Impact of the aging process of black carbon aerosols on their spatial distribution, hygroscopicity, and radiative forcing in a global climate model

Abstract

Abstract. Black carbon (BC) absorbs shortwave radiation more strongly than any other type of aerosol, and an accurate simulation of the aging processes of BC-containing particle is required to properly predict aerosol radiative forcing (ARF) and climate change. However, BC aging processes have been simplified in general circulation models (GCMs) due to limited computational resources. In particular, differences in the representation of the mixing states of BC-containing particles between GCMs constitute one of main reasons for the uncertainty in ARF estimates. To understand an impact of the BC aging processes and the mixing state of BC on the spatial distribution of BC and ARF caused by BC (BC-ARF), we implemented three different methods of incorporating BC aging processes into a global aerosol transport model, SPRINTARS: (1) the "AGV" method, using variable conversion rates of BC aging based on a new type of parameterization depending on both BC amount and sulfuric acid; (2) the "AGF" method, using a constant conversion rate used worldwide in GCMs; and (3) the "ORIG" method, which is used in the original SPRINTARS. First, we found that these different methods produced different BC burden within 10% over industrial areas and 50% over remote oceans. Second, a ratio of water-insoluble BC to total BC (WIBC ratio) was very different among the three methods. Near the BC source region, for example, the WIBC ratios were estimated to be 80–90% (AGV and AGF) and 50–60% (ORIG). Third, although the BC aging process in GCMs had small impacts on the BC burden, they had a large impact on BC-ARF through a change in both the WIBC ratio and non-BC compounds coating on BC cores. As a result, possible differences in the treatment of the BC aging process between aerosol modeling studies can produce a difference of approximately 0.3 Wm −2 in the magnitude of BC-ARF, which is comparable to the uncertainty suggested by results from a global aerosol modeling intercomparison project, AeroCom. The surface aerosol forcing efficiencies normalized by aerosol optical thickness and by BC burden varied greatly with region in the AGV method, which allowed for the existence of internally mixed BC and sulfate, whereas these were not varied with region in the AGF method. These results suggest that the efficiencies of BC-ARF obtained by previous studies using the AGF method are significantly underestimated.