Preprints
https://doi.org/10.5194/egusphere-2024-4157
© Author(s) 2025. This work is distributed under
the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
https://doi.org/10.5194/egusphere-2024-4157
© Author(s) 2025. This work is distributed under
the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
17 Jan 2025
| 17 Jan 2025
A novel European windstorm dataset based on ERA5 reanalysis from 1940 to present
Abstract. In this work, we present and preliminarily evaluate a novel dataset of European windstorms associated with extratropical cyclones (ETCs) based on the whole ERA5 reanalysis period (1940–present). This dataset is produced within the Copernicus Climate Change Service (C3S) Enhanced Operational Windstorm Service (EWS), to promote a knowledge-based assessment of the nature and temporal evolution of European windstorms associated with ETC. Such a dataset is primarily thought to provide high-quality, standardized data on windstorms that support various industries, particularly insurance and risk management, by offering insights into the intensity, density spatial patterns, and, if coupled downstream, with vulnerability and exposure information, the impact of windstorms. EWS includes two datasets: windstorm tracks, based on two tracking algorithms (TRACK and TempestExtremes), and windstorm footprints, produced considering both original-resolution ERA5 variables and statistically downscaled ERA5 variables, with a target grid at 1 km resolution. A preliminary analysis of the datasets shows increasing number of cold-semester windstorms and the associated footprint wind gusts magnitude over a portion of the European territory. The choice of the tracking algorithm is shown to be an important factor in the decision-making process, as it results in non-negligible uncertainties in main windstorm statistics.
How to cite. Sangelantoni, L., Tibaldi, S., Cavicchia, L., Scoccimarro, E., Vidale, P. L., Hodges, K., Mavel, V., Almansi, M., Cagnazzo, C., and Almond, S.: A novel European windstorm dataset based on ERA5 reanalysis from 1940 to present, EGUsphere [preprint], https://doi.org/10.5194/egusphere-2024-4157, 2025.
Publisher's note: Copernicus Publications remains neutral with regard to jurisdictional claims made in the text, published maps, institutional affiliations, or any other geographical representation in this preprint. The responsibility to include appropriate place names lies with the authors.
Download & links
- Preprint
(8254 KB) - Metadata XML
-
Supplement
(141 KB) - BibTeX
- EndNote
Status: closed
Comment types: AC – author | RC – referee | CC – community | EC – editor | CEC – chief editor
| : Report abuse
-
RC1: 'Comment on egusphere-2024-4157', Mukesh Kumar, 07 Feb 2025
- AC2: 'Reply on RC1', Lorenzo Sangelantoni, 21 Jun 2025
-
RC2: 'Comment on egusphere-2024-4157', Anonymous Referee #2, 12 May 2025
The authors present an overview of the latest C3S windstorm information service. While I appreciate that this service is useful and presents a fantastic resource for the users highlighted in the manuscript, I have many issues with the presentation of the manuscript in its current form and do not believe it can be published as is. I believe the writing is quite lazy and overly long in places, with necessary details being left out. Below i detail my major comments, and also minor points. Once these are addressed I think the authors should re-submit to the journal for another consideration.
Major comments
- My biggest comment involves section 4 on the trend quantification. I believe that these results are highly inrepresentative of changes to windstorms across Europe, and while the authors present significant trends, consideration should be made as to whether to even include this analysis. It has been shown that trends from long reanalysis products are highly questionable (e.g. Wohland et al., 2019; Bloomfield et al., 2018; https://agupubs.onlinelibrary.wiley.com/doi/full/10.1029/2018JD030083; https://iopscience.iop.org/article/10.1088/1748-9326/aad5c5), so this poses as to whether the windstorms in the early part of the ERA5 catalogue are representative of true climate. Furthermore, variability in European windstorms is highly non-linear and dominated by peaks in the 90s (see Cusack, 2023, fig 9; https://nhess.copernicus.org/articles/23/2841/2023/) and therefore the authors application of a linear trend is not appropriate. I believe the authors should not use this data to quantify any potential trends, but instead present this as a resource for analysis in a manner the end-user deems appropriate. This section should be (in my opinion) removed for the resubmission.
- My other major comment is around the statistical downscaling used to generate the high resolution footprints. It appears from the equation at lines 208-209 that the only information at the 1km-scale is the local terrain. Therefore, all the other terms are there as re-scalings and then the information to downscale is just an orography scaling. The reason i take issue for this is in situations when you may have something like a downslope windstorm. You would expect the strongest gusts to be at the lowest elevation, yet following this statistical approach the strongest gusts would be at the highest elevations? This is surely unintuitive. What validation of this downscaling has gone on, and why was this chosen over a dynamical downscaling approach that happened in the previous C3S windstorm product WISC?
- The choice of the 990 hPa track point threshold is one i question. My opinion is that often frontal wind gusts may occur when the cyclone core pressure is >990 hPa. What is the authors justification to this, and what impact does this choice have on the footprints that they are creating?
Minor comments
- L20 - "an increasing"
- Use season instead of semester throughout
- L20 "associated footprint wind gusts magnitude" doesn't read well and should be rephrased
- L21 "portion of the European territory" should be more specific
- L29 - "During recent decades". THis is not a new phenomena and this makes it seem as if it is. Rephrase
- L31 - reference for the €5 billion per year
- L48 - also consider referencing the recent CMIP6 study (https://rmets.onlinelibrary.wiley.com/doi/10.1002/qj.4849)
- L58 - you use "innovation" here, but other windstorm assessments have been performed in the past. You mention XWS, but also there should be some recognition of WISC as this was the predecessor of this project and provided all this information but instead with ERA-Interim. Perhaps use "resource" instead of "innovation".
- L75-77 - there are >>2 tracking schemes, so surely just using these two does not represent the full uncertainty from tracking? See recent paper by Flaounas et al. (https://wcd.copernicus.org/articles/4/639/2023/) as to the tracking uncertainty. If the authors are only using these two, some justification as to why these two is required.
- L125/126 - you state earlier that you require winds within a 3 degree radius, but only mention 5 degrees here. This is either an additional step that needs mentioning, or needs rectifying here.
- L144-146 - so these tracks must essentially pass over Europe. Please make this clearer here as this is quite a clunky section and i found it hard to follow these simple regional criteria.
- L153-154 - this sentence doesn't make the most grammatical sense, please rephrase more simply to just mention TRACK and TE in (a) and (b) respectively.
- Figure 1 - this would be better as a track density plot and differences. It's very hard to see any differences in the top panels with this density of lines.
- L164 - "Moreover, TRACK is quite consistent with itself", what do you mean by this? Of course it is consistent with itself!
- Figure 2 - please add a colorbar for the shading. Also, the black contours are not consistent across panels, so it makes it very difficult to compare the values, especially when looking at the panels on the top row.
- You need to state in your methods what months are you "cold" and "warm" seasons
- L172 - why are you using a 4x4 degree binning of data for your statistics? This is incredibly coarse and also much coarser than the native ERA5 data. It seems unnecessary and should be done on at least 1x1 degree surely as ERA5 is 0.25x0.25 degree
- Figure 4 - it would be good to see a difference plot between these two footprints. As the tracks are different, but the gusts the same i question quite why we need to see both, but it would be good to see how this affects the resultant footprints
- L235 - please mention dates of these storms
- L237 - 'traces' should be 'tracks'?
- Figure 7 - it feels unnecessary to have 9 panels here, can you not plot all the tracks for a storm on one panel but with different colours?
- L247 - how do you propose someone would combine both algorithms?
- Figures 8/9, i recommend somehow combining figures 8 and 9, or just using figure 9, as this is in some ways duplicate information and means that the manuscript is unnecessarily long.
- Page 14 - this is a very long paragraph - consider reducing and breaking up the text throughout the manuscript
- L279/280 - how much better though? Is it possible to compare the mean bias of the downscaled ERA5 to the original ERA5 relative to the obs for selected (or all) storms. This will demonstrate the benefit of the downscaling.
- L345-350 - this SSI formulation is somewhat different to what is commonly used (see Leckebusch et al., 2008). As SSI is very dependent on the formulation i recommend using the most commonly adopted formulation (e.g. Karremann et al., 2014; https://iopscience.iop.org/article/10.1088/1748-9326/9/12/124016).
Citation: https://doi.org/10.5194/egusphere-2024-4157-RC2 - AC1: 'Reply on RC2', Lorenzo Sangelantoni, 21 Jun 2025
Status: closed
Comment types: AC – author | RC – referee | CC – community | EC – editor | CEC – chief editor
| : Report abuse
-
RC1: 'Comment on egusphere-2024-4157', Mukesh Kumar, 07 Feb 2025
- AC2: 'Reply on RC1', Lorenzo Sangelantoni, 21 Jun 2025
-
RC2: 'Comment on egusphere-2024-4157', Anonymous Referee #2, 12 May 2025
The authors present an overview of the latest C3S windstorm information service. While I appreciate that this service is useful and presents a fantastic resource for the users highlighted in the manuscript, I have many issues with the presentation of the manuscript in its current form and do not believe it can be published as is. I believe the writing is quite lazy and overly long in places, with necessary details being left out. Below i detail my major comments, and also minor points. Once these are addressed I think the authors should re-submit to the journal for another consideration.
Major comments
- My biggest comment involves section 4 on the trend quantification. I believe that these results are highly inrepresentative of changes to windstorms across Europe, and while the authors present significant trends, consideration should be made as to whether to even include this analysis. It has been shown that trends from long reanalysis products are highly questionable (e.g. Wohland et al., 2019; Bloomfield et al., 2018; https://agupubs.onlinelibrary.wiley.com/doi/full/10.1029/2018JD030083; https://iopscience.iop.org/article/10.1088/1748-9326/aad5c5), so this poses as to whether the windstorms in the early part of the ERA5 catalogue are representative of true climate. Furthermore, variability in European windstorms is highly non-linear and dominated by peaks in the 90s (see Cusack, 2023, fig 9; https://nhess.copernicus.org/articles/23/2841/2023/) and therefore the authors application of a linear trend is not appropriate. I believe the authors should not use this data to quantify any potential trends, but instead present this as a resource for analysis in a manner the end-user deems appropriate. This section should be (in my opinion) removed for the resubmission.
- My other major comment is around the statistical downscaling used to generate the high resolution footprints. It appears from the equation at lines 208-209 that the only information at the 1km-scale is the local terrain. Therefore, all the other terms are there as re-scalings and then the information to downscale is just an orography scaling. The reason i take issue for this is in situations when you may have something like a downslope windstorm. You would expect the strongest gusts to be at the lowest elevation, yet following this statistical approach the strongest gusts would be at the highest elevations? This is surely unintuitive. What validation of this downscaling has gone on, and why was this chosen over a dynamical downscaling approach that happened in the previous C3S windstorm product WISC?
- The choice of the 990 hPa track point threshold is one i question. My opinion is that often frontal wind gusts may occur when the cyclone core pressure is >990 hPa. What is the authors justification to this, and what impact does this choice have on the footprints that they are creating?
Minor comments
- L20 - "an increasing"
- Use season instead of semester throughout
- L20 "associated footprint wind gusts magnitude" doesn't read well and should be rephrased
- L21 "portion of the European territory" should be more specific
- L29 - "During recent decades". THis is not a new phenomena and this makes it seem as if it is. Rephrase
- L31 - reference for the €5 billion per year
- L48 - also consider referencing the recent CMIP6 study (https://rmets.onlinelibrary.wiley.com/doi/10.1002/qj.4849)
- L58 - you use "innovation" here, but other windstorm assessments have been performed in the past. You mention XWS, but also there should be some recognition of WISC as this was the predecessor of this project and provided all this information but instead with ERA-Interim. Perhaps use "resource" instead of "innovation".
- L75-77 - there are >>2 tracking schemes, so surely just using these two does not represent the full uncertainty from tracking? See recent paper by Flaounas et al. (https://wcd.copernicus.org/articles/4/639/2023/) as to the tracking uncertainty. If the authors are only using these two, some justification as to why these two is required.
- L125/126 - you state earlier that you require winds within a 3 degree radius, but only mention 5 degrees here. This is either an additional step that needs mentioning, or needs rectifying here.
- L144-146 - so these tracks must essentially pass over Europe. Please make this clearer here as this is quite a clunky section and i found it hard to follow these simple regional criteria.
- L153-154 - this sentence doesn't make the most grammatical sense, please rephrase more simply to just mention TRACK and TE in (a) and (b) respectively.
- Figure 1 - this would be better as a track density plot and differences. It's very hard to see any differences in the top panels with this density of lines.
- L164 - "Moreover, TRACK is quite consistent with itself", what do you mean by this? Of course it is consistent with itself!
- Figure 2 - please add a colorbar for the shading. Also, the black contours are not consistent across panels, so it makes it very difficult to compare the values, especially when looking at the panels on the top row.
- You need to state in your methods what months are you "cold" and "warm" seasons
- L172 - why are you using a 4x4 degree binning of data for your statistics? This is incredibly coarse and also much coarser than the native ERA5 data. It seems unnecessary and should be done on at least 1x1 degree surely as ERA5 is 0.25x0.25 degree
- Figure 4 - it would be good to see a difference plot between these two footprints. As the tracks are different, but the gusts the same i question quite why we need to see both, but it would be good to see how this affects the resultant footprints
- L235 - please mention dates of these storms
- L237 - 'traces' should be 'tracks'?
- Figure 7 - it feels unnecessary to have 9 panels here, can you not plot all the tracks for a storm on one panel but with different colours?
- L247 - how do you propose someone would combine both algorithms?
- Figures 8/9, i recommend somehow combining figures 8 and 9, or just using figure 9, as this is in some ways duplicate information and means that the manuscript is unnecessarily long.
- Page 14 - this is a very long paragraph - consider reducing and breaking up the text throughout the manuscript
- L279/280 - how much better though? Is it possible to compare the mean bias of the downscaled ERA5 to the original ERA5 relative to the obs for selected (or all) storms. This will demonstrate the benefit of the downscaling.
- L345-350 - this SSI formulation is somewhat different to what is commonly used (see Leckebusch et al., 2008). As SSI is very dependent on the formulation i recommend using the most commonly adopted formulation (e.g. Karremann et al., 2014; https://iopscience.iop.org/article/10.1088/1748-9326/9/12/124016).
Citation: https://doi.org/10.5194/egusphere-2024-4157-RC2 - AC1: 'Reply on RC2', Lorenzo Sangelantoni, 21 Jun 2025
Viewed
Total article views: 507 (including HTML, PDF, and XML)
Cumulative views and downloads
(calculated since 17 Jan 2025)
| HTML | XML | Total | Supplement | BibTeX | EndNote | |
|---|---|---|---|---|---|---|
| 372 | 116 | 19 | 507 | 36 | 12 | 28 |
- HTML: 372
- PDF: 116
- XML: 19
- Total: 507
- Supplement: 36
- BibTeX: 12
- EndNote: 28
Viewed (geographical distribution)
Total article views: 496 (including HTML, PDF, and XML)
Thereof 496 with geography defined
and 0 with unknown origin.
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1
1
Latest update: 15 Jul 2025
Lorenzo Sangelantoni
CORRESPONDING AUTHOR
CMCC Foundation - Euro-Mediterranean Center on Climate Change, Bologna, Italy
Stefano Tibaldi
CMCC Foundation - Euro-Mediterranean Center on Climate Change, Bologna, Italy
Leone Cavicchia
CMCC Foundation - Euro-Mediterranean Center on Climate Change, Bologna, Italy
Enrico Scoccimarro
CMCC Foundation - Euro-Mediterranean Center on Climate Change, Bologna, Italy
Pier Luigi Vidale
National Centre for Atmospheric Science, Dept. of Meteorology, University of Reading, Reading, UK
Kevin Hodges
National Centre for Atmospheric Science, Dept. of Meteorology, University of Reading, Reading, UK
Vivien Mavel
B‐Open Solutions srl, Rome, Italy
Mattia Almansi
B‐Open Solutions srl, Rome, Italy
Chiara Cagnazzo
ECMWF, Bonn, Germany
Samuel Almond
ECMWF, Bonn, Germany
Short summary
We introduce a new dataset of European windstorms linked to extratropical cyclones, spanning whole ERA5 reanalysis period (1940–present). Developed under Copernicus Climate Change Service, the dataset provides standardized, high-quality information on windstorm tracks and footprints for industries like insurance and risk management. Preliminary findings show an increase in cold-season windstorms and their impacts in parts of Europe. Tracking methods contribute to uncertainties in key statistics.
We introduce a new dataset of European windstorms linked to extratropical cyclones, spanning...