Dear Chun Hang Chau, dear co-authors,

I am now in receipt of the two additional reviews of your manuscript " Simulated mixing in the UTLS by small-scale turbulence using multi-scale chemistry-climate model MECO(n)". I would like to apologize you for the delays in this round of reviews, but I had certain doubts about your responses to Ref#2 and Ref#3 comments from the previous round of reviews and wanted to have the referees own assessment.

As you can see, the Ref#2 thinks that all of his comments were successfully addressed and rates your revised manuscript as excellent on all fronts. There is only a minor suggestion for the further revision that I urge you to adopt in the next round.

However, the Anonymous referee #3, who made some fundamental critical comments in the interactive discussion stage, still does not think (rightfully, according to my judgement) that you have been able to resolve them. This reviewer also appears to remain less convinced about the quality and significance of your manuscript.

I must say that I fully support her/his original comment 1 about the model set-up validation and that from the current state of the manuscript it is not clear, whether the model resolution is adequate for studying the turbulent mixing. This comment is absolutely crucial in my eyes and I also agree that this cannot be addressed by comparing the parameterized TKE and a turbulence index computed from a similar set of variables from the same source. Further, the referee in his new review summarizes 4 points that you need to throughly address (and reflect in the manuscript) for the manuscript to be publishable in ACP. Note that the reviewer mostly does not demand you to redo the analysis, but urges you to clearly identify the limits of your analysis and modify the interpretation of the results accordingly. We should both thank the reviewer for this and the intention to improve the manuscript on all fronts.

To conclude, an editorial decision of Reconsider after major revisions has been reached. When I receive the second revision, I will send it back to the Ref#3, who graciously offers us her/his help also in the next review round and read it carefully myself. Hopefully, we will both find that all of the comments have been successfully resolved and reflected in the manuscript.

Sincerely,

RNDr. Petr Šácha, Ph.D.
Editor
Atmospheric Chemistry and Physics

Dear authors,

considering the last comments and rating of your manuscript by referee#3 and based on my own editorial review I regret to inform you that major revisions are needed before your manuscript can be considered for publication in ACP.
In addition to the comment by referee#3, I list below a set of major and minor comments that should help you to improve the manuscript. Please take your time to prepare a significantly revised version, where all the major comments will truly be reflected. After we receive the revised version, I will send it to two new referees for a review.
I am sorry for a lengthy and frustrating review process, but I hope that the constructive criticism will help you in improving the manuscript significantly.

Best regards, Petr.

Major comments:
1) Motivation and scientific question - As it comes now, the manuscript does not convey a sufficiently new or important message. The importance of UTLS composition, of the role of turbulent mixing in shaping it and of the issue of how well high-resolution simulations can capture it, is beyond any question. However, after reading your manuscript I am struggling to disentangle the motivation and goal of your manuscript. Was the aim to produce a detailed case study of turbulent events above Scandinavia? To test the skill of high-resolution simulations for turbulent mixing in UTLS? To test the performance of the parameteriation and get more insight in how the turbulence scheme by Doms et al works? Or to introduce a novel delta tracer methodology for analysing the mixing in the models?

For the first two questions, you would need to bring observations in for validating the model. For the latter two questions, the manuscript would have to be reconstructed with a focus on the type of a scientific question and a more detailed analysis in this regard would be needed.
Moreover, I think that there is a possibility to incorporate the question of mixing sensitivity to resolution from the supplement to the main manuscript and elaborate more on it.

2) Do I understand correctly that you are interested only in the part of mixing that is unresolved by the model and is mediated by the Doms et al. (2018) parameterization scheme? If so, the scheme needs to be detailly introduced in the paper, including all the underlying physics and technical details and review of previous validation efforts of the scheme.

3) Thorough meteorological analysis of the case studies needs to be done, including identification of regions of dynamic (e.g. Ri) or static instabilities (e.g. using potential temperature contours).

4) The methodology of computation of tracer-tracer and delta tracer correlations needs to be clarified, because right now it is not clear how they were constructed from the model fields (e.g. sampled over fixed volume regions over time of the simulation?)

5) Section 3.1 does not have any good meaning and should be omitted from the revision.
Definitely, the analysis presented cannot be used for the argumentation (however vague):" To conclude, the model is able to represent turbulence at a reasonable position (and time)."

Or was the intention to test the skill of the Ellrod index in detecting the turbulence in the model?


Minor and technical comments (chronological ordering):
L29 local changes in the energy budget
L30 affecting-> affect?
L67 the MESSy-fied - MESSy abbreviation has not been defined yet
L129 conducted in the supplement
L131 ..hence the mixing - a verb is missing?
L141 .. the Ellrod index does not fully representing?

Figure 5 and a majority of the following figures - the units are missing! Also what is the meaning of the (kg-1kg-1) unit? If the unit is indeed correct, why not to write kg-2?

Caption of Fig. 5:.. difference (vdiff on - off) (c) Inverted O3-like tracers,
(d)O3-like tracers at 2016-02-05 18:00. - Please rephrase for clarity.

L180-181:.. as the cross section of Figure 5. - in Figure 5?
Btw. where is the cross-section located?

L187 Figures 6c and 6d show
Sections 3.3.2-3.3.4. define the tracer differences and the whole delta tracer-tracer correlation by equation and discuss the results more rigorously.
L205.. The delta tracer-tracer correlation can also be used to determine the direction of vertical mixing. - but later in the text you mention only 1) no vertical mixing and 2) bi-directional mixing scenarios!
Please pay attention to proper definition and discussion of mixing regimes, including balanced/unbalanced and reversible/irreversible regimes invoked later in the text.

The part of the manuscript showing the results of delta tracer-tracer decomposed according to dynamical quantities and then normalized by the gradient is the strongest part of the paper that can elucidate how the turbulent parameterization works. But, more detail is needed how the plots were constructed.
L243 ..figure 9d....figure 7d of case 1 have a similar N2 on both ends - You mean 11d and 9d?
L255 2->two
L256 s most likely due to the advection, considering the completely different wind field in Figure 8 and tropopause in Figure S17 - This feature has to be disentangled (maybe a 3D resolved overturn encompasses the unresolved turbulence region) and the methodology suggested by Ref#3 can help here.

L262 ... such that the mixing is almost equally balanced. What does this statement about the mixing type mean?

L272-273 To conclude, vertical turbulent mixing by CAT in the model simulations leads to an enhanced and significant tracer mixing in the UTLS region. - Is there any novelty in this statement? This seems to come from the definition of the turbulence parameterization.

L275-276 Strong dynamical forcing like vertical wind shear could lead to mixing even
in the stable atmosphere with a typical stratospheric N2 value. - This is a textbook knowledge, intensively studied under the term dynamic instability?

L280 provides a reliable tool - this was not assessed in the manuscript
L283 ..MECO(n) is presented.-> MECO(n), is presented.
L286 .. balanced and imbalanced bi-direction mixing - balanced/imbalanced mixing has not been defined/introduced before. I had the impression that the motivation was to analyze the direction of mixing. Nevertheless, there is a great potential to define the different mixing regimes robustly and provide a systematic analysis on these aspects in the revision
L287-289 The simulated turbulent kinetic energy (TKE) is spatially and temporally well matched with the (post-simulation) diagnosed Ellrod Index, showing the model is able to generate turbulence in the UTLS in agreement with the gridscale wind field data from the model output. - > I assume that the Doms et al. parameterization has been validated before against observations.

L293-294 ...which located near the tropopause (case 1) experiencing the strongest mixing considering the high vertical wind shear and tracer gradient. -> please rephrase.
L295.. when measurement data is available. - you mean that there are not any data even from targeted campaigns existing that would be suitable for validating the model?
L300 .. both, the forcing or the pre-existing tracer gradients are the dominant drivers for the exchange. ->please rephrase
L302-305 These events can be irreversible, i.e., the exchange
of tracers happens along the diagonal of a delta tracer-tracer correlation, leading to a disturbance of typical stratospheric or tropospheric chemical compositions in the respective parts of the atmosphere with implications for climate, e.g., via the
radiative impact of exchanged species. - From the definition of a delta tracer-tracer correlation, these events are unresolved in the model and mediated by the parameterization, hence inherently irreversible. However, is this unresolved process the only irreversible mixing process in the model given that your resolution may be sufficient for resolving parts of the overturning?

Dear authors,

let me congratulate you for successfully resolving all the remaining concerns that in my eyes prevented publication of your manuscript.
Especially, I like that the presentation of the method stands now in the forefront of the paper.
Hence, after reading the revision, the editorial decision of "Publish as is" has been reached.
Due to the technical nature of the paper, I have not proposed it for a highlight paper.

Finally, I would like to apologize you for a lengthy review process. Hopefully, you have found it beneficial in the end as well.

Thank you very much again for publishing with ACP
and wishing you all the best for your future work.
Best regards,
Petr Šácha.