1National Center for Atmospheric Research, Boulder, CO, USA
2Dept. of Chemistry and Biochemistry, and CIRES, Univ. of Colorado, Boulder, CO, USA
3Aerodyne Research, Billerica, MA, USA
4Dept. of Atmospheric and Oceanic Science, Univ. of Colorado, Boulder, CO, USA
5Brookhaven National Laboratory, Upton, New York, USA
6Pacific Northwest National Laboratory, Richland, WA, USA
*now at: Laboratory of Atmospheric Chemistry, Paul Scherrer Institut, Villigen, Switzerland
Abstract. It has been established that observed local and regional levels of secondary organic aerosols (SOA) in polluted areas cannot be explained by the oxidation and partitioning of anthropogenic and biogenic VOC precursors, at least using current mechanisms and parameterizations. In this study, the 3-D regional air quality model CHIMERE is applied to quantify the contribution to SOA formation of recently identified semi-volatile and intermediate volatility organic vapors (S/IVOC) in and around Mexico City for the MILAGRO field experiment during March 2006. The model has been updated to include explicitly the volatility distribution of primary organic aerosols (POA), their gas-particle partitioning and the gas-phase oxidation of the vapors. Two recently proposed parameterizations, those of Robinson et al. (2007) ("ROB") and Grieshop et al. (2009) ("GRI") are compared and evaluated against surface and aircraft measurements. The 3-D model results are assessed by comparing with the concentrations of OA components from Positive Matrix Factorization of Aerosol Mass Spectrometer (AMS) data, and for the first time also with oxygen-to-carbon ratios derived from high-resolution AMS measurements.
The results show a substantial enhancement in predicted SOA concentrations (3–6 times) with respect to the previously published base case without S/IVOCs (Hodzic et al., 2009), both within and downwind of the city leading to much reduced discrepancies with the total OA measurements. The predicted anthropogenic POA levels are found to agree within 20% with the observed HOA concentrations for both the ROB and GRI simulations, consistent with the interpretation of the emissions inventory by previous studies. The impact of biomass burning POA within the city is underestimated in comparison to the AMS BBOA, presumably due to insufficient nighttime smoldering emissions. Model improvements in OA predictions are associated with the better-captured SOA magnitude and diurnal variability. The predicted production from anthropogenic and biomass burning S/IVOC represents 40–60% of the total SOA at the surface during the day and is somewhat larger than that from aromatics, especially at the T1 site at the edge of the city. The SOA production from the continued multi-generation S/IVOC oxidation products continues actively downwind. Similar to aircraft observations, the predicted OA/ΔCO ratio for the ROB case increases from 20–30 μg sm−3 ppm−1 up to 60–70 μg sm−3 ppm−1 between a fresh and 1-day aged air mass, while the GRI case produces a 30–40% higher OA growth than observed. The predicted average O/C ratio of total OA for the ROB case is 0.16 at T0, substantially below observed value of 0.5. A much better agreement for O/C ratios and temporal variability (R2=0.63) is achieved with the updated GRI treatment. Both treatments show a deficiency in regard to POA evolution with a tendency to over-evaporate POA upon dilution of the urban plume suggesting that atmospheric HOA may be less volatile than assumed in these parameterizations. This study highlights the important potential role of S/IVOC chemistry in the SOA budget in this region, and highlights the need for improvements in current parameterizations. We note that our simulations did not include other proposed pathways of SOA formation such as formation from very volatile species like glyoxal, which can also contribute SOA mass and especially increase the O/C ratio.