The authors apply an ozone tagging method in a global chemistry-climate model to attribute the origin of surface ozone pollution in Europe with focus on the Po Valley and the Benelux.

The work is carefully done but it’s not clear to me that there is anything new in the methods or results.  I felt that I was reading a technical report rather than a scientific paper, with a tedious deluge of numbers and figures but no real new insight about the origin of ozone.  The source attribution for ozone is consistent with what has been documented in many previous papers. The model is not particularly successful at reproducing observations, so it’s not clear to me that the source attribution here deserves any more confidence than previous studies. I don’t think that this paper is up to the scientific standards of ACP. Maybe I’m missing new scientific insights coming from the paper because they are not properly advertised and/or buried. I couldn’t find them in the abstract. I would suggest that the authors submit a much shorter paper focused on what is scientifically new in their results, and including proper citation to the literature.

Specific comments:

1. The introduction discusses at length the difference between perturbation analysis and attribution by tagging. This is an old story and I don’t find it particularly interesting.  There’s nothing wrong with tagging, it just shouldn’t be interpreted as a linear response to a perturbation, and we can leave it at that.  It would seem more appropriate for the intro to review past relevant studies on attribution of ozone pollution in Europe - this is lacking.
2. Although the writing is generally fine (albeit tedious), there are a lot of minor grammatical mistakes and typos that could be corrected by a copy editor.
3. Page 10, line 12: under- rather than overestimated? There is general ambiguity in referring to frequencies as being under- or over-estimated.
4. Page 10, line 19: surface ozone overestimate is attributed to excessive downward mixing but the model also seems to be too high in the free troposphere based on the aircraft comparisons.
5. Page 20, line 26: I’m surprised that ozone would be produced from ships by ship NMHCs. My understanding is that the ozone production efficiency from ship NOx emissions in models is very high because the chemistry is strongly NOx-limited, unless some specific model parameterization is used to age the NOx faster but that doesn’t seem to be used here. I don’t think that ship NMHC emissions are needed – there is plenty of CO and methane around for ozone production in the NOx-limited regime. I may be wrong but a reference would be helpful.

This manuscript describes an attribution of summertime ozone over Europe to anthropogenic, natural and transport sources using a tagging approach in the MECO(n) model. The study is competently performed and described and the results are potentially useful for regional ozone mitigation efforts. However, the paper is largely descriptive rather than analytical, and does not provide much new insight into either the source contributions or their broader context, and this reduces its value greatly. While the paper should ultimately be appropriate for publication, it is not suitable in its current form. The results require a deeper level of interpretation to explain their consequences and why they matter, and the paper needs a clearer and more distinct message (preferably demonstrating some originality) that sets it apart from previous studies and makes it worthy of publication.

General Comments

The quantification of the different sources over Europe is interesting, and the focus on two important but contrasting regions of Europe is valuable. The focus on a range of metrics, including the upper tail of the ozone distribution and the responses of MDA8 are particularly valuable and are to be commended. However, the results are not fully exploited, and a greater degree of interpretation is required in the text about why the regions differ and what the consequences of this are. What are the implications of the attribution for mitigation efforts, are past source changes evident in recent ozone trends (given the attribution that has been derived), and what are the likely contributions to future ozone changes?

The treatment of uncertainty in the paper is weak. Contributions are typically given in the form of large ranges (whether this is spatial or temporal variance is not clear) for a single summer with a single emissions inventory. How sensitive are the results likely to be to changes in meteorology or to the reliability of the emissions used? Quantification (or at least estimation) of these uncertainties would give the reader greater confidence in the results presented. There is some comparison of results with previous tagging studies, although this is unsatisfying given the differing approaches used; how do the results compare with other estimates of source contributions based on observational estimates or non-tagging approaches?

The discussion somewhat undermines the application of tagging approaches for source attribution through highlighting the differing attribution to NOx or VOC sources dependent on the method used (e.g., comparison with the Butler approach). Under these circumstances, the perturbation approach appears more useful, as it provides a clear, unique ozone response to an applied change.

The conclusions need sharpening up. The current text summarizes results from each part of the analysis, but does not bring them together well to generate a clear and coherent message from the study. The summary of results needs to be cut back (by about half?) and some synthesis of findings should be added.

Specific Comments

The aims are clearly stated in the introduction, but I would like to see a stronger statement about why the study matters. It is valuable to understand how ozone may change in contrasting parts of Europe in future under different mitigation scenarios, but this isn't stated as a motivation. Why is source attribution important?

P.6:  Why is lightning neglected for the regional models? This may have a non-negligible impact on source attribution in summer, particularly in southern Europe, so some justification is needed.

Fig 1: Subtraction of the mean bias in panel (c) is potentially misleading and not physically meaningful; I suggest that the full bias is plotted, with an appropriate monochromatic (not dichromatic) color scale that emphasizes the key features of interest.

Page 9: What do we learn from the TOAR evaluation? There are substantial biases in the model simulations; while this is true for most models, the reader needs to understand why this is the case and how this is likely to impact the source attribution generated in the study. The same is also true for the HALO evaluation.

Figs 5 and 6 should be combined; so should Figs 8 and 9. Please consider combining Figs 4-9 into just two figures, either by species (Fig 4/5/6, Fig 7/8/9) or, better, by orientation (Fig 4/7, Fig 5/6/8/9). This would help the reader get a clearer overview of the comparisons.  Similarly Figs 10-13 should be combined into a single figure.  Using different colors (consistently) for NOy and O3 would aid the reader in interpretation.

Page 19, line 3: roughly what contribution do these sectors make?

Page 19, lines 9-11: show the key results first, before referring to the supplementary results.

Page 19, line 17: Why was Spain omitted? A short explanation is needed.

Fig 14: Please use the same color scale for the contribution plots (a-e) so that the reader can compare the contributions directly.

Figs 16-17: These would be clearer if colors were chosen to provide harmonization across a specific sector (land transport, non-traffic) or region (ROW,EU,NA,EA) using similar hue but contrasting saturation (for example). The same is true for Figs S10-S13 in the Supplement.

Page 21, line 14: Figs S21 and S22 are useful, but it would help the reader to quantify the differences in soil NOx and biogenic isoprene emissions in the text, e.g., by providing regional JJA average fluxes over each region.

Page 21, line 17: remind the reader what is included in the "others" category

Page 27, line 1: "high ozone concentrations" would be clearer in the title here (and in the text) than "large ozone values".

Figure 18: For direct comparison of these panels, it would be useful to have the same scale on the Y-axis (0-25 or 0-30 ppb). This is also true for the three lower panels in Figure 19, where the scale could be 0-40 ppb.

Page 31, lines 12-13: "various assumptions", "considerable uncertainties": please be specific here.  The discussion of uncertainties here is weak and qualitative, and a more thorough and quantitative assessment is needed here.

Page 33, lines 28-32: this paragraph undermines the study by casting doubt on the value of the results. A more quantitative approach to tackling uncertainties would allow these issues to be addressed, and would provide more confidence for the reader on the value of the results presented.

Page 34, lines 1-2: It would be clearer to say that the approach adopted here is not practical for use with a large number of regions.

Page 34, lines 32-33: combine this with previous paragraph (the topic is the same)

Typos and minor issues

There are a relatively large number of typographical errors that need to be cleaned up.

p.1,l.4:  remove "a"

p.1,l.7:  mode -> model

p.1,l.10: Worl -> World

p.1,l.12: data = observations?

p.1,l.19: in the range -> are in the range

p.1,l.20: emission sare -> emissions are

p.2,l.9:  citation format incorrect

p.2,l.24: grammar: "both" should be "these"

p.2,l.29: grammar: "rather" should be "more"

p.3,l.13: form -> from

p.3,l.14: up to -> at up to

p.3,l.28: asses -> assess

p.3,l.30: remove "a"

p.3,l.34: rest -> Rest

p.5,l.22: "interest" would be better as "concern" (the reader may be interested, but the distinction here isn't important for this study)

p.5,l.29: add "reactions of" before acetonitrile

p.5,l.31: part of -> included in

p.6, Table 1 caption: row -> column.

p.7,l.2-3: preprocessed in a way... (unclear, please rephrase)

p.7,l.7:  citation format (also on l.10)

p.10,l.7: sides -> sites

p.10,l.13: frequency of NOx -> frequency of high NOx

p.10,l.16: Means -> Mean

p.10,l.19: night -> the night  (is is -> is)

p.20,l.20: as in -> than in   (and on l.21)

p.21,l.1: rephrase sentence: the ozone contribution is about twice as large as in the Benelux region, 7 nmol/mol vs 3 nmol/mol.

p.21,l.15: add "these" before "different"

p.21,l.27: "similar to or greater than" (note that the transport times are also longer than the lifetime of ozone itself)

p.22,l.7: "contributes 27-37% of ozone."

p.27,l.5: whole Europe -> whole of Europe

p.27,l.8: analysis -> analyses

p.27,l.9: )) -> )

p.28,l.11: confirm -> confirms

p.31,l.26: artifical -> artificial; contribuions -> contributions

p.32,l.2: as the -> than the

p.33,l.19: seperate -> separate

p.33,l.21: oder -> or

p.33,l.27: corresponds to -> agrees with

p.34,l.9: it's -> its

p.34,l.13: the summer -> summer

p.34,l.18: by -> from

p.34,l.31: larger insulation -> higher insolation

p.35,l.3: hardly contribute -> contribute little

Page 37: EEA is an agency name, not a personal name, so should appear under "E" in the reference list.

Table S2: The higher elevation emissions from sector 9 have been placed under sector 8, please check and correct the table.