Source Identification and Characterization
Aerosol emissions and their interactions with airborne pollutants directly influence air quality and have climatic impacts. Source identification and characterization aims to understand the properties of aerosols at a location of interest, known as the receptor site, in terms of the contributions and characteristics of emissions from different sources. In this area of research, statistical models are used to estimate the relative contributions of emissions from sources based on measurements made at receptor sites. In addition, measurements are being used to characterize aerosols released from known sources; vehicles are of particular interest in this regard. Air quality management is dependent on the identification of aerosol emission sources, and the ability to quantitatively relate emissions to the characteristics of the aerosols at a receptor site.
Data Analysis and Source Apportionment of PM2.5 in Golden, British Columbia using Positive Matrix Factorization (PM) Contract report, C-H. Jeong, G.J. Evans, CR-WB-2007-02Show Abstract >>
This report outlines the methodology and findings from a three-month research contract to interpret particulate matter speciation data from Golden, British Columbia. The focus of this project was the application of receptor modeling techniques to identify the origins of the particulate matter (PM). Positive Matrix Factorization (PMF), a useful factor analysis method, was applied to 24-hour PM speciation data collected between November 2004 and August 2006 to identify possible sources of PM and determine the contribution of each identified source to ambient PM concentrations in Golden. PMF apportioned the PM2.5 mass into seven factors identified as road salt, secondary sulphate, wood burning, wood processing, crustal material, traffic, and winter heating. The most important sources affecting ambient air quality in Golden were wood burning and winter heating factors. In order to obtain the quantitative contributions of the sources, their relative contributions were normalized through multiple linear regression against the aerosol mass concentration. In addition, seasonal trends of the PMF-resolved sources were characterized. The identity of the sources was further elucidated by exploring their relationship with other measured parameters; a correlation study was conducted to evaluate relationships between the sources, volatile organic compounds, gaseous pollutants, and meteorological variables.
Advanced Factor Analysis of Chemically Speciated Aerosols using Non-Refractory Mass Spectra and Trace Refractory Elemental Compositions, M. McGuire1, G.J. Evans1, J. Brook2, C. Jeong, G. Lu2, J. Ondov3, Conference Presentation, CP-WB-2007-01
1Southern Ontario Centre for Atmospheric Aerosol Research, Department of Chemical Engineering and Applied Chemistry, University of Toronto, 2Meteorological Service of Canada, Environment Canada, 3University of MarylandShow Abstract >>
Factor analysis using Positive Matrix Factorization (PMF 2) was performed on the mass spectra of non-refractory aerosol components measured by an Aerosol Mass Spectrometer, as well as refractory elemental components acquired using a Semi-Continous Elements in Aerosol System (SEAS). A comprehensive data set was captured by Environment Canada’s CRUISER mobile air pollution sampling laboratory which was deployed in MicMac Park in Windsor, Ontario for three weeks during January and February 2005. MicMac Park was chosen for its proximity to a variety of local sources, namely continuous trans-border traffic, and industrial complexes in both Windsor and Detroit. Regional sources also impacted Windsor given its proximity to Toronto as well as power generating stations in the Ohio valley. Variability in local and regional source concentrations at the receptor site was induced by shifting wind direction and speeds. CRUISER was equipped with an Aerosol Mass Spectrometer sampling every 15 minutes, an Aethelometer, a GRIMM dust monitor as well as a Condensation Particle Counter. Gas monitors measured NOx, SO2, O3 and CO concentrations. Trace refractory elemental contributions were collected by the SEAS on-site every half-hour and stored for off-site Inductively Coupled Plasma Mass Spectrometry. Factor analysis using PMF 2 was performed first on the mass spectra of the non-refractory aerosol components generated by the Aerosol Mass Spectrometer. Advanced factor analysis was then performed by incorporating the trace refractory elemental compositions collected using the SEAS along with the aerosol mass spectra to yield increased factor resolution. Resolved source profiles from both approaches were compared to reference spectra and their source contributions were compared with gas and particle concentrations, meteorological trends and air mass back-trajectories generated using the NOAA HYSPLIT model. The presentation will present the results of the source apportionment on the Windsor airshed and compare and contrast the two factor analysis approaches.
ATOFMS ART2a Analysis and IT Support for SOCAAR, E. McDermott, Summer Research Report, SRA-WB-2006-05Show Abstract >>
The TSI Aerosol Time-of-Flight Mass Spectrometer (ATOFMS) Series 3800 is a state-of-the-art instrument capable of characterizing individual aerosol particles. This study discusses the development of a method for analyzing the ATOFMS output, which produces 1 to 8 gigabytes of data per day, as rated by the manufacturer. Additionally, the software bundled with the ATOFMS relies on extensive operator interaction and a peak-finding algorithm that is computationally expensive, and therefore slow. These shortcomings necessitated a method in which data could be quickly and roughly divided into meaningful categories for further analysis, preferably with little user input. After considering DBSCAN and recognizing its scaling restrictions, it was decided that the ART2a algorithm would be used to build a tool to analyze this large data set. Scripts for ATOFMS data syncing, merging ATOFMS datasets, retrieving meteorological data, and server backup processes were also developed.
Investigation of the Secondary Organic Aerosol (SOA) Fraction, Summer 2005 British Columbia, Andrew J. Knox1, Greg J. Evans1, Jeffrey R. Brook1,2, Yayne-Abeba Aklilu2, Patrick Lee2, Gang Lu2, Conference Presentation, CP-WB-2006-03
1Department of Chemical Engineering and Applied Chemistry, University of Toronto, M5S3E5, Toronto, Canada, 2Air Quality Research Division, Environment Canada, M3H 5T4, Toronto, CanadaShow Abstract >>
Nearly all carbonaceous aerosols can be classified as either Black Carbon (BC) or organic Carbon (OC). BC is a primary aerosol generated by combustion only, whereas OC, the carbon fraction of Organic Matter (OM), can have both primary and secondary sources. The contribution of Secondary Organic Aerosol (SOA) to the OM portion of carbonaceous aerosols must be well understood. Such an understanding is vital for assessing the projected impact of emissions control legislation, given that reductions in emissions may not have an easily predicted effect on SOA levels. SOA can also have an impact on the optical properties of BC, altering the radiative forcing of BC containing aerosols (Schnaiter, et al., 2005). This impact of SOA would have substantial ramifications for the carbon crisis. This study was performed as part of an effort to quantify the SOA fraction in urban areas of British Columbia and to characterize its sources.
From May to July of 2005 Environment Canada’s mobile lab, CRUISER (Canadian Regional and Urban Investigation System for Environmental Research), collected data on BC, OM and other aerosols in Vancouver and Abbotsford, British Columbia. A model AE-20 dual wavelength aethalometer, a Particle Soot Absorption Photometer (PSAP), a
Multiangle Absorption Photometer (MAAP), and a Photoacoustic spectrometer (PA) monitored BC, while an Aerosol Mass Spectrometer (AMS) collected OM data. Other aerosol measurement instruments were in operation, as were a suite of meteorological data collection units. Sampling was also carried out on the grounds of British Columbia’s Pacific Environmental Science Centre (PESC).
Trends in BC and OM, air mass behaviour, and meteorological conditions were used, with BC as an indicator of combustion sources, to determine the contribution of secondary sources to ambient OM aerosol. Specific compounds often associated with SOA were noted. CRUISER’s mobility was exploited to collect data at various ranges from sources and to place the instrumentation at locations where the most telling data was expected to be available. Meteorological and location data were correlated with aerosol measurements.
Interpretation of Pollutant Data, S. Basma, Summer Research Report, SRA-WB-2005-01Show Abstract >>
Urban air pollution is a serious health concern in Canada. Adverse health effects related to exposure have motivated researchers to study air pollution in three major ways: tracking the origin of pollutants in the air, understanding the chemical and physical processes that influence their concentration, and developing techniques to eliminate them. Measurements of SO2, Ultra Fine Particles (UFP), and NO were taken in downtown Toronto during 2003. The first part of this report examines the influence of emissions from Lakeview Generating Station on ambient SO2 concentration. The highest average concentrations of SO2 were observed when the wind direction was between 220 and 260 degrees (south-west) suggesting Lakeview Generating Station is likely a contributor to SO2 events. The second part of the report studies the trends of Ultra Fine Particles (UFP) by examining daily, monthly, and seasonal profiles. UFP counts showed a much higher average concentration during weekdays than during weekends. Observations demonstrated an overall annual trend of lower levels of UFP in the summer and higher levels in the winter. Also, the correlations of UFP with other pollutants and physical atmospheric conditions were analyzed in order to identify the key characteristics of UFP. The diurnal profile of NO concentration showed a morning peak comparable to UFP profile, which is attributed to morning rush hour traffic. Regional air quality does not seem to be a key factor contributing to the UFP concentration. Further research is necessary to investigate the link between smog alert days and UFP concentration.
Application of Advanced Receptor Modelling to Highly Time Resolved Aerosol Data, K. C. Buset, M.A.Sc. Thesis, MASc-WB-2005-01Show Abstract >>
This study investigates the application of an advanced receptor model to highly time resolved aerosol concentration data. Aerosol mass concentration data, ambient gas measurements, and meteorological data were collected from August 20 to September 25, 2003 in Toronto, Ontario, Canada. A suite of instruments, including an aerosol mass spectrometer (AMS), was incorporated into the study which produced a diverse air quality data set. A Multilinear Engine receptor modelling script was developed to combine aerosol mass concentration data, meteorological data and time-of-day information so as to resolve sources contributing to Toronto’s aerosol pollution. The inclusion of particle size resolved organic mass concentrations into the receptor modelling allowed for improved interpretation of the modelling results. The combination of a highly time resolved and diverse air quality data set, and the application of advanced receptor modelling, allowed for a more complete description of contributing aerosol sources as compared to previous studies.