Please see the attached document for my comments!

This is a well written paper. The analysis is thorough and supports the conclusions made. The presentation quality is good. The authors first conduct AERI and HiSRAMS retrievals for ground-based observations of clear sky only, which compares the capability of the two instruments in terms of information content on temperature and water vapor. Then they demonstrate a joint retrieval between ground-based AERI and air-borne HiSRAMS clear sky and show the benefit of combining the two instruments. I don't think this paper delivers new science, but it does present a retrieval algorithm for atmospheric temperature and water vapor sounding from IR and MW. I have the following comments for authors to consider:

1. The value of this paper is on combining air-borne microwave and ground-based infrared sounding measurements, not much on whether or not the MW and IR combined could provide a better "all-sky" retrievals. I think it is better to point this out in your title, abstract, and literature survey, and center them on your major findings.

2. The retrieved profiles agree with the radiosonde surprisingly well, but clear differences with the first guess can be seen. Are these results all for a single case? How does the algorithm perform in general? It is hard to reach conclusions based on a single case.

3. The retrieval bias around the sharp vertical features is very large, which is due to the lack of vertical resolution. This could be demonstrated by calculating the bias by comparing retrievals with the radiosondes that are vertically smoothed by the averaging kernel. Perhaps it is more interesting to ask what kind of instrument designs could resolve such feature, new channels, lower measurement error, etc.

This paper is well written, with good explanations of the OEM

methodology and analysis that supports the conclusions. One could

argue that the Optimal Estimation methodology is well known, but the

paper has practical examples that accompany the explanations.



Having said that, not much new or unexpected conclusions are in this

paper, but it can find a place in an Atmospheric Measurements and

Techniques paper if some fixes are made.



1) Line 470 : it is very hard to tell if the results presented truly

are statistical N >> 1, or if only one case is done for the

instruments alone on the ground, and one other case is done for the

HISRAMS/AERI synergy. Your conclusions are too general if only one

case is considered, especially since what you find is not surprising

(AERI and HISRAMS together give "better" results (closer to truth)

than one instrument alone, and a top/bottom sandwich brings a good

deal of information together).



2) Similarly, are you really doing an clear-sky retrieval. How do you

know there are no clouds? Your paper gets confusing at the beginning

when you mention "clear sky" in the title and then talk about the use

of HISRAMS in allsky conditions. Can you clarify?



3) Figure 2,6 : Please show mean(observations - calculations) and

std(observations - calculations) for both instruments. You could be

getting great results while having poor spectral fits (ie the bias and

std. dev are larger than the NeDT of the instruments).



In particular could you indicate the surface channels of the HISRAMS

when showing the biases?



4) The paper could be much better organized. For example instead of

"sandwich" you could use the more traditional "synergy" term.



5) Figures S5 - S8 are ..?



6) Lines 155-162, and lines 174-180, should be combined together

rather than being separated.  Then again line 189 the n_{level} is

mentioned.



7) Similarly lines 434 to 439 could be moved to the conclusion