Evaluation of recently-proposed secondary organic aerosol models for a case study in Mexico City
- ISSN: 1680-7316
Recent field studies have found large discrepancies in the measured\nvs. modeled SOA mass loadings in both urban and regional polluted\natmospheres. The reasons for these large differences are unclear.\nHere we revisit a case study of SOA formation in Mexico City described\nby Volkamer et al. (2006), during a photochemically active period\nwhen the impact of regional biomass burning is minor or negligible,\nand show that the observed increase in OA/Delta CO is consistent\nwith results from several groups during MILAGRO 2006. Then we use\nthe case study to evaluate three new SOA models: 1) the update of\naromatic SOA yields from recent chamber experiments (Ng et al., 2007);\n2) the formation of SOA from glyoxal (Volkamer et al., 2007a); and\n3) the formation of SOA from primary semivolatile and intermediate\nvolatility species (P-S/IVOC) (Robinson et al., 2007). We also evaluate\nthe effect of reduced partitioning of SOA into POA (Song et al.,\n2007). Traditional SOA precursors (mainly aromatics) by themselves\nstill fail to produce enough SOA to match the observations by a factor\nof similar to similar to 7. The new low-NOx aromatic pathways with\nvery high SOA yields make a very small contribution in this high-NOx\nurban environment as the RO2 center dot+NO reaction dominates the\nfate of the RO2 center dot radicals. Glyoxal contributes several\nmu g m(-3) to SOA formation, with similar timing as the measurements.\nP-S/IVOC are estimated from equilibrium with emitted POA, and introduce\na large amount of gas-phase oxidizable carbon that was not in models\nbefore. With the formulation in Robinson et al. (2007) these species\nhave a high SOA yield, and this mechanism can close the gap in SOA\nmass between measurements and models in our case study. However the\nvolatility of SOA produced in the model is too high and the O/C ratio\nis somewhat lower than observations. Glyoxal SOA helps to bring the\nO/C ratio of predicted and observed SOA into better agreement. The\nsensitivities of the model to some key uncertain parameters are evaluated.