Representing Aging of Biomass Burning Aerosol in the Unified Model

Biomass burning is a major source of organic aerosols (OA) and black carbon (BC) in the atmosphere which impact climate through direct, semidirect, and indirect radiative effects. A recent analysis of ORACLES data over the Southeast Atlantic ocean observed decreased in OA:BC ratios over time from biomass burning aerosols, a change the authors attributed to photochemical OA loss processes. The OA:BC ratio influences the single scattering albedo of the aerosol and therefore whether it warms or cools the atmosphere. However most climate models do not reproduce this decrease in OA:BC ratio with chemical age well. Far from biomass burning emissions sources (> 10km), the change in OA:BC ratio is likely due to chemical or photolytic reactions while in the near field, we hypothesize effects of OA volatility are important and can reduce the rate of gas molecule evaporation from aerosols.

Model treatment of OA varies due to complex physical and chemical processes involved in OA partitioning and estimates of radiative forcing from OA range substantially partly for this reason. One common approach assumes organic species instantaneously equilibrate between their gas and aerosol phases which allows effective treatment of OA volatility, but this mass-based approach typically results in enhanced OA growth on large particles. Another kinetically-based method assumes OA condenses on any available surface which effectively distributes OA by particle size, but this method often assumes OA is completely nonvolatile once condensed.

To better represent OA:BC ratios, we simulate particle-phase OA loss in the UK Met Office Unified Model (UM), which is thought to occur through evaporation and heterogeneous oxidation as observed in the ORACLES campaign. We also present a simple semivolatile OA treatment within the UM that limits dependance on under-constrained parameters, using a reduced version of the popular volatility basis set approach. The rate of OA mass transfer is defined for each lognormal aerosol mode in the UM with a kinetic approach to better mimic observed OA size distributions. We first show the capabilities of this scheme using a box model and then compare results in the climate model to ORACLES flight campaign data over the Southeast Atlantic ocean. Though simpler than some approaches, this method introduces treatment of OA volatility to the UM while limiting the number of new parameters that are not well-constrained to prevent overtuning. We investigate how OA loss via aging and volatility affect the biomass burning plumes at different distances from the emissions sources.