Time-Series PM2.5 Source Apportionment – Value of the Measurements

Josh Hemann (1), Ricardo Piedrahita (2), Steve Dutton (3), Sverre Vedal (4), Shelly Miller (2), Jana Milford (2), MIKE HANNIGAN (2)

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

     Abstract Number: 259
     Last modified: November 9, 2009

Abstract
The overall objective of the Denver Aerosol Sources and Health (DASH) study has been to identify emission sources and constituents of fine particulate matter (PM2.5) that cause adverse health effects. Sources of air pollution were apportioned using a combination of detailed chemical analyses on PM2.5 filters and factor analysis modeling including positive matrix factorization (PMF). This work will be used to determine whether short-term changes in the contributions of specific emission sources to ambient PM2.5 were associated with an array of adverse health effects, including: (1) increased daily mortality, (2) increased daily hospitalizations for cardio-respiratory illnesses, and (3) worsening of asthma control in a panel of moderately severe asthmatic children. The detailed chemical analysis of the filters included gravimetric analysis for total PM2.5 mass concentration, ion chromatography for sulfate, nitrate, ammonium, calcium, magnesium, and potassium concentrations, inductively coupled plasma-mass spectrometry for 45 water-soluble metal species concentrations, thermal optical transmission for elemental carbon, total organic carbon, and carbon fraction concentrations, and gas chromatography-mass spectrometry for 72 organic molecular marker concentrations. This effort resulted in a time-series of PM2.5 speciation concentrations consisting of 350+ days of water-soluble metals, 1000+ days of organic molecular markers, and 2100+ days for the bulk species (mass, ions and carbon).
In the past, researchers have used either metals, bulk species (including carbon fractions), or organic molecular markers as input to factor analysis models to determine source contributions to the ambient PM2.5 mass concentration (or some portion of the PM2.5 mass concentration). The DASH speciation time-series has allowed us to explore the utility, or value, of these suites of species in combination and in smaller subsets. An increase in data’s utility is seen here as an increase in the number of factors resolved, or a reduction in the uncertainty of resolved factors. We will present the results of this comparison with specific focus on the value that each suite of species added to the source apportionment and the limitations we ran into using various combinations of species. This will provide valuable guidance on study design for future PM2.5 source apportionment studies.
This is an abstract for a proposed presentation/poster and does not necessarily reflect the policies of the U.S. EPA.