A22A-02:
Modeling Elemental Composition of Organic Aerosol: Exploiting Laboratory and Ambient Measurement and the Implications of the Gap Between Them
Tuesday, 16 December 2014: 10:35 AM
Qi Chen1, Colette L Heald1, Jose L Jimenez2, Manjula R Canagaratna3, Qi Zhang4, Ling-Yan He5, Xiao-Feng Huang5 and Pedro Campuzano Jost6, (1)Massachusetts Institute of Technology, Civil and Environmental Engineering, Cambridge, MA, United States, (2)University of Colorado at Boulder, Dept. of Chemistry and Biochemistry, Boulder, CO, United States, (3)Aerodyne Research Inc., Billerica, MA, United States, (4)University of California Davis, Davis, CA, United States, (5)Peking University Shenzhen Graduate School, Key Laboratory for Urban Habitat Environmental Science and Technology, Shenzhen, China, (6)University of Colorado Boulder, Boulder, CO, United States
Abstract:
Global chemical transport models have been unable to capture the magnitude and variability of the mass concentrations of organic aerosol (OA). Uncertainty remains in the simulations, including the identification of primary sources and secondary tracers, the understanding of the formation mechanisms, and the representation of the atmospheric evolution of OA. There have been limited ambient measurements available to test simulations that use elemental composition to constrain the sources and aging of OA. In this study, a large dataset including both surface, aircraft, and laboratory observations of the atomic oxygen-to-carbon (O:C) and hydrogen-to-carbon (H:C) ratios of OA is synthesized and corrected for the bias of general Aerosol Mass Spectrometer elemental analysis. Mean observed O:C and H:C ratios range from 0.3 to 0.9 and 1.3 to 1.9, respectively, for the ground sites. Aircraft measurements show more oxidized OA with a vertical-level mean O:C of 1.2 and H:C of 1.4. We developed a global model simulation for the elemental composition of OA based on laboratory measurements. The standard GEOS-Chem simulation underestimates the O:C ratios, with the largest model bias in remote regions. Model performance is greatly improved by the addition of a laboratory-based oxidative-aging scheme. The revised simulations are best able to capture the observed variability of O:C in remote regions when the heterogeneous aging of secondary organic aerosol is introduced. The simulations underestimate the H:C ratios due to the gap between ambient and laboratory data. This suggests that that we may be missing sources and pathways which increase H:C, or alternatively, that laboratory experiments do not adequately mimic the ambient environment, and thus that their application in models may not reproduce field observations.
