Full text
| |
| Author(s): |
Beatriz Sayuri Oyama
Total Authors: 1
|
| Document type: | Master's Dissertation |
| Press: | São Paulo. |
| Institution: | Universidade de São Paulo (USP). Instituto Astronômico e Geofísico (IAG/SBD) |
| Defense date: | 2010-05-03 |
| Examining board members: |
Maria de Fatima Andrade;
Alexandre Lima Correia;
Adalgiza Fornaro
|
| Advisor: | Maria de Fatima Andrade |
| Abstract | |
Several studies have been developed in order to describe the gaseous phase of atmospheric constituents in the Metropolitan Region of Sao Paulo (RMSP). However, the aerosol description remains simplified in chemical models, despite the knowledge acquired in its characterization and composition analyses. Facing these limitations, the objective of this work is to identify the main emission sources of fine particulate matter, specially the vehicular ones that present a lot of difficulties due to the fact that the characteristic trace elements are unknown for these sources. It was used in this work 201 samples collected in 24-hour period each at Dr. Arnaldo Avenue, a large and busy avenue in the city of São Paulo, from June 2007 to August 2008. The source identification was accomplished considering the samples composition and using receptor models: Factor Analysis (FA) and Positive Matrix Factorization (PMF) techniques. PMF was a new statistical tool in the study of particulates in the city of São Paulo. The number of sources identified by these two models was different. The FA technique identified 4 factors, (soil, fuel burning, and 2 factors combining in light and heavy-duty vehicles), whereas PMF identified 6, the same as FA (light and heavier vehicles differentiated) and biomass burning. There was concordance between the two techniques, considering that both found that vehicular emission is the major contribution for concentration. The comparison between the models indicated that PMF model present a better source classification, mainly for the vehicular identification. The PMF technique considers the error of each sample in the analysis, weighting the variables and imposing that all the factors must be positive. This mechanism provides a better characterization of sources linking the results with the physics of the process. (AU) | |