I very much appreciate this study.  I reviewed Gao et al. (2021), and while I thought it was a clever idea and worth discussion in the literature, it was easy to see some potential problems with adding absorbing aerosols to the stratosphere.  The present study provides much of that context and does a nice job with it.  Furthering the discourse is exactly what needs to happen.  I am recommending revisions, some of which could take a bit of work.

General comment:  None of the figures has statistical significance calculations, so it is difficult for me to understand whether the values I’m looking at are important.  (I suspect they are, but I’d rather not guess.)

Introduction:  Some discussion of the work of Wake Smith is warranted.

Lines 154-155:  I wonder if you’re missing a few mechanisms here.  Ozone is a greenhouse gas, and it will certainly change in G6abs.  Stratospheric water vapor is too, and G6abs has strong heating of the tropopause cold point.  And now that I’m looking through the paper, I don’t see any mention of stratospheric water vapor, which seems like an important oversight.

Lines 182ff:  How do these results compare with the hydrological cycle impacts of tropospheric BC and sulfate?  Are the mechanisms related?  Also, you don’t provide a mechanism –Bala et al. (2008, PNAS) seems be relevant here.

Line 281:  This is a bit of an overstep.  I agree that if the models get all of the dynamics right, they should all show this response.  I have to imagine there are some models that have problems here.

Figure 8:  Can you report slopes and R2 values?

Section 5:  Your results for the QBO are dependent upon these being equatorial injections – see Richter et al. (2017, 2018) and Kravitz et al. (2019).  I’m not disputing your results, but I do think your description needs this qualification.

Review of Haywood et al., Assessing the consequences of including aerosol absorption in potential Stratospheric Aerosol Injection Climate Intervention Strategies, https://doi.org/10.5194/acp-2021-1032

This study is examining impacts of absorbing aerosol injected with sulfate (which also partially absorbed) for stratospheric geoengineering.  The responses noted are not new, in that some of these have been noted in papers using much more absorbing aerosol than employed here.

One example (which is referenced here) is the paper by Ben Kravitz (Kravitz, B., A. Robock, D. T. Shindell, and M. A. Miller, 2012. Sensitivity of stratospheric geoengineering with black carbon to aerosol size and altitude of injection. J. Geophys. Res. 117, D09203 (2012).) that discussed how using large amounts of black carbon for SAI purposes had potentially detrimental impacts on climate. However, Kravitz et al. used black carbon as the SAI material, and used a large amount, 1 Tg BC per year.  As noted by references in this paper, it has been previouslyl established that using BC as the SAI material causes detrimental effects, and, the Gao et al. proposal being criticized here also notes that issue.  The Gao et al. paper uses .01 Tg/yr.  The authors of this current paper note that they are using a factor of 10 more BC material in their assessment than Gao et al. did, and effectively introduce this study as a criticism of the Gao et al. proposal.  However, because of that factor of 10, it’s really not a fair criticism of the method. What would be a useful addition to this study is something akin to a toxicity study, where one determines at what concentration a substance is toxic.  I recommend examining consequences at 1X, 3X and 10Z the .01 Tg/yr case.   This study is not including quite enough for BC to be the SAI material (like the 1 Tg/yr used by Kravity), but it used way too much to be considered as a lofting material as discussed in Gao et al. Consider something like a medical supplement; a small amount of a vitamin is beneficial for health, but a large amount is detrimental.  You would not then recommend not taking the vitamin at all because a large dosage is detrimental.

The other point that these authors missed in regards to the Gao et al. paper is that they used BC as an example, but also suggested using Brown Carbon for the initial lofting, which would then break down in the stratosphere, and the heating effects would be different. 

A sensitivity study looking at varying amount from very small (.01 Tg/yr) to the values assumed in this study would be useful.  Are the feedbacks and consequences really linear with the forcing?  A sensitivity study could examine whether you get 10X the response using 10X the amount of BC.  It could be smaller, or larger, and that would be useful to know.

Overall, it seems this paper was inspired by the Gao et al. proposal, but really isn’t definitively proving or disproving whether the proposal is actually harmful or not.  Something akin to a toxicology assessment would be really useful, and I’d encourage the authors to consider doing something like that.