Estimation of Aromatic Secondary Organic Aerosol Using a Molecular Tracer-A Chemical Transport Model Assessment

A modified community multiscale air quality model, which can simulate the regional distributions of 2,3-dihydroxy-4-oxopentanoic acid (DHOPA), a marker species for monoaromatic secondary organic aerosol (SOA), was applied to assess the applicability of using the DHOPA to aromatic SOA mass ratio (f_SOA) from smog chamber experiments to estimate aromatic SOA during a three-week wintertime air quality campaign in urban Shanghai. The modeled daily DHOPA concentrations based on the chamber-derived mass yields agree well with the organic marker field measurements (R = 0.79; MFB = 0.152; and MFE = 0.440). Two-thirds of the DHOPA are from the oxidation of ARO1 (lumped less-reactive aromatic species; mostly toluene), with the rest from ARO2 (lumped more-reactive aromatic species; mostly xylenes). Modeled DHOPA is mainly in the particle phase under ambient organic aerosol (OA) loading but could exhibit significant gas-particle partitioning when a higher estimation of the DHOPA vapor pressure is used. The modeled f_SOA shows a strong dependence on the OA loading when only semivolatile aromatic SOA components are included in the f_SOA calculations. However, this OA dependence becomes weaker when non-volatile oligomers and dicarbonyl SOA products are considered. A constant f_SOA value of ∼0.002 is determined when all aromatic SOA components are included, which is a factor of 2 smaller than the commonly applied chamber-based f_SOA value of 0.004 for toluene. This model-derived f_SOA value does not show much spatial variation and is not sensitive to alternative estimates of DHOPA vapor pressures and SOA yields, and thus provides an appropriate scaling factor to assess aromatic SOA from DHOPA measurements. This result helps refine the quantification of SOA attributable to monoaromatic hydrocarbons in urban environments and thereby facilitates the evaluation of control measures targeting these specific precursors.