Improving Secondary Organic Aerosol Simulations in the Chemical Transport Model Using Multigenerational Semi-volatile Organic Carbon Reactions

Jaemeen Baek (1)(2), Yongtao Hu (2), M. Talat Odman (2), Armistead G. Russell (2)

(1) University of Iowa, Iowa City (2) Georgia Institute of Technology, Atlanta

     Abstract Number: 740
     Last modified: May 14, 2010

Abstract
Chemical transport models historically have simulated organic aerosol concentrations that are lower than observed, likely due to an underestimate in the formation of secondary organic aerosol. Here, an SOA module in CMAQ version 4.5 is updated using detailed monoterpene species emissions, additional SOA formation from isoprene, sesquiterpenes and following the multigenerational photochemical oxidation of semi-volatile organic carbon (SVOC). The updated CMAQ was applied to both a summer and winter episode (2001 July and 2002 January) over a domain covering much of North America to elucidate how the various changes impact SOA formation and to assess if the revised version better simulates observed organic levels. Further, the revised model is now operational in the high resolution air quality forecasting (Hi-Res) system in the Southeast since May 2009 .
SOA formation had both seasonal and regional responses to the different updates. Isoprene and sesquiterpene emissions are abundant in the southeastern U.S., resulting in some regional increases in SOA due to the more detailed representation of biogenic VOCs. Oxidation of SVOC produces multigenerational SVOCs that partition into particle phase and contribute more than half of SOA. Using a lower enthalpy of vaporization (40 kJ/mol vs. 156 kJ/mol) reduces SOA concentrations as the overall aerosol yield from the first generation of oxidation falls by 70% in July. Overall, the updated SOA module significantly improved CMAQ performance on a daily basis, mainly due to the increased SOA formation from aged aerosol. Adding multigenerational SVOCs and isoprene SOA improved diurnal variations of simulated organic carbon as well. Further, the revised CMAQ has shown improved PM2.5 simulations in the Hi-Res forecasting system. Before revision, the Hi-Res System typically underestimated PM2.5 levels, in summer.