Dear Authors,

Thank you very much for your letter and the revised manuscript.

I will not respond to all of your points individually, but let me stress that editorial standards require that the review is concerned exclusively with the present manuscript.

The revised version of your manuscript takes many suggestions into account and the reviewers will have their say whether the new version responds to their different requests. The reviews of your submission were very positive and unanimous about the scientific relevance of the paper. They have fully confirmed that your contribution is suitable for publication in "Atmospheric Measurement Techniques" (AMT) and that it will constitute an important contribution to the scientific community. Note, however, that all three final review reports as well as a fourth independent review of an earlier version of your paper were very critical about the paper length, to the point that the presentation quality of the paper has been seriously questioned. This unanimous analysis cannot be ignored even though AMT does not impose a fixed page limit. It is a profound concern, because presentation quality is one of three key criteria for publication in AMT. Based on these independent expert reviews, I can only consider the article for publication after a major revision concerning the length of the article and the clarity of the presentation. Note that I agree with the analysis of referee #2 (who has provided the longest list of specific comments) and summarized that the paper "is unnecessarily long (which distracts from the main content) and it is full of typos, minor inconsistencies and little errors. Many sections make the impression of hastily written drafts that were not proof read." Referee #4 comes to the same conclusion by stating "I feel that the paper might be easier to follow and make a stronger case for the conclusion if content in the main body were limited to what is needed to support the conclusion, and use references or supplemental material otherwise."

In your accompanying letter as well as in the replies to the referee comments you present arguments that do not seem to allow shortening the manuscript. One particular important point is the upright strive for providing complete and reproducible science. Let me point out that shortening the manuscript should not be incompatible with complete and reproducible science.

First, you have been worrying that the article would not be complete and self-consistent if you were to shorten the manuscript. Note that for means of documentation, or reasons of completeness, content (such as paragraph 2.3.3, see further below) can be presented in the supplementary materials or in appendices. This includes text, figures and tables. Such choice would allow to streamlining the narrative in the main part of the manuscript, increase the presentation quality of the article and keep your submission complete. Note also that both the discussion version and the final version will be conserved and made accessible on the AMT website. They will both have a doi and can both be cited whenever necessary. If, in the future, some content that you would like to cite is only in the discussion paper, but not in the final version, you can always refer to the discussion paper.

Second, a complete and self-consistent article does not have to be exhaustive in all aspects. Parts of the current article can well be shortened without impacting the message and compromising the integrity of the paper. For example, while not necessarily prohibited, one needs to question the usefulness of the paragraphs on Earth and Mars (2.1.3) as well as on ground-based observations (2.1.1) in a paper entitled "The use of O2 1.27 μm absorption band revisited for GHG monitoring from space and application to MicroCarb". Clearly, these paragraphs can be omitted. This issue has already been raised by the reviewers.

Third, as discussed by referees #1 and #4, section 2.3 could be shortened significantly, as parts of the presentation is already present in the literature. I do not insist that the section has to be removed completely, just because equations 14-15 can be derived from the equations in the Sun et al. (2018) paper, but you should 1/ considerably shorten the paragraph (or defer large parts into a supplementary section), 2/ refer as much as possible to the literature and to published work (see also referee #4) and 3/ point out that eq 15 can also be derived from equations in Sun et al. (which shows that the two approaches are consistent). At this point it needs to be stated very clearly that AMT is open to researchers from all countries and nationalities and that it follows universal publication standards. As a consequence, already published work should not be repeated without obvious reason and must be clearly credited whether this work has been published by any other journal.

Fourth, some paragraphs contain superfluous and misleading information and have not been proofread carefully. For example, a significant portion of paragraph 2.3.3 is concerned with the calculation of Q'. As far as I can see, this quantity has no practical use and does not provide any particular physical insight. What does it mean that "Q'/Q is approximately between 0.1 and 1", apart from checking that you likely (but not surely) have correctly calculated the internal partition sum of a linear rigid rotor molecule? The lack of importance of Q' (for the resent paper) can be seen from the fact that the quantity is not at all referred to in any calculation/formula later on. Why should the reader try to understand this quantity that is presented over one page of the document, given that the same quantity is not used any more? Another source of confusion in the paragraph is the unnecessary hopping between state energy and transition. First, eq 10 is defined as a sum over v,J states (with state energies). Then it is pointed out that "The summation must be not over all the transitions, but over all energy levels, because from a given energy level having a certain population N2i, there are several transitions going down to the lower level with different A21." This remark is not only redundant, because the formula and the text already explain that the sum extends over all states, but the reasoning "because" is completely incomprehensible. On the next page in the caption of Fig. 4, the confusion goes on. The plot shows the individual terms that contribute to the partition function as a function of individual states with quantum number J', but the legend discusses transitions (between states). Once more, "the distribution of O2* molecules among the various energy levels" which is the title of the paragraph has nothing to do with transitions to other states (if the distribution within rotational states is given by thermodynamic equilibrium). In the same veins it is unclear why equations 7 and 8 are listed in this section. Apart from notation, radiative processes do not at all impact the distribution of rotational states in the vibrational ground state of O2*. These processes (and this is also clear from the equation used in the paragraph) only become important in section 2.3.4., where the deactivation rate is calculated. In any way, it seems that much of the calculation/presentation can be either completely removed and/or deferred to the supplementary material. The important approach in section 2 is the derivation of Eq (15) which is used in paragraph 2.3.7 to calculate the emission rate of O2* from the absorption spectrum of O2.

In conclusion of this exemplary discussion, there are various ways to considerably shorten the length of the article without compromising on the message/content and on the impact of the article. To the contrary, visibility and readability of the article will greatly gain from such a reduction in length and you should try to point out the essential features (I have tried to argue that a considerable amount of statements of your presentation are not very relevant). The only non-trivial part in the calculation of Q' is that J'=2 is the lowest accessible state and that even and odd J' states can be populated. This is different from O2 in its electronic ground state. But no reference is provided for this case. The different reviewers have also made various valuable suggestions to shorten the article and to focus on the main content. Apart from the above propositions and the reviewer suggstions, it is your choice as the authors of the article to select the most appropriate paragraphs. As a rule, the strive for reproducibilty should not have a negative impact on the readibility of the paper. Calculations that are given for reasons of completeness and not for providing insight are ideal candidates for the supplementary materials.

In any case, the article should also be proofread very carefully. The above cited sentence on the summation is only one example of several phrases that don't make any clear sense. Another example for an ambiguous phrase in the same section is "they will soon re-equilibrate among the various rotational upper levels because the lifetime is quite long versus the collision time with ambient molecules, without changing ΣN2i/N." The sentence is misleading, because without collisions the rotational states cannot equilibrate. You probably meant to say that the lifetime of collisional re-equilibrqation is much shorter than quenching into the electronic ground state.

I conclude my remarks by reminding that the issue of length has already been raised at the earliest review stages. I sincerely hope that the three detailed and independent review reports (and even four when also counting the reviewer who in the meanwhile has ceased to further review the paper due to time constraints) have convinced you to considerably shorten the manuscript.

With kind regards,

Christof JANSSEN
Associate editor
Atmospheric Measurement Techniques

Dear Authors,

While most of the referees have given a very positive evaluation and the scientific quality of the manuscript is out of question, one referee has still expressed serious concerns about the presentation quality of your manuscript. Given that I also have stumbled over many details (see my individual remarks further below),
I request that the paper gets revised once more before being published. The paper also concludes on results that are only presented in the Appendix. Please include the parts that are essential for your conclusions and stay as concise as possible. I agree with referee 3 that Appendix A can be suppressed altogether.

The manuscript is still hard to read at times and it is therefore recommended that you improve on the readability of the paper.

When revising, please take into account the remarks of all referees.

p 2 l 16 : The atmospheric fraction is the ratio of the atmospheric increase of CO2 mass to the mass of CO2 anthropogenic emission
p 2 l 33 : CEOS 2018 ???
p 2 l 35 : build -> built
p 2 l 35 : give name of the chinese space agency instead of country
p 3 l 15 : the acronym TCCON is not defined
p 4 l 10 : carry -> record
p 5 l 13 : phrase is redundant with figure caption
p 6 l 1 : Delete We have made use ...
p 6 : Number appendices in order of appearance. Describe the structure of the paper first and Appendices later.
p 7 l 11 : fundamental -> electronic ground
p 7 l 37 : form -> from
p 8 l 5 : what is an absorption emission spectrum ?
p 8 l 17 : as a part of the OSIRIS instead of a part of OSIRIS
p 8 l 21 : spectral capability (resolution l/dl~850) -> resolving power (L/dl~850) of the
p 9 l 2 : reducing cross-track -> reducing the cross-track
p 10 l 11 : change cm3 -> cm2, s1 -> s
p 10 Fig. 4 The figure should go into Appendix. It is neither discussed nor does is seem to be particularly important to what is developed in the paper.
p 12 l 11 : nearer 1 for larger -> closer to 1 for higher
p 19 figure 11: There are no panel numbers on the Figure (a, b, c). please denote by top middle and bottom panel. Corect the text accordingly.
p 22 figure 13: The figure does not comply with publication standards.
p 23 l 11 : Delete second part of title. This is explained in the text.
p 24 l 14 : Delete phrase : The wavelength unit pm ...
p 25 l 26++ : Use _a instead of _aero. Avoid using _aer and _aero interchangeably.
p 27 figure 16 : panels are interchanged. top is high albedo, bottom is low albedo. Avoid cryptic notation that is not used in the publication (Lmax, Mmin)
p 27 l 12 : Tr is used as a symbol for transmission, whereas it is T later on (section 6.2.1)
p 27 l 16 : replace changing continuous by continuous
p 28 l 24 : The header (LATMOS inversion breadboard) is not very informative about what is done in this section. It should be chosen differently
p 29 l 21 : symbol TH2O is not explained
p 29 l 12 : please explain why water enters into the modeling. So far the article has been concerned with O2 and all simulations have been done without H2O.
p 29 l 22 : intensity is not the same as an albedo.
p 29 l 20 : why does K5 appear as exponent in eq (11) ?
p 30 figure 17: Avoid SCIA as an abbreviation for SCIAMACHY. Note that you have already used CIA as abbreviation for collision induced absorption. You have used SCIAMACHY everywhere else.
p 30 l 11 : intensity -> intensities
p 31 l 14,15 : g should be a subscript. Use same notation of singlet oxygen everywhere.
p 31 l 12: values haigher -> higher values
p 33 figure 20: One can hardly distinguish dark violet points (morning) from black points (Reprobus)
p 34 l 19 : so far, you have been concerned with the determination of the O2 column. Psurf shuld only be introduced as an auxiliary, not as the main parameter.
p 35 l 8 : similarly use O2 column instead of psurf.
p 35 l 18: This airglow is mainly due to photolysis of ozone in the mesosphere, letting an O2 molecule in an excited state O2* which spontaneously de-excite. -> This airglow is mainly due to the spontaneous relaxation of excited oxygen (O2 Delta) that is formed in the photolysis of ozone.
p 35 l 23: absolute amount seems to be in contradiction to contamination. replace "absolute amount of ... contamination with "airglow radiance".
p 36 l 2 : In summary, we have found that the intensity of the O2* airglow is well organized. Probably you mean something like well behaved. The word organized is very confusing here.
p 36 l 5 : what is a meteo field ? meteorological or meteorology field ?
p 36 l 7+: This is a new result and should not suddenly appear in the conclusion section without having been presented as a result before.
p 36 l 38: It is not clear which simulations you refer to : "This allowed performing some simulation exercises with good confidence about their ability to represent reality, showing that with the resolving power of 25,000 of the MicroCarb instrument, it is indeed possible to disentangle the airglow emission from the O2 absorption in the O2 band at 1.27 μm."
p 37 l 4 : These are new results that have not been discussed in the previous part of the document, but rather in Appendix E. It seems to be necessary to include parts of Appendix E in the main section.
Appendix : Check references to equations in text. There are plain numbers in the text whereas equations are numerated by the Appendix Number + Equation Number.
p 65 l 12 : ratio emission/absorption -> emission/absorption ratio
p 65 l 23 : Q_tot^up is not defined

Dear Authors,

Thank you very much for completely addressing all issues that have been brought up in the previous round of refereeing. All referees agree that your paper is original and should be published in Atmospheric Measurement Techniques and I congratulate you for this accomplishment.

However, in rereading the latest version of your article, I still have found some technical issues that must imperatively be addressed before the article can go into the publication phase.

Please correct these :

p 1 Lines 6 - 12. It seems that addresses 1 and 6 are effectively the same. If so, please remove affiliation 6.

p 7 l 6&7. Please delete ", or the light emission when the molecule deexcites spontaneously." The phrase is not compatible with common rules. In the chemical literature (ACS style book (p. 262)) the star designates an electronically excited state and not the process of de-excitation involving such a state. Note that both EGU and AGU embrace recommendations on notation from the chemical societies, and from ACS in particular. Note also that you mostly write O2* emission or O2* airglow instead of simply O2* when you refer to the process of photon emission via de-excitation of O2*.

p 7. The symbol for the rayleigh is R. If used with SI prefixes, megarayleigh should be MR and kilorayleigh kR.

p 8 l 12 : Replace "joule "by "J"

p 9 l 1,3 : remove "," in the wavelength specification. The use of a thousand separator is inconsistent with usage elsewhere. It is thus confusing.

p 9 l 24 : Please remove square brackets to designate wavelength domains.

p 16 - 21 : 3 or 4 different referees have all criticised the confusing use of different notations for O2 in its first electronic excited state (O2*, O2(Δ), O2(1Δ), O2(a1Δ), O2(a1Δg)). While the use of these different notations is not incorrect in the first place, it is a possible source of confusion for the general reader if different notations are used in an arbitrary fashion. Since AMT imposes the consistent use of symbols/notation, I request to replace all instances of O2(a^1Δ_g)) on pages 15 - 21 by O2*, except for appearances in equations, Figures or on p 16 l 35 (which refers to Fig 9). In addition, O2(a1Δ) should be replaced throughout the whole manuscript by O2(1Δ). This would remove one notation and bring the text consistent with your explanatory remark on the use of notation on the top of p 7, where O2(a1Δ) is not listed.

p 21 Fig 12: The curve labelling outside the legend box is not readable. Since figures are already given in the legend box, the labels should be suppressed altogether. Also change the x-scale so that SCIAMACHY data for SZA = 33.3° are fully shown. The color bar on the RHS is completely superficial given the much more accessible information provided by the legend box. Please remove.

p 27 l 15: "Igor Pro" is a commercial software. It should be cited using the correct name and you should possibly give the version number that you have used. Since it is a company who owns copyright of the product, its name should also be mentioned in the text or given in the reference list.

p 50 Fig A3: The wavelength label "1.27 μm" corresponds to a transition, not to an energy; please remove from graph.

p 74 l 5 : Remove "s" at end of line