AAAR 37th Annual Conference October 14 - October 18, 2019 Oregon Convention Center Portland, Oregon, USA
Abstract View
Simulation on IVOC Emissions and SOA Formation in Los Angeles during CalNex Study Using Updated Mobile Source Emission Profiles and SOA Parameterization
Quanyang Lu, Benjamin Murphy, Peter Adams, Yunliang Zhao, Momei Qin, Havala Pye, Christos Efstathiou, Chris Allen, ALLEN ROBINSON, Carnegie Mellon University
Abstract Number: 686 Working Group: Urban Aerosols
Abstract In this talk, we describe simulations using Community Multiscale Air Quality model version 5.3 (CMAQ v5.3) to investigate the contribution of intermediate volatile organic compounds (IVOCs) to secondary organic aerosol formation (SOA) in Southern California during the CalNex study. We first derive a simplified parameterization for SOA formation from IVOCs that accounts for both differences in volatility and molecular structure. We also implement new mobile emission profiles that quantitatively include IVOCs based on direct measurements. In the Los Angeles region, gasoline sources emit four times more non-methane organic gases (NMOG) than diesels, but diesel emit roughly three times more IVOCs on an absolute basis. All mobile sources (including on- and off-road gasoline, aircraft and on- and off-road diesel) are predicted to contribute 2.4 μg/m3 of hydrocarbon IVOCs in Pasadena, CA, which corresponds to 38% of the concentrations measured during the CalNex campaign. Therefore, mobile sources emissions do not explain the total measured hydrocarbon IVOCs. To close the mass balance in Pasadena, IVOC would need to contribute 12% NMOG emissions from non-mobile sources. Mobile and non-mobile source IVOCs increase the model predicted SOA in Pasadena by a factor of 5.2. In a scenario with IVOC equal to 26.8% NMOG for non-mobile sources, modelled OA agrees well with the AMS-measured average OA diurnal pattern with noon peak SOA of 8.6 μg/m3. In areas downwind of L.A., on average, 70% of predicted OA mass is SOA formed from oxidation of IVOC emissions. Evaluating the model predicted OA using 24-hr average OC concentrations from the Chemical Speciation Network (CSN, assuming OA-to-OC ratio = 1.8) shows that, with added IVOC emissions, the overall Fractional Bias (FB) is reduced from -0.59 to -0.10 and the Fractional Error (FE) is reduced from 0.67 to 0.42 for 7 sites in California.