Verifikace a zpřesňování výsledků rozptylových modelů pomocí prostorových analýz

Abstract

This diploma thesis deals with the analysis of data on sources and average annual concentrations of PM10 measured in 2015. The data set that was the input to the modeling was created within the Air Tritia project. Air quality was monitored in three countries in this study, the Czech Republic, Slovakia and Poland, together with the creation of a spatial information database. The factors used into the model were assumed to have a possible effect on the average annual concentrations of PM10. One of the methods used for creating models of distribution of pollutant concentrations was the Land use regression method, which uses linear regression for statistical analysis of data and model development. An alternative to this process was nonlinear regression using neural networks. The aim of this work was to compare the results of these methods with the results of mathematical modeling SYMOS '97 and to try to specify the results of poluttion dispersion models.