General Comments

The paper uses the ERA5 reanalysis dataset to obtain 40 years worth of Lagrangian meteorological fields for extratropical cyclones crossing into Northern Europe. Since the paper uses a new reanalysis data set to examine extratropical cyclones and their precursors over Northern Europe, it is novel. I find the composite results interesting and useful not only for specialists in the field, but also for a more general audience.

In the paper, composite analysis is used to characterize and understand Northern European extratropical cyclones, while ensemble sensitivity analysis is used to find correlations between precursor fields and the selected response functions: Minimum Mean Sea-Level Pressure (MSLP) and 10m wind gusts. Four precursor fields were selected: 850 hPa potential temperature, total column water vapor, and 300 hPa wind speed and potential vorticity. Regressions were computed for 1, 2, and 3 days before extrema in MSLP and 10m wind gusts.

The results are presented in a series of 18 figures that are legible, well-labeled, and generally of a high quality. Some notable results in the context of cyclones affecting Northern Europe include; no significant trends in cyclones were found over the 1980-2019 period; summer cyclones are less intense than winter ones; 1995-2005 lull in cyclone activity, which can be attributed to anomalously lower numbers of windstorms over the period; windstorms are more common in the winter; overall, numbers of cyclones do not show an appreciable seasonal cycle over Northern Europe; windstorms have distinct genesis regions compared to non-windstorm cyclones; windstorms preferentially genesis over the sea while non-windstorms that affect Northern Europe genesis over Northern Europe; summer tracks are further poleward than winter tracks, and summer cyclones live longer.

The paper reads well, and it also establishes the scientific relevance of the study well. The paper also embeds its results in the established literature. I can recommend this paper for publication after some major revisions are made.

Major Comments

While the nature and characteristics of extratropical cyclones are of interest to practitioners and specialists in the field, I found the ensemble sensitivity part of the paper quite weak. The paper makes causal inferences when the method as is does not support causal inferences. The causal inferences from the Dacre et al., 2019 study were valid because the authors isolated the precursor fields from the effects of the cyclone itself before the regression. In the present study, all that can be said is that stronger baroclinic and moisture conditions in a developing cyclone are associated with stronger maximum cyclones. When put like that, the results of the ensemble sensitivity analysis don’t seem as exciting. Thus all causal inferences (like L328 or L339-L342) should be removed prior to publication. Changing "leads to" or "causes" to "is associated with" may address this problem, but a broader rethink of how the method is used could lead to a far superior manuscript.

Minor Comments

Figure 2 a,c,e: Make the origin non-zero so that the bars aren't unusally long and so that the variation among bars, not length of bars, is emphasized.

Figure 3: Label the y-axis in the figure.

L209: A citation is required for the Mann-Kendall reference.

L210-211: I would end the discussion at no statistically significant trend was found because this line reads as if the experimenter is beholden to a particular finding.

Is there any explanation for the drop in windstorms between the years 1995 and 2000?

Section 6. It may be helpful to indicate the direction of cyclone propagation on figures 7 and 8 (left to right?).

L322: This causal statement does not follow from the results presented.

L325: This causal statement does not follow from the results presented.

L326: This causal statement does not follow from the results presented. 

L329: This causal statement does not follow from the results presented.

L337: This causal statement does not follow from the results presented.

L357: This causal statement (and the others that follow) does not follow from the results presented.

L430: less -> fewer

L456: Sentence unclear

This paper investigates extratropical cyclones over Northern Europe, which are typically less studied in the literature, but have big impacts on the region. The cyclones are split into two groups – those that are classed as windstorms (i.e. having strong winds within the region) and those that are non-windstorms. The authors analyse the seasonality of these storms, their structure, lifecycle, and correlations with precursor fields (using ensemble sensitivity analysis). An interesting result is that the minimum MSLP of the windstorms are more strongly correlated with precursor fields than the maximum 10-m wind gusts. However, the strong link between the minimum MSLP and the wind gusts gives extra potential predictability.

The paper is very well written and enjoyable to read, with some minor grammatical issues (detailed below).

My two main criticisms are:

1. The way in which the “windstorms” are selected gives (as the authors mention) mostly the extratropical cyclones that occur over the sea, due to decreased friction. It would be valuable to know whether storms that have the biggest impacts (i.e. the strongest winds over land) have the same precursors. This would certainly be more relevant when considering how these features might change in the future. However, if the authors feel that this is outside the scope of the paper, a stronger justification of the selection of the storms would be good.
2. The ensemble sensitivity analysis is based on correlations between fields, and does not necessarily imply causation. Many of the phrases “leads to” or “causes” (e.g. line 185), should be changed to reflect this.

Minor comments

1. Line 15: “than” -> “as”.
2. Line 30: “and” -> “with”.
3. Line 44: “to” -> “of”.
4. Line 61: “of” -> “for”.
5. Lines 51-66: The inclusion of some references would be useful here.
6. Lines 79-80: This gives the impression that this study will also consider climate change. Perhaps it could be made clearer that this is a possibility for future study.
7. Introduction: For some of the introduction, the terms “extratropical cyclones” and “windstorms” seem to be interchangeable until line 81. Then the difference is not made clear before the research questions. I suggest including a brief definition of extratropical cyclones versus windstorms.
8. Line 90: “How the spatial…” -> “How does the spatial…”.
9. Line 111: Many other studies using TRACK use a truncation of T42 before identifying extratropical cyclones. Is this because you are using MSLP rather than vorticity? It would be useful to mention this difference.
10. Line 116: “with a 6…” -> “within a 6…”.
11. Line 128: “differ from…” -> “go from…”.
12. Line 138: “which” -> “whose”.
13. Figure 2: I’m not sure you need both the left and right panels. Perhaps on the left panels you could just add a dashed line to indicate the mean.
14. Line 251: “less” -> “fewer”.
15. Lines 313, 317, 379 and others: The use of the term “offset times” is confused in places. Line 317, for example, I would refer to the time of minimum MSLP as the “central” time (or similar). Then the offset is the difference from that central time.
16. Line 335 and throughout: “similar than” should be “similar to” everywhere it appears.
17. Line 347: Why not just say “minimum MSLP” instead of “maximum intensity in terms of MSLP”?
18. Line 360: I think the sentence could end after “cold season”.
19. Line 408: Perhaps the first sentence could be turned around. E.g. “The 10-m wind gusts are weakly sensitive to the 300-hPa PV”.