Elevated biomass burning aerosols interact with clouds thermodynamically by perturbing the lower atmosphere thermal structure as well as microphysically. These effects strongly depend on the vertical distribution and optical properties of aerosols in the long-distance transport (Feng et al., 2016). On the other hand, the evolution of biomass burning aerosols is also modulated by clouds. Global climate models have a poor representation of the vertical extent of the aerosol layers when transported far from the biomass burning sources in South Africa (Das et al. 2017). In this study, we examine vertical and seasonal variability of biomass burning aerosols simulated by the global climate model CAM5, compared with the ARM Mobile Facility 1 (AMF-1) observations from the DOE LASIC campaign (2016-2017) on Ascension Island (AI).

During the peak biomass burning season (June-August), smoke is present within and above the boundary layer, as expressed with large extinction values in the Micro-Pulse Lider (MPL) retrievals on AI. The MPL-retrieved monthly mean aerosol extinction in the boundary layer is up to 100~200 Mm^-1, compared to low values around 20 Mm^-1 in the clean marine environment. Surface aerosol optical measurements indicate that the increased aerosol extinction and absorption are dominated by the fine-mode aerosols coincident with the high black carbon concentrations (>1000 ng/m³), suggesting non-marine sources. This temporal variability is simulated consistently by CAM5, although the model underestimates the surface absorption by about 50%, possibly related to emission sources. Incorporation of enhanced absorption by BrC improves the model-simulated aerosol absorption and spectral dependence. In the free troposphere, MPL shows that the smoke layer is present between 1.5 to 3 km in July and extends higher to about 4 km in August. The retrieved extinctions from MPL are lower in the elevated smoke layer (below 50 Mm^-1) than in the boundary layer possibly due to lower relative humidity. CAM5 is able to capture the elevated smoke layer between 2-4 km with the maximum value of about 40 Mm^-1 around 3 km over this region, similar to the MPL data. In addition, the model predictions indicate biomass burning aerosols at 3 km are more absorbing than the boundary layer aerosols over AI, e.g., the monthly mean absorption coefficient for September is 9 Mm^-1 at 3 km and 3.5 Mm^-1 at surface in the model. These aerosol-induced changes in the lower-troposphere heating rate profile are examined with observed and modeled extinctions. We also compare the MPL extinction profiles with satellite CALIPSO and CATS retrievals along the plume transport tracks during LASIC. Direct radiative effects of biomass burning aerosols are estimated.