AAAR 35th Annual Conference October 17 - October 21, 2016 Oregon Convention Center Portland, Oregon, USA
Abstract View
AuRo-SoFi: The Gold Standard for Organic Aerosol Source Apportionment Using ACSM Data
ANDRE PRÉVÔT, Francesco Canonaco, Yuliya Sosedova, Kaspar Rudolf Daellenbach, Imad El Haddad, Carlo Bozzetti, Athanasia Vlachou, Jay Slowik, Urs Baltensperger, Paul Scherrer Institute
Abstract Number: 318 Working Group: Source Apportionment
Abstract Long-term online monitoring with the aerosol chemical speciation monitor (ACSM, Aerodyne Research, Inc.) and subsequent source apportionment analysis reveals major organic aerosol sources. During the past decade the positive matrix factorization algorithm that creates linear combinations of static factor profiles (source profiles) and corresponding contributions (source contributions) has been extensively used for the source apportionment study of organic aerosol particles usually for short campaigns or seasonally divided. One way to overcome static factor profiles and assess in an improved way the uncertainties is the so-called Automated Rolling (AuRo) technique implemented in the Source Finder (SoFi) software developed at PSI. We have implemented the possibility to run thousands of constrained runs and assess the solution space a posteriori in a new way. Thereby we run constrained PMF analyses for windows of 1-2 months and move this window throughout the year leaving us with various solutions for the same day. Various criteria for the choice of “good solutions” are tested and will be presented. As mentioned, a big advantage of such exercises will be that the robustness of the chosen solutions or in other words the uncertainty of the solution can be assessed to some degree in such an approach. We will show results of some European datasets that were obtained within the European ACTRIS program, discussing the uncertainties of the source apportionment and the variability of the secondary organic aerosol factor spectra throughout the year.