The manuscript submitted by Zhang et al. uses the global Earth system model ECHAM6-HAM to investigate how idealized perturbations of BC and SO4 influence precipitation. They divide precipitation responses into the rapid and the slow surface-temperature-mediated response, and find – as earlier work – that BC-induced precipitation changes are dominated by the rapid response while SO4-induced changes are driven mostly by the slow response. Authors analyze both rapid and slow responses in terms of the energy budget, and further decompose every term into contributions from clouds, aerosols, and clean-clear-sky. This approach allows for a thorough understanding of which processes underly the different changes. The paper is very well written and based on sound scientific analyses. While the results are unsurprising, I believe that the clear scope, the thorough method, and the systematic and accurate account of results makes the paper a good contribution to the field, and I recommend the paper to be published after relatively minor revisions.

General comment:

As the manuscript is so well worked through and so well written my comments are few.

Specific comments:

• L 79: “lacks agreement in both model simulations and observations”. Please clarify – is there little agreement between typical model results and what observations show, or is there both deviations between models and observations as well as large inter-model disagreement?
• L 88: “Fast and slow responses are essential in determining changes in precipitation” This one is a bit unclear as to what the authors mean. Do you perhaps mean that identifying fast and slow responses is essential to determining the causes or processes behind the precipitation change? Consider rewording this sentence.
• L 128: “on different scales” – Please clarify, is it different timescales (fast vs slow) you mean here?
• L 198: I’m not familiar with ECHAM6-HAM – does this model include any microphysical effects of BC, allowing BC to function either as a condensation or as an ice nuclei?
• L 285: Please fix the subscript in “Rresp”
• L 311: Just for easier reading: please add “fast” before “precipitation” in the beginning of this line
• L 326: What causes the slow BC-induced change in atmospheric cooling?
• L 343: Again, just for easier reading it would be good to remind the reader that this sentence is regarding SO4, not BC (which is mentioned on the previous line).
• L 490: “largely balance and less balanced” – does this refer to global and extratropical respectively? Please consider rewording the sentence to make this more clear.
• FIG 2: Consider adding a short sentence in the caption explaining what values of e.g. -1, 1 and 0 would mean, to make the figure easier to understand without reading the manuscript text.
• FIG 3 and 4: Please increase the font size of the legend in panels e)
• FIG 5: In the legend, you use e.g. ARCaero and ARCcc, while in the manuscript text and also in the figure caption it differs whether you use ARCaerosol /ARCaero or ARCcc/ARCclean,clear. Please make this consistent throughout the text and figures. Also, please increase the font size of this legend.
• FIG 6 and 7: Could you, to help the reader, consider adding a short sentence in the caption, explaining that red/blue is descent/ascent in d-f?

The global aerosol-climate model ECHAM6-HAM is used to investigate the response of precipitation to perturbations of black carbon and sulfate based on examining changes in atmospheric energy budget on global and regional scales. Precipitation responses are divided into the fast response due to black carbon and the slow response due to sulfate. Overall, the manuscript is well written and good analyses have been provided. However, there is an important concern regarding the approach of the study, which includes application of a general circulation model (GCM) for unraveling precipitation responses on regional scales to aerosol perturbations. Further comments are listed below.  

Although the manuscript is generally well written, some minor syntax errors need to be fixed. In addition, long sentences should be avoided. For example, in line 60-62.

Lines 31-38. The uncertainty of aerosol indirect effects or, maybe more precisely, different or even opposite conclusions, which have been obtained for the indirect effects of aerosols in different studies, are related to different environmental conditions, including relative humidity, vertical wind shear and cloud types. This needs to be clarified in this paragraph and the works of Khain et al. (2008), Khain (2009) and Alizadeh-Choobari (2018) can be cited.

Line 76. The recent related work of Keshtgar et al. (2020) can be also cited here.

Line 105-108. Here the lack of regional studies is emphasized to convince conduction of the study. However, a GCM) with a relatively low resolution (1.9 degree) is applied in this study, which is not appropriate to demonstrate precipitation responses on regional scales. Considering the fact that precipitation highly varies on regional scales, the authors should explain how application of a GCM can be helpful to unravel responses of precipitation on regional scales to aerosol perturbations.   

Line 153. It should be explained why BC and Sulfur dioxide have been increased differently, and how it can impact the obtained results.  

References

Alizadeh-Choobari O (2018) Impact of aerosol number concentration on precipitation under different precipitation rates. Meteorological Applications 25, 596-605

Khain, AP (2009) Notes on state-of-the-art investigations of aerosol effects on precipitation: a critical review. Environmental Research Letters, 4, 015004.

Khain, AP et al. (2008) Factors determining the impact of aerosols on surface precipitation from clouds: an attempt at classification. Journal of the Atmospheric Sciences, 65, 1721–1748.

Keshtgar et al. (2020) Seasonal and interannual variations of the intertropical convergence zone over the Indian Ocean based on an energetic perspective. Climate Dynamics 54, 3627-3639

Review of Zhang et al. 2020 ACP

General comments

This study evaluated the impact of artificially perturbed aerosol emissions on precipitation in ECHAM-HAM and investigated the underlying mechanisms by looking at individual energetic terms. Following an earlier study, the authors separated the fast and slow responses using atmosphere-only and slab ocean simulations. The major conclusion on global mean precipitation response is similar to previous studies, i.e. both BC and sulphate emission increases reduce the precipitation, but the impacts are through different processes (fast process for BC, but slow process for sulphate). They further compared the decomposed terms of atmosphere radiative cooling and found the largest terms are from aerosol radiative effect for BC and clear-clean sky effect (mainly due to reduced water vapor) for sulphate. Overall I think this study is useful and it provides new information to understand the climate impact of aerosols. Nevertheless, I think the description of slab ocean simulations should be improved and the interpretation of the results needs some clarifications. Below please find my specific comments. 

Specific comments

Page 1, abstract, line 24: Non-absorbing aerosols can decrease surface temperature by scattering shortwave radiation, which is a fast process. Do you mean a different mechanism? 

Page 3, line 84-85: do you mean the atmosphere-only model? 

Page 3, line 102: “which is” -> “which are”

Page 3, line 104: is LW radiative cooling the only term? How about SW changes due to aerosol-cloud interactions? 

Page 5, line 153: is the emission perturbation for all or anthropogenic sources only? 

Page 5, line 158: please provide more information on how the MLO model was setup in your simulations. For example, how were the QFLUX data derived for MLO? 

Page 5, line 160: the time when a slab ocean model reaches equilibrium state is dependent on the model physics and how large the model response (to the perturbation) is. Different models might need different time to reach equilibrium. It’s better to check it in your model (e.g. following the method used in Samset al. 2016). 

Page 5, line 170: it would be useful for the readers learn more details on how the energy transport term is diagnosed, e.g. did you calculate it online or offline? 

Page 6, section 3.1, table 1: please check the numbers in the table. It seems to me that L&P is not equal to &ARC - &SH for most cases. Also, it’s better to use same unit for the terms shown in the table and in the figure. Right now one (table) in W/m2, and the other (figure) in mm/d. 

Page 11, line 401-402: based on the discussion on the fast/slow response, the reduced surface temperature is mainly due to the slow response. The sentence here reads like the dimming effect of aerosols (fast process) leads to reduced surface temperature and the major cooling in NH shown Figure 7a. Is this really what the author meant?