AAAR 32nd Annual Conference
September 30 - October 4, 2013
Oregon Convention Center
Portland, Oregon, USA
Abstract View
Characterization of Coarse Aerosol in St. Louis and Phoenix: Results from EPA’s 2010-2011 Pilot Study
STEVEN BROWN, Jay Turner, Hilary Minor, Joann Rice, Paul Roberts, James B. Flanagan, Jeffrey Nichol, Sonoma Technology, Inc.
Abstract Number: 488 Working Group: Urban Aerosols
Abstract The coarse PM (PMc) pilot speciation study included one year of 1-in-3 day sampling at sites in Phoenix (AZ) and East St. Louis (IL), from June 2010 through May 2011. Among the instruments at each site were two Thermo 2025D sequential dichotomous (dichot) samplers, one Thermo 2025 sequential PM10 Federal Reference Method (FRM) sampler, and one Thermo 2025 sequential PM2.5 FRM sampler. Samples were collected on Teflon filters with Nylon filter backups for analysis of elements by XRF and of ions by ion chromatography, respectively, or quartz/quartz (Q/Q) filter sandwiches for analysis of carbon via thermal-optical analysis. PMc was directly measured via the dichot, and by the difference between PM10 and PM2.5 FRM concentrations.
On average, PMc at both sites was predominantly composed of crustal oxides, with 15% of the mass attributed to organic carbon (OC) and less than 5% of the mass attributed to other species such as sulfate and nitrate. Average PMc mass was 9.4 and 18.8 µg/m3, compared to fine PM mass of 10.7 and 7.6 µg/m3, at St Louis and Phoenix, respectively. Sum of species, assuming an OM/OC ratio of 1.6 and using standard equations for calculating crustal mass, exceeded measured mass at Phoenix by 13% and was 1% higher at St Louis. This finding that the PMc reconstructed mass is biased high, especially in Phoenix, suggests systematic errors in the estimation methodology, such as improper multipliers for estimating crustal PMc from elemental concentrations or x-ray fluorescence (XRF) attenuation corrections for light elements (e.g., Al, Si, Ca) that are too large. Further analysis of a subset of filters by ICP-MS suggested that the XRF attenuation corrections for crustal elements are too large, and will likely need to be revised prior to full-scale deployment of a national, routine monitoring network.