performed in a small-scale area (Napajedla) and large-scale area (Moravian-Silesian region). During the Napajedla measurement campaign, the possibility of focusing only on the prominent markers in the source apportionment studies was tested which would, in future, lead to facilitation of the time-consuming analytical techniques. The method was found to be reliable for the small-scale area. Moreover, using only the selected markers in the multivariate methods allowed the distinction of the pollution sources between the individual sites even more accurately than using all the determined compounds. During the Moravian-Silesian region campaign, identification of the most important energy pollution sources during the winter period of worsened air conditions was the main goal while the study was focused on the smog situation. High diversity of the identified organic markers in the PM10 samples between the individual sampling sites was to be addressed in this campaign, hence, the data had to be unified to extract as complete dataset as possible – with losing a minimal information. The results shown that each sampling site is specific by different prevalent source of pollution, some were, however, similar. In the cluster partitioning, the factor of site was found to be the most important during the smog season, whereas during the non-smog season, the sampling day was the most important factor. The most specific sites during both seasons were sites Poruba and Radvanice indicating that these sites are influenced by stable local sources and are not significantly influenced by e.g. change in wind direction.