Spatial and Temporal Variation of Speciated PM2.5 in Denver, CO

Teresa Coons (1), STEVEN DUTTON (2), Jennifer Peel (3), Sverre Vedal (4), Mike Hannigan (1)

(1) University of Colorado, Boulder (2) EPA, NCEA, RTP, (3) Colorado State University, (4) University of Washington, Seattle

     Abstract Number: 257
     Last modified: November 9, 2009

Abstract
As part of the Denver Aerosol Sources and Health (DASH) study, a six-year time-series of daily fine particulate matter was collected at one centrally-located site in Denver, Colorado. This site was located on the rooftop of an elementary school that was not near any major localized PM2.5 emission sources. The collected samples were chemically speciated for ions, elemental carbon, total organic carbon, water soluble metal species, and organic molecular markers. The resulting time-series of chemical species has been used in a source apportionment effort. Both the chemical species and source contribution time-series are being compared to daily health data collected from nearby hospitals including mortality and hospitalization rates in an effort to identify associations between sources (or chemical components) and health effects. As with any epidemiology study, there was a concern about the representativeness of one site to characterize the exposure in the entire urban area.
In an effort to examine the representativeness of the one DASH site, three additional sites were set up to the north, west and south of the existing site. The site to the west was similar to the primary site in that it was located on another elementary school roof in a neighborhood far from significant localized PM2.5 emission sources. The site to the south was co-located with the Denver NCore site in an industrial area adjacent to the major north-south interstate highway in Denver. The site to the north was co-located with the Denver regulatory chemical speciation monitoring site near another major interstate highway.
For one year, samples were collected every sixth day in the same manner as the primary site. These spatial samples were chemically speciated using the same methods as the daily DASH study. The PM2.5 collected on Teflon filters was analyzed for total PM2.5 mass concentration via a gravimetric technique and ionic species (sulfate, nitrate, ammonium, calcium, magnesium, and potassium) concentrations via ion chromatography. Quartz fiber filters were analyzed for elemental and organic carbon via thermal optical transmission and organic molecular markers via gas chromatography/mass spectrometry. This speciation dataset allowed us to explore the spatial variation of the composition of PM2.5 and investigate any seasonal dependence in the spatial variation. For example, nitrate and sulfate concentrations around Denver were very consistent at all locations with pair-wise correlation coefficients ranging from 0.88 to 0.99 and slopes ranging from 0.95 to 1.15 suggesting long-range or regional pollution sources for these ionic species. In addition to correlation statistics on annual species concentration, we will present broad spatial variation analyses using the coefficients of divergence and of variation. Similar analyses of seasonal subsets of the speciation dataset will be presented. Finally, we will discuss implications of these results to time-series health studies.
This is an abstract for a proposed presentation/poster and does not necessarily reflect the policies of the U.S. EPA.