AAAR 34th Annual Conference
October 12 - October 16, 2015
Hyatt Regency
Minneapolis, Minnesota, USA
Abstract View
Toward the Minimal Representation of the Aerosol Mixing State
LAURA FIERCE, Nicole Riemer, Tami Bond, University of Illinois at Urbana-Champaign
Abstract Number: 566 Working Group: Aerosols, Clouds, and Climate
Abstract One challenge in the simulation of atmospheric aerosols is the representation of particle composition. Tracking particle-level composition is computationally expensive, so global models approximate the representation of the aerosol mixing state, that is, the manner in which chemical species are distributed across the particle population. In this study, we used a particle-resolved aerosol model to evaluate representations of the aerosol mixing state that are commonly applied in global models.
The evolution of the aerosol mixing state was simulated in a series of scenarios with PartMC-MOSIAC. Each simulation started with an external mixture of two particle types: particles containing black and organic carbon, representing fresh combustion emissions, and hygroscopic background aerosol comprised of ammonium sulfate. As the simulations proceeded, we computed cloud condensation nuclei concentrations and aerosol absorption for the same particle population, but with different representations of the aerosol mixing state. The size distribution and bulk aerosol composition was the same for all treatments, and they differed only in the assumed distribution in chemical species between individual particles.
The full internal mixture approximation, the simplest representation of the mixing state, was suitable for modeling cloud condensation nuclei activity of aged aerosol. We show that the timescale for particles to become internally mixed with respect to their hygroscopic properties depends on the characteristics of the chemical environment, with timescales on the order of hours in urban areas. On the other hand, the full internal mixture was not suitable for modeling aerosol absorption, even for highly aged populations. We found absorption by aged aerosol populations was overestimated for both the full internal mixture and sectional treatments of aerosol composition. Error in modeled absorption was lowest if the composition of separate particle types was tracked, which is possible in modal models.