Dear authors,

Thank you for submitting your revised manuscript. I sent it to the three referees of the prior submission; two felt it was now acceptable for publication, while the other (referee 2) felt that the revisions did not completely address their comments. After a careful read, I agree with this reviewer and would like to request some further minor revisions. I have reproduced the reviewer comments below:

“The authors reply to my first major concern well. Although the authors indicate that they do not include the material in the paper, I suggest adding one section and including this result in the paper. The title of the section would be “testing selection of wavelengths for aerosol applications” or something similar and discuss how the selection of wavelengths made by Barker et al. performs for aerosols.

In replying to my second major concern, the authors compared CALIPSO and RXS-expanded attenuated backscatter profiles with ground-based observations. The relative error of RXS-expanded profile is slightly worse than the relative error of CALIPSO profile. This is not surprising since aerosol properties are relatively horizontally uniform compare with cloud properties. But this result does not address my point that I initially raised. What is the motivation of applying the construction algorithm to aerosols? The authors state in the reply that MODIS does not provide vertical profiles. But CALIPSO does. What scientific questions the authors try to answer by applying the construction algorithm initially developed for clouds to aerosols? The authors might say that the objective of this work is to test the construction algorithm for aerosols. And the objective of this study is mentioned on page 3. But for what purpose and why do the authors think the algorithm is needed? Does the construction algorithm gives a better knowledge of aerosols that MODIS (horizontal coverage) and CALIPSO (vertical profiles) cannot provide? If so please clearly state and demonstrate in the paper.”

On the first point, you had performed the test but omitted the results from the manuscript to keep the method selection concise; I appreciate the additional effort here. However I agree with the referee that there is value in adding this information briefly into the revised manuscript), in case other readers wonder about wavelength selection. I do not think this would make that part of the paper too long.

On the second point, this is perhaps in part a philosophical matter. My understanding from reading your paper and response is that while the AOD you get is not as good as MODIS standard, and the profile has some limitations compared to CALIOP standard, what you see as the main benefit is getting some profile information over a broader instantaneous swath than available from CALIOP alone. So, while both MODIS and CALIOP provide near-global coverage on a long-term basis, the important thing here is the gain in profile information on an instantaneous basis, which might be important for some instantaneous applications (e.g. feature tracking, air quality). Is that right? I suggest adding an additional sentence or two signposting this more clearly. I do not think it needs to be demonstrated on those applications in this paper, but direct statements would be useful.

Please let me know if you have any questions; I look forward to receiving your next revision. I expect this will not need to go out for another round of reviews, as the above are fairly minor comments.

Best wishes,

Andrew

Dear authors,

Thank you for your further revision. I appreciate the additions you made in this round, and after reading through am pleased to accept your study for publication in AMT.

While checking this latest version I noticed that on page 3, line 8 the word "boarder" should instead read "broader". This typo can be fixed when you upload your files for the Production Office.

Best wishes,

Andrew