Contribution of primary carbonaceous aerosol to cloud condensation nuclei: processes and uncertainties evaluated with a global aerosol microphysics model

This paper explores the impacts of primary carbonaceous aerosol on cloud condensation nuclei (CCN) concentrations in a global climate model with size-resolved aerosol microphysics. Organic matter (OM) and elemental carbon (EC) from two emissions inventories were incorporated into a preexisting model with sulfate and sea-salt aerosol. The addition of primary carbonaceous aerosol increased CCN(0.2%) concentrations by 65-90% in the globally averaged surface layer depending on the carbonaceous emissions inventory used. Sensitivity studies were performed to determine the relative importance of organic solubility/hygroscopicity in predicting CCN. In a sensitivity study where carbonaceous aerosol was assumed to be completely insoluble, concentrations of CCN(0.2%) still increased by 40-50% globally over the no carbonaceous simulation because primary carbonaceous emissions were able to become CCN via condensation of sulfuric acid. This shows that approximately half of the contribution of primary carbonaceous particles to CCN in our model comes from the addition of new particles (seeding effect) and half from the contribution of organic solute (solute effect). The solute effect tends to dominate more in areas where there is less inorganic aerosol than organic aerosol and the seeding effect tends to dominate in areas where there is more inorganic aerosol than organic aerosol. It was found that an accurate simulation of the number size distribution is necessary to predict the CCN concentration but assuming an average chemical composition will generally give a CCN concentration within a factor of 2. If a 'typical' size distribution is assumed for each species when calculating CCN, such as is done in bulk aerosol models, the mean error relative to a simulation with size resolved microphysics is on the order of 35%. Predicted values of carbonaceous aerosol mass and aerosol number were compared to observations and the model showed average errors of a factor of 3 for carbonaceous mass and a factor of 4 for total aerosol number; however, errors in the accumulation mode concentrations were found to be lower in comparisons with European and marine observations. The errors in CN and carbonaceous mass may be reduced by improving the emission size distributions of both primary sulfate and primary carbonaceous aerosol.