Biomass burning (BB) fractions estimated with the aethalometer model differ from 0 even for pure aerosols from fossil fuel (FF) combustion. We used black carbon (BC) aerosols partially‐coated with non‐absorbing materials to represent typical FF aerosols, and the BB fractions were determined with the aethalometer model. Thus, the estimated BB fractions are the fractions that FF aerosols are incorrectly attributed to BB aerosols. The BC morphology and mixing state have significant effects on the estimation of BB fractions. For freshly emitted BC, the BB fractions do not deviate significantly from 0, and the BB fraction is generally in the range of −1%–10% based on an FF absorption Ångström exponent (AAE) of 1 and a BB AAE of 2. The BB fraction deviates substantially from 0 when BC becomes compact and is coated. The absolute values of the deviations sometimes can be close to 100% for heavily coated BC. The BB fraction is generally greater than 0 for fluffy BC while compact BC generally exhibits a negative BB fraction. The BB fractions of partially‐coated BC are very sensitive to the size distribution and coating ratio, which is consistent with the results of the core‐shell sphere model. We have also performed a series of studies with different configurations for BC morphologies, coating ratios, and mixing states, which show the variations of the estimated BB fractions with atmospheric aging. We also explain the reasons for the uncertainties in the BB fraction estimates and provide suggestions for using different AAE pairs.

Abstract. The absorption Ångström exponent (AAE)-based methods are widely used to estimate brown carbon (BrC) absorption, and the estimated BrC absorption can be significantly different from 0 even for pure black carbon (BC). However, few studies have systematically quantified the effects of BC microphysical properties. Moreover, it is still unclear under which conditions the AAE-based method is applicable. In this work, we used BC models partially coated with non-absorbing materials to calculate the total absorption. Since the total absorption is entirely from BC, the estimated BrC absorption should be 0 if the retrieval methods are accurate. Thus, the estimated BrC absorption (ABSBrC) should be the absorption from BC that is incorrectly attributed to BrC. The results show that a BC AAE of 1 can generally provide reasonable estimates for freshly emitted BC, since at this time ABSBrC is generally in the range of -3 % to 4.5 %. However, when BC aerosols are aged, ABSBrC of about 35 % could be observed. The WDA method does not necessarily improve the estimates, sometimes a negative ABSBrC of -40 % can be found for partially coated BC. By combining simulations of a global chemical transport model, this work also quantified the effects of BC microphysical properties on BrC global optical absorption aerosol depth (AAOD) estimates. The AAE = 1 method could sometimes lead to a misassigned global mean AAOD of about -0.4 – 0.5 × 10-3 if BC aerosols have a complex morphology, leading to a global mean direct radiation factor (DRF) of about -0.068 ± 0.0172 to +0.085 ± 0.0215 W/m2 from BC, which is incorrectly assigned to BrC. The WDA method does not necessarily improve the estimates. In our cases, the WDA methods based on the spherical models can lead to a range of about -0.9 – 0.05 × 10-3 of misassigned AAOD, which could lead to a global mean DRF error range of -0.153 ± 0.0387 to +0.0085 ± 0.0022 W/m2. At the regional scale, the AAE = 1 method in East Asia sometimes leads to a distributed AAOD of over 3 × 10-3, resulting in a BC DRF of about +0.51 ± 0.129 W/m2, which is incorrectly attributed to BrC. Mie theory-based WDA methods would lead to an estimated AAOD error of more than 6 × 10-3 in some regions (e.g., East Asia), resulting in an estimated misattributed DRF of +1.0 ± 0.258 W/m2.