the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 24 Mar 2025
North Atlantic seasonal climate variability significantly modulates extreme winter Euro-Atlantic extratropical cyclone hazards
Abstract. North Atlantic extratropical cyclones (ETCs) cause significant financial losses in Europe, particularly in winter. Previous work has shown seasonal relationships between ETC hazards and modes of North Atlantic atmospheric variability, including the North Atlantic Oscillation (NAO; PC1) and East Atlantic Pattern (EAP; PC2). This study examines the relationship between the most extreme ETC hazards experienced at a given location in a winter season with the NAO and EAP, focusing on the winter maximum 10 metre wind gust and coastal wave swell height and the maximum daily total precipitation. We examine compound effects where PC1 or PC2 have signals in multiple hazard types at the same location. Positive PC1 exhibits coincident increases in winter maximum wind gust and wave swell hazards around most coastal regions in northern Europe. Positive PC2 exhibits coincident increases in winter maximum wind gust and daily total precipitation hazards over land areas in southern UK, Portugal and Spain, with an additional compound effect from increased wave swell near southern UK, northern France and Spain coasts. We also consider compound effects where PC1 and PC2 show coincident signals in the same hazard at a given location, potentially indicating an elevated hazard likelihood when circulation anomalies project onto both modes concurrently. PC1 and PC2 have coincident signals for wind gusts in southern Ireland, southern UK, Portugal and Scandinavian coast. For wave swell height, PC1 and PC2 have coincident signals around the Scandinavian, southern UK and Ireland and Northern Portugal coasts. This study shows that large-scale modes of seasonal North Atlantic climate variability modulate the exposure to extreme ETC hazards in many parts of Europe. The results have the potential to be combined with skilful seasonal climate forecasts of PC1 and PC2 to inform the insurance sector.
- Preprint
(3813 KB) - Metadata XML
-
Supplement
(3550 KB) - BibTeX
- EndNote
Status: open (until 07 May 2025)
-
RC1: 'Comment on egusphere-2025-1131', Mika Rantanen, 16 Apr 2025
reply
Review of NHESSD manuscript egusphere-2025-1131 “North Atlantic seasonal climate variability significantly modulates extreme winter Euro-Atlantic extratropical cyclone hazards” by Maycock et al.
This paper investigates the relationship between large-scale climate modes and ETC-related hazards in Europe. The authors track cyclones, calculate their hazard footprints and then use linear regression to find how much the ETC-related extremes change with respect to PC1 (North Atlantic Oscillation) and PC2 (East Atlantic pattern). The key result is that PC1 or PC2 alone exhibit increases in hazards in relatively different geographical areas (i.e. PC1 in the east and north of the UK, but PC2 mainly in the west and south). In addition, there are areas which exhibit signals for several hazards at the same time, and areas where both PC1 and PC2 affect simultaneously.
I like the research idea and I think this is definitely something which is worth publishing in NHESS. The used datasets are appropriate for conducting this kind of study. I also liked that negative results (SDI) were mentioned.
However, I had some concerns related to how the key results are presented. I think this could have been done in a more explicit/quantitative way (see comment 1). In addition, I’m afraid that the daily precipitation associated with the ETCs might be overestimated (see comment 2). I hope that the authors could address these concerns before the publication of this study.
Major comments:
- The presentation of the results. I don't know how to really formulate this, but I got the feeling that presenting the main results only in a rather qualitative way with Figs. 3-7 leaves the results a bit incomplete. Now you go through the regions rather subjectively (i.e. increase here, decrease there and so on). Could this be done in a more quantitative way, for example selecting beforehand relevant regions (domains) from Europe, e.g. countries or more wider regions such as Scandinavia, Western Europe, etc. And then calculate the regional statistics of how NAO and EAP affect the ETC hazards. These could be presented for example with boxplots which compare the climatology and then a unit increase of PC1/PC2. For example, the climatological daily ETC-precipitation in Scandinavia is this, but when NAO is positive, it’s this, and so on. This would provide more quantitative information on the regional distribution of the results. I hope you get the idea!
- Precipitation footprints. I was not entirely convinced by the way the daily precipitation is assigned to the ETCs. Most importantly, you consider daily (24h) precipitation, but the passage of an ETC can last much less than 24 hours and can occur during two consecutive calendar days. For instance, if an area of ETC (i.e. the 10° circle) passes over a particular grid cell in 12 hours, say from 18 UTC to 6 UTC. What is the period used for calculating the daily precipitation that is attributed to the ETC? Is it a moving 24 hour window, i.e. the previous 24 hours after the passage, or some fixed time interval, like 00-00UTC? This can cause problems especially at the outer edges of the 10° circles, which are only briefly affected by the passage of the distant ETC, but the precipitation is still counted from a 24-hour duration, resulting in exaggerated ETC-related precipitation values. Or am I missing something here?
Other comments:
Section 2. It seems that the whole analysis is restricted to the NH winter but it would be good to mention the months (Dec-Feb) explicitly in the Methods section. Currently, this is mentioned only in Section 2.3 but I guess it applies to the whole analysis. Which leads me to the 2nd question. Why only DJF? At least in Fennoscandia, November is often a very active month in terms of windstorm hazards.
Section 2.3. North Atlantic modes of variability. I think section 2.3 is a bit incomplete. It lacks justification why you chose the domain which you chose (90W-40E, 20-80N). Furthermore, I think this area is often called the Euro-Atlantic sector as it extends up to 40E, but you talk about the North Atlantic sector which is slightly misleading, given the area. Also, some studies (e.g. https://rmets.onlinelibrary.wiley.com/doi/10.1002/qj.3341) call the 2nd EOF as Scandinavian or European blocking. Is the East Atlantic pattern the same as Scandinavian blocking? If not, it might be reasonable to mention this in the text. It might help if the regression/correlation patterns against MSLP are shown, for example in supplementary material.
Section 4. Please consider writing something about the limitations of your analysis. For example, the linear model does not naturally explain all the variability in ETC-related hazards. What other factors are there which add the variability? How could you improve your work in the future?
L46 and thereafter. You often cite Degenhardt et al. 2022 but there is only Degenhardt et al. 2023 and 2024 in the reference list.
L136. This should be Section 2.4
L155: these metrics? which metrics?
L157: Here you mention that linear regression performs poorly if the data is non-linear and contains lots of zeros. But isn’t that the case for ETC-hazards too, for those regions which infrequently see ETCs during DJF months? So how do you deal with those regions that are far from storm tracks, and might not see ETCs every winter? Are there those regions at all?
L182: are shown
L200: show a reduction? How can you see this as the colour bars in Fig. 3 only show positive values? I see that the absolute anomalies in Fig. S3 also have negative values, but I don't understand why the percentage anomalies in the ΔPC1 and ΔPC2 maps in Fig. 3 only show positive changes?
L287. Previous work. Here it would be good to cite the actual previous work.
Fig. 5 and 6: the titles show 1981-2010, should it be 2020? And why does Fig. 7 have 2021 in its title?
Citation: https://doi.org/10.5194/egusphere-2025-1131-RC1 -
RC2: 'Comment on egusphere-2025-1131', Anonymous Referee #2, 28 Apr 2025
reply
Summary: I think there is potential for this work to be developed into a very useful contribution to our understanding of the connections between weather hazards and variability on larger (both spatial and temporal) scales. To move towards that potential, a couple of important elements in the current manuscript need to be addressed, as follows:
Main Comment
Reading through the analysis, the question I kept asking myself is: what is gained in the study by including ETCs? For Figs. 3 – 6, it is not clear to me how the ETC footprint step is useful. For instance, you could look at the relationship between the extremes for each of those hazards (regardless of ETC association) and the NAO and EAP. You might even have more data, which would make the analysis less noisy. So the question is: what does the utilization of ETCs add to the story? Does the link to the ETCs give us longer timescales in predictability? Or do they provide something else? This needs to be explained.Related to this: In Section 3.2 (Line 211), based on the title, my hope was that you would focus on the compound events caused by individual ETCs, but you actually go in the opposite direction (based on my interpretation of what you write on Line 216). You might want to consider including an analysis that is focused on describing the ETCs that cause the compound hazards.
Minor CommentsUnfortunately, I think the title is too general of a statement. For instance, when I read that title, I think: I have high confidence in that statement, without doing any research. So, I suggest you come up with a new title that demonstrates more of the knowledge gained by this research.
Here is a possible replacement (mostly taken from text that you wrote in the introduction): “The Observed relationships between ETC hazards and the NAO and EAP on Seasonal Timescales.”
Line 19: Throughout the abstract and the text, I think it will be more intuitive for the reader if you replace “PC1” with NAO and “PC2” with EAP. I appreciate that PC1 and PC2 are more precise, but they offer less connotation with physics.
Line 46: In this section, you may also want to refer to Pinto et al. 2009 (DOI 10.1007/s00382-008-0396-4)
Line 51-52: You write: “A positive EAP phase is associated with an increase in cumulative winter storm severity in the UK, which is weaker than for the NAO …”
What is weaker, the storm severity or the association? Please re-write the sentence to clarify.Line 82 - 86: You have provided some discussion on the potential bias in the reanalysis. Please expand a bit more on this. Please add some discussion specifically about the wave swell. And for winds and precipitation, some examples that you could reference:
Ramon J, Lledó L, Torralba V, Soret A, Doblas-Reyes FJ. What global reanalysis best represents near-surface winds?. Q J R Meteorol Soc. 2019; 145: 3236–3251. https://doi.org/10.1002/qj.3616
Chen, T.-C., Collet, F., & Di Luca, A. (2024). Evaluation of ERA5 precipitation and 10-m wind speed associated with extratropical cyclones using station data over North America. International Journal of Climatology, 44(3), 729–747. https://doi.org/10.1002/joc.8339Line 107: What is “Rx1 day metric”? I tried searching within your doc and did not find Rx anywhere else. Sorry if I missed it.
Line 194: You write: “The lack of significant relationship with PC1 could be a result of the relatively noisy data, since at each gridpoint we are taking the wettest day in the winter associated with any ETC and regressing this against the seasonal NAO.” Have you tried interpolating the seasonal NAO to daily? Or is there a reason you have to use the seasonal NAO? If so, remind the reader of that reasoning. Also, have you tried relaxing the definition of the “extreme” precipitation to include more data (i.e., instead of using the maximum, use the top N percentile)?
Line 213-214: You write:
“Here we consider the overlap of the shaded areas for each hazard in Fig 3 separately for PC1 and PC2 to determine the relative exposure to multiple ETC hazards at a given location.”
Please explain more clearly the method for capturing the overlap and explain what you interpret this overlap to mean.Line 237: You write:
“… we next consider the overlap between the PC1 and PC2 patterns for each variable separately, …” How do you do this analysis. Explain it clearly.Citation: https://doi.org/10.5194/egusphere-2025-1131-RC2 -
RC3: 'Comment on egusphere-2025-1131', Lisa Degenhardt, 29 Apr 2025
reply
Attached you can find my review to the manuscript.
Viewed
| HTML | XML | Total | Supplement | BibTeX | EndNote | |
|---|---|---|---|---|---|---|
| 118 | 16 | 6 | 140 | 13 | 7 | 8 |
- HTML: 118
- PDF: 16
- XML: 6
- Total: 140
- Supplement: 13
- BibTeX: 7
- EndNote: 8
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1