AAAR 37th Annual Conference October 14 - October 18, 2019 Oregon Convention Center Portland, Oregon, USA
Abstract View
Using Highly Time-resolved Data to Improve the Lake Michigan Ozone Study: Particle Size Distributions and VOCs at a Coastal Site
MEGAN CHRISTIANSEN, Austin Doak, Dagen Hughes, Charles Stanier, Elizabeth Stone, Dylan Millet, Hariprasad Alwe, University of Iowa
Abstract Number: 782 Working Group: Source Apportionment
Abstract The collaborative field campaign, Lake Michigan Ozone Study 2017, occurred between May 22-June 22, 2017. The campaign combined ground site, mobile, aircraft, and ship measurements to better understand the ozone episodes around Lake Michigan. Particle size distributions (PSD) were measured at the main ground site (Zion, IL) with a Scanning Mobility Particle Sizer (SMPS) and Aerodynamic Particle Sizer (APS) and volatile organic compounds (VOCs) with a proton-transfer-reaction mass spectrometer (PTR-MS). The high time resolution of these data allows for the source receptor modeling to resolve local sources that periodically impact the site.
The Zion site is influenced by several local sources of air pollution within 1 km: traffic and cooking in the state park, commuter rail, and an arterial roadway. Transport to the site is heavily influenced by a lake breeze flow from Lake Michigan during early summer, especially on days with ozone above 70 ppb. PM2.5 mass reconstructed from the PSD compared favorably with collocated gravimetric filters (5.06 vs. 6.81 µgm-3). The average number concentration from 3 nm – 10 µm was 8258 cm-3 and from 3 – 10 nm was 1108 cm-3. Particle number and mass were positively correlated with ozone and its precursors. Preliminary Positive Matrix Factorization (PMF) identifies a nighttime aromatic hydrocarbon factor to the west, an acetonitrile rich factor from the southwest, and several other factors. Preliminary PMF of PSD show nuclei mode in the lake breeze and several other directionally specific ultrafine factors.
Diel patterns and wind direction dependence will be used to locate nearby sources, while potential source contribution functions (PSCF) and Concentration Weighted Trajectory (CWT) will be used to locate sources within the University of Iowa 4x4 km LMOS 2017 WRF-Chem modeling grid. PSCF uses hourly back trajectories processed in HYSPLIT to locate probable source grid cells.