Black carbon (BC) is one of the major air pollutions severely threatening public health despite of its relatively low contribution to total PM mass, not only due to its toxicity but also its nature of light absorption. Because of this, BC is able to alter boundary layer stability and structure, which further influence the ventilation of pollutants. Previous studies have demonstrated the importance of this BC light absorption effect in surface layer pollution in many China megacities. This study use APEC Blue period as a natural lab to further quantify this impact using a fully coupled model. The scope fit well with ACP. The manuscript is well written, the results are scientific interesting and politically meaningful supported by sound methodology. I feel this work is suitable for publication after addressing a few minor concerns.

Minor concerns:

1) Calculation of aerosol optical properties has been well described for the external mixture and internal homogeneous mixture (volume-weighted average). I feel the calculation details of core-shell mixture should also be elaborated, given core-shell is the main mixture style discussed in this paper. Please also provide the information of how complex reflective index is defined for each component, especially for organics.

2) There are many ways/definitions of boundary layer top. How boundary layer top is defined and hence its height is estimated?

3) line 283. Do you mean reduce PBLH by 8.2m on average?

4) There is some nice discussion about the PBL-PM2.5-O3 interactions. I think some chemical reasons also influence ozone. Such as, reduce of PM2.5 could enhance the surface layer photolysis therefore increase ozone especially for heave polluted area/periods, and the co-reduction of NOx and the regime of NOx (Chen et al., 2021).

5) Just curious that in Fig. 5h, why reduction of emission could lead to a strong enhance of pollutants in the northwest? Is there some interactions between PM and dynamic lead to the re-distribution of pollutants? This may be out of the scope of this study, therefore do not expect authors’ full answer here. Some discussion would be appreciated and may be an interesting topic for future study.

References:

Chen, Y., Beig, G., Archer-Nicholls, S., Drysdale, W., Acton, J., Lowe, D., Nelson, B. S., Lee, J. D., Ran, L., Wang, Y., Wu, Z., Sahu, S. K., Sokhi, R. S., Singh, V., Gadi, R., Hewitt, C. N., Nemitz, E., Archibald, A., McFiggins, G., and Wild, O.: Avoiding high ozone pollution in Delhi, India, Faraday Discussions, 10.1039/D0FD00079E, 2021.

General comments:

The current work investigated the impacts of changes in BC mixing state during APEC on air quality and meteorology in November 2014 in Beijing. The scope of the research is of interest and well suited to the current journal, Atmos. Chem. Phys. However, the manuscript is not acceptable in its current form, because the model they used cannot answer the authors’ question (please see Comment #1). The manuscript is well written and the logics is fine. It proves that the authors are certainly experts of air quality modeling and aerosol feedback processes. However, they may not well understand aerosol mixing state modeling. The simulation setup is not appropriate and so the recalculation is required. It is recommended for the authors to invite additional expert(s) for the support of relevant simulation setup. Anyway, the manuscript will be accepted in the current journal, after the authors adequately address the following comments, conduct additional simulations, and revise the manuscript accordingly. The general comments are as follows:

1. The simulation setup what the authors named as “core-shell” is assuming 100% of all components mixed with BC, which never happens. This is not the case of Beijing but in Tokyo, only a small fraction (15%) of sulfate and nitrate mixed with BC (Miyakawa et al., 2014; https://doi.org/10.1080/02786826.2014.937477). Thus, there is even a possibility that the reality could be closer to what they named as “externally mixing (0% of aerosols mixed with BC)” than “core-shell (in fact 100% aerosols mixed with BC)”. Anyway, the model they used cannot simulate the fractions of secondary particles mixed with BC. The reviewer recommends to use other models which can consider/resolve BC mixing state or assume fractions of BC mixing for the optical calculation of WRF-Chem that are obtained from observation in Beijing (Zhang et al., 2018) or simulation by other BC-mixing-state-considering/resolving models.

Specific comments:

Introduction

2. 44-45: “In addition to contributing considerably to particulate matter and degraded air quality”. BC usually contributes only 10% or less to the total PM mass. Can it be “considerably”?

3. Fourth paragraph of Introduction: There are plenty of studies regarding BC mixing state in Nicole Riemer’s group using PartMC/MOSAIC. For the three-dimensional modeling, the author might as well cite Matsui et al. (2012) https://doi.org/10.1029/2012JD018446, as the model domain they used are close to those of the current study. There is a mixing-state resolving model called MOSAIC-mix, Ching et al. (2016) https://doi.org/10.1002/2015JD024323, which is the same aerosol module as the authors used. I have read an article reporting that PartMC/MOSAIC is coupled with WRF-Chem. The reviewer does not know if these mixing state resolving/considering models are open to community, but if so, these can be an option for the authors to tackle with the issue raised in the study.

Method

4. Sect. 2.2: Eqs. (11)-(13) are just a general description of the calculation method of aerosol optical properties, which are less important. The last sentences of Sect. 2.2 (Ln.168-170) should be rather elaborated using equations, as it is essentially important for the current study.

5. Judging from Eq.(14), dust is included in the simulation, which is also a light-absorber. How do the authors treat the light absorption of mineral/anthropogenic dust particles?

6. Sect. 2.3: Which wave lengths of AAOD (observation and simulation) did the authors use in their study?

Results

7. Sect. 3.1: Better to compare AOD also, which might be able to partly answer why simulated AAOD was underestimated.

8. Sect. 3.2: rBC is not an indicator of BC aging, but just the relative concentrations of BC and components other than BC. Aging of BC depends not only on abundance of secondary components but also on relative abundance of BC-containing and BC-free particles. If BC-free particles exist more, the secondary components condensed more toward the BC-free particles, which do not contribute aging of BC.

9. Sect. 3.3: E_ab is not indicting the light-absorption enhancement of BC, but indicting the difference of AAOD between 100% internal mixture with BC assumption and 0%.

Throughout the manuscript

10. The paper assumed clear-sky conditions for the whole analysis period. If there are no cloud during the period, please describe it. The reviewer is curious to see the simulated (with different sensitivity tests) and observed solar radiations during the period.