the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 02 Jan 2023
Thunderstorm Types in Europe
Deborah Morgenstern
Isabell Stucke
Georg J. Mayr
Achim Zeileis
Thorsten Simon
Abstract. Lightning characteristics in all seasons are investigated across Europe because it is observed that lightning strikes to tall infrastructure have no or only a weak annual cycle whereas lightning in general has a pronounced annual cycle. Using cluster analysis on ERA5 reanalysis data and EUCLID lightning data, two major thunderstorm types are found: Wind-field thunderstorms characterized by increased wind speeds, strong updrafts, and high shear occurring mainly in winter. And mass-field thunderstorms characterized by increased mass-field variables such as large CAPE values, high dewpoint temperatures, and elevated isotherm heights, occurring mostly in summer. Several sub-types of these two main thunderstorm types exist. Using principal component analysis, four topographically distinct regions in Europe are identified that share similar thunderstorm characteristics: The mediterranean, alpine-central, continental, and coastal regions, respectively. Based on these results it is possible to differentiate lightning in different seasons without a static threshold or a seasonal criterion.
-
Notice on discussion status
The requested preprint has a corresponding peer-reviewed final revised paper. You are encouraged to refer to the final revised version.
-
Preprint
(1321 KB)
-
The requested preprint has a corresponding peer-reviewed final revised paper. You are encouraged to refer to the final revised version.
- Preprint
(1321 KB) - BibTeX
- EndNote
- Final revised paper
Journal article(s) based on this preprint
Deborah Morgenstern et al.
Interactive discussion
Status: closed
-
RC1: 'Comment on egusphere-2022-1453', Anonymous Referee #1, 18 Jan 2023
The overall quality of this paper is good. The research uses objective statistical methods (PCA and K-means clustering) to simplify atmospheric variables, thunderstorm types and spatially similar regions to good effect. This enhances the understanding of thunderstorm types present in different regions of Europe and explores their seasonal variations. In particular, this method allows the identification of less frequently occurring thunderstorm types which pose a risk to particular types of infrastructures. The thunderstorm types identified by this method had not yet been defined in this way across a wide and variable geographic area, so this identification, therefore, complements and enhances the current understanding of thunderstorm behaviour in Europe. The methodology to identify thunderstorm types may now be used at regional resolution and in other areas to continue to add to knowledge.
The Abstract – consider explaining more clearly the rationale for undertaking this research. The strikes to tall infrastructure are mentioned in the first sentence as the main reason for undertaking the research. Tall infrastructure is then not mentioned very often in the main body of the paper. It is suggested that either the authors include more discussion of the relevance of their results to tall infrastructure so it is more prominent in the mind of the reader. Alternatively, the abstract could be reworded so that the emphasis is taken off tall infrastructure and it could be mentioned merely as an example of the potential benefits / end users of this research.
The discussion makes good points and relates this research back to the literature well. However, some greater depth would enhance this part of the paper nicely. Examples of points which could be expanded:
-The identification of the thunderstorm types could be related back to the results of previous studies with similar domains (coastal, continental) and seasonal variations in lightning occurence – the Baltic and the UK are mentioned in particular and perhaps the authors could extrapolate from their results which thunderstorm type and by extension atmospheric conditions, might produce a risk from lightning in the regions / seasons identified in these countries (where there might be wind farms located). This should then fit in nicely with the author’s statement that these already identified regions compliment their research by bringing together all this information in one place.
-The authors mention that this research is limited by the resolution of atmospheric data. Do the authors have any suggestions to tackle this for further research? Are there any potential higher-resolution features (mountain ranges) within the identified domains / regions that may influence the thunderstorm types as a result?
The text of the paper does contain some grammatical errors which are noted where identified by the reviewer. However, there may be others which have not been identified, so the authors should check again the text for grammar using relevant grammar assistance technology if necessary. The authors may also wish to check whether it is the convention in their field/this journal to use American-English or English-English as there are some instances of “summarize” instead of “summarise” or “fall” instead of “autumn” which the journal may or may not have a preference for.
L11 – Perhaps provide examples of meteorological settings
L12 – Suggest replacing “Plenty” with “Numerous”. If “Plenty” is to remain suggest amendment to “Plenty of”.
L34 – Suggest replacing “at land and advected to the sea” with ”over land and advected out to sea” or “onshore and advected out to sea”
L35 – Perhaps instead of “it endures longer” which suggests a singular long-lasting lightning flash, replace with “where lightning activity endures longer”
L35 – Suggest replace “water surface is” with “sea surface temperatures are”
L45/46 - Suggest inclusion of study location where the correlation of SST and lightning strokes were discovered.
L46 – Suggest replace “to use the” with “the use of”
L56 – Do you mean to ask the question of how thunderstorm characteristics vary by meteorological conditions across Europe or how thunderstorm occurrence varies by meteorological characteristics across Europe… Please make the wording of this question a little clearer.
L61 – suggest change “second one (Sect. 3)” to “second (Sect. 3)” The word “one” is unnecessary and removal may improve the flow of the sentence.
L64 – Suggest change “The found thunderstorm types” to “These thunderstorm types”
L87/88 – Suggest review of sentence grammar / rewording here to make clearer. The use of the word “As” to start the sentence does not seem quite right.
L100 – “France and Belgium, being a less homogeneous but very representative domain.” Representative of what? Please consider making this clearer for the reader.
L123 – “In the following, only one sample is discussed, as all repetitions led to qualitatively the same results.
“ Please consider providing greater detail on how the repetitions were compared and found to be the same.
L157 – consider changing “are considered as ‘baselines’ there.” To “are considered as ‘baselines’ “ The word “there” does not seem to be necessary.
L160 – “Meteorological similar domains gather close to one another” perhaps this should read something like “Domains with similar meteorological characteristics are represented in close proximity to one another within this diagram”
Figure 3 – This is a nice way to represent these results, however it is a little busy which makes interpretation more difficult. If possible, some improvements should be considered to simplify the plot for the benefit of the reader. Suggest experimenting with; removing or making smaller the point symbols, removing the domain letter labels to the right of the chart which can extend the x-axis (these can be moved to figure caption instead), increasing the size of the y-axis.
L188 – Suggest change from “With this the spatial different thunderstorm conditions in Europe” to “These spatially different thunderstorm conditions in Europe”
Figure 4 is really good, shows the results very clearly for the reader to then understand the main text discussion. Suggest making the CP, noMF labels a bit more prominent so the distinction between the types is immediately obvious to the reader.
L208 – Style wise it is not usual to include we, or I in a research paper and the tone is a little informal. Suggest changing “Now we dive deeper into the characteristics of the found thunderstorm types” to something like “More detailed analysis of the thunderstorm type characteristics is undertaken”
Figure 5 – Similar to figure 3 in places it is difficult to interpret due to being quite busy. Wind field noMF in particular has a rather faint colour which becomes obscured by others. Please consider again removing or reducing the size of the point symbols or changing the colour scheme to make the graph clearer to read.
241 – “Now that the thunderstorm types are found” tone is a bit informal. Suggest rewording this sentence.
L246 – Where it states “there is often a bigger one” consider replacing the word “one” with something clearer or rewording the sentence, perhaps “there is often a dominant type”
Figure 6 – include y-axis labels or y-axis explainer in the figure caption. To make it clear whether this refers to absolute counts, is a percentage or scaled like earlier plots.
Figure 7 – J, K and L domains with their barplots are a little cramped, suggest investigating offsetting the barplots to edge of the figure and indicating their domain using arrows to see if this improves things a little.
L277 – Consider replacing “little thunderstorms” with “fewer thunderstorms”
L300 – It is not ideal to start a sentence with the word “And”. Suggest rewording this.
L313 – Consider changing “lightning is climatological unlikely” to “lightning is climatologically unlikely”
L317 – Please remove full stop from: Europe?”.
L336 – Consider changing “origins” to “originates”
Please consider the comments made regarding the figures in the main text of the paper with respect to the figures in the appendix.
Citation: https://doi.org/10.5194/egusphere-2022-1453-RC1 -
RC2: 'Comment on egusphere-2022-1453', Anonymous Referee #2, 09 Feb 2023
According to the title and the abstract, the present work “Thunderstorm Types in Europe” by Deborah Morgenstern, Isabell Stucke, Georg J. Mayr, Achim Zeileis, and Thorsten Simon shall provide a possibility to distinguish thunderstorms independent of the season and fixed thresholds (like CAPE). The work is mainly motivated by the fact that thunderstorms also occur in winter and for high infrastructure even no significant seasonal correlation can be identified. For this purpose, model parameters of ERA5 reanalysis data and EUCLID lightning data are statistically analyzed and it has been possible to describe significant differences between at least two situations in which thunderstorms occur. The results suggest that other atmospheric processes are important for the development of winter thunderstorms compared to thunderstorms in summer. The results allow to compare thunderstorm types for different regions independent of the season.
The results presented are relevant on this point, showing that obviously different atmospheric processes enable thunderstorms. However, there are also important limitations. For example, the analyzed data are only coarsely resolved and accordingly do not allow to resolve the thunderstorms. At best, they reflect the large-scale environment and parameterization of thunderstorms. Moreover, the parameter selection is also arbitrary (line 79 and following is not well-founded enough in this context) and accordingly not necessarily suitable to characterize thunderstorm situations. The results are accordingly only of limited use for further research, since, contrary to what is suggested in the abstract and summary, by no means only fixed thresholds or seasonal parameters are used for thunderstorm forecasting, especially not when using high-resolution weather models.
This is accompanied by the most important points of criticism of the submitted manuscript. Both the title and the abstract are misleading, since they give the impression that different types of thunderstorms are actually treated, and moreover, even the differences in lightning physics are discussed. However, to differentiate between thunderstorm types, one would examine smaller scale characteristics, such as subdivision into mesoscale systems or isolated thunderstorm cells. Also, the vulnerability of high infrastructure as a motivation and the dissimilarity of winter and summer thunderstorms in terms of their lightning characteristics is not discussed further, except that it is mentioned prominently and is dismissed as "beyond the scope of this paper" in the course. This short-coming is already noticeable in the introduction, when there is a strong break in the content between general thunderstorm climatology and the forced-seeming motivation for this paper in lines 40-44. In this respect, more depth may be added to the text.
In this sense, the paper can disappoint the reader. Consideration should at least be given to changing the title, abstract, and also the summary accordingly to match the results of the paper. These are essentially described above and can be understood as an indication that thunderstorms are not exclusively caused by strong seasonal heating of the subsurface (solid or liquid), but also independent dynamic factors, called "wind-field" in this paper, play a role.
Despite these short-comings, the overall quality of the paper is good.
Citation: https://doi.org/10.5194/egusphere-2022-1453-RC2 -
EC1: 'Comment on egusphere-2022-1453', Johannes Dahl, 17 Feb 2023
Dear authors,
I would like to add a comment: The nomenclature of "wind-field" thunderstorms vs. "mass field" thunderstorms is potentially misleading I think. To some readers, it might suggest that wind-field storms are somehow driven by the wind field, rather than by CAPE (because you contrast them with "mass-field" thunderstorms). Really, what you are saying is that in some regimes, CAPE is a poor predictor for whether or not lightning occurs. But this appears to be a result of CAPE being generally small in these regimes (and hence, it has only small variability). Maybe you could add some clarification in the revised manuscript.Citation: https://doi.org/10.5194/egusphere-2022-1453-EC1 - AC1: 'Comment on egusphere-2022-1453', Deborah Morgenstern, 03 Mar 2023
Interactive discussion
Status: closed
-
RC1: 'Comment on egusphere-2022-1453', Anonymous Referee #1, 18 Jan 2023
The overall quality of this paper is good. The research uses objective statistical methods (PCA and K-means clustering) to simplify atmospheric variables, thunderstorm types and spatially similar regions to good effect. This enhances the understanding of thunderstorm types present in different regions of Europe and explores their seasonal variations. In particular, this method allows the identification of less frequently occurring thunderstorm types which pose a risk to particular types of infrastructures. The thunderstorm types identified by this method had not yet been defined in this way across a wide and variable geographic area, so this identification, therefore, complements and enhances the current understanding of thunderstorm behaviour in Europe. The methodology to identify thunderstorm types may now be used at regional resolution and in other areas to continue to add to knowledge.
The Abstract – consider explaining more clearly the rationale for undertaking this research. The strikes to tall infrastructure are mentioned in the first sentence as the main reason for undertaking the research. Tall infrastructure is then not mentioned very often in the main body of the paper. It is suggested that either the authors include more discussion of the relevance of their results to tall infrastructure so it is more prominent in the mind of the reader. Alternatively, the abstract could be reworded so that the emphasis is taken off tall infrastructure and it could be mentioned merely as an example of the potential benefits / end users of this research.
The discussion makes good points and relates this research back to the literature well. However, some greater depth would enhance this part of the paper nicely. Examples of points which could be expanded:
-The identification of the thunderstorm types could be related back to the results of previous studies with similar domains (coastal, continental) and seasonal variations in lightning occurence – the Baltic and the UK are mentioned in particular and perhaps the authors could extrapolate from their results which thunderstorm type and by extension atmospheric conditions, might produce a risk from lightning in the regions / seasons identified in these countries (where there might be wind farms located). This should then fit in nicely with the author’s statement that these already identified regions compliment their research by bringing together all this information in one place.
-The authors mention that this research is limited by the resolution of atmospheric data. Do the authors have any suggestions to tackle this for further research? Are there any potential higher-resolution features (mountain ranges) within the identified domains / regions that may influence the thunderstorm types as a result?
The text of the paper does contain some grammatical errors which are noted where identified by the reviewer. However, there may be others which have not been identified, so the authors should check again the text for grammar using relevant grammar assistance technology if necessary. The authors may also wish to check whether it is the convention in their field/this journal to use American-English or English-English as there are some instances of “summarize” instead of “summarise” or “fall” instead of “autumn” which the journal may or may not have a preference for.
L11 – Perhaps provide examples of meteorological settings
L12 – Suggest replacing “Plenty” with “Numerous”. If “Plenty” is to remain suggest amendment to “Plenty of”.
L34 – Suggest replacing “at land and advected to the sea” with ”over land and advected out to sea” or “onshore and advected out to sea”
L35 – Perhaps instead of “it endures longer” which suggests a singular long-lasting lightning flash, replace with “where lightning activity endures longer”
L35 – Suggest replace “water surface is” with “sea surface temperatures are”
L45/46 - Suggest inclusion of study location where the correlation of SST and lightning strokes were discovered.
L46 – Suggest replace “to use the” with “the use of”
L56 – Do you mean to ask the question of how thunderstorm characteristics vary by meteorological conditions across Europe or how thunderstorm occurrence varies by meteorological characteristics across Europe… Please make the wording of this question a little clearer.
L61 – suggest change “second one (Sect. 3)” to “second (Sect. 3)” The word “one” is unnecessary and removal may improve the flow of the sentence.
L64 – Suggest change “The found thunderstorm types” to “These thunderstorm types”
L87/88 – Suggest review of sentence grammar / rewording here to make clearer. The use of the word “As” to start the sentence does not seem quite right.
L100 – “France and Belgium, being a less homogeneous but very representative domain.” Representative of what? Please consider making this clearer for the reader.
L123 – “In the following, only one sample is discussed, as all repetitions led to qualitatively the same results.
“ Please consider providing greater detail on how the repetitions were compared and found to be the same.
L157 – consider changing “are considered as ‘baselines’ there.” To “are considered as ‘baselines’ “ The word “there” does not seem to be necessary.
L160 – “Meteorological similar domains gather close to one another” perhaps this should read something like “Domains with similar meteorological characteristics are represented in close proximity to one another within this diagram”
Figure 3 – This is a nice way to represent these results, however it is a little busy which makes interpretation more difficult. If possible, some improvements should be considered to simplify the plot for the benefit of the reader. Suggest experimenting with; removing or making smaller the point symbols, removing the domain letter labels to the right of the chart which can extend the x-axis (these can be moved to figure caption instead), increasing the size of the y-axis.
L188 – Suggest change from “With this the spatial different thunderstorm conditions in Europe” to “These spatially different thunderstorm conditions in Europe”
Figure 4 is really good, shows the results very clearly for the reader to then understand the main text discussion. Suggest making the CP, noMF labels a bit more prominent so the distinction between the types is immediately obvious to the reader.
L208 – Style wise it is not usual to include we, or I in a research paper and the tone is a little informal. Suggest changing “Now we dive deeper into the characteristics of the found thunderstorm types” to something like “More detailed analysis of the thunderstorm type characteristics is undertaken”
Figure 5 – Similar to figure 3 in places it is difficult to interpret due to being quite busy. Wind field noMF in particular has a rather faint colour which becomes obscured by others. Please consider again removing or reducing the size of the point symbols or changing the colour scheme to make the graph clearer to read.
241 – “Now that the thunderstorm types are found” tone is a bit informal. Suggest rewording this sentence.
L246 – Where it states “there is often a bigger one” consider replacing the word “one” with something clearer or rewording the sentence, perhaps “there is often a dominant type”
Figure 6 – include y-axis labels or y-axis explainer in the figure caption. To make it clear whether this refers to absolute counts, is a percentage or scaled like earlier plots.
Figure 7 – J, K and L domains with their barplots are a little cramped, suggest investigating offsetting the barplots to edge of the figure and indicating their domain using arrows to see if this improves things a little.
L277 – Consider replacing “little thunderstorms” with “fewer thunderstorms”
L300 – It is not ideal to start a sentence with the word “And”. Suggest rewording this.
L313 – Consider changing “lightning is climatological unlikely” to “lightning is climatologically unlikely”
L317 – Please remove full stop from: Europe?”.
L336 – Consider changing “origins” to “originates”
Please consider the comments made regarding the figures in the main text of the paper with respect to the figures in the appendix.
Citation: https://doi.org/10.5194/egusphere-2022-1453-RC1 -
RC2: 'Comment on egusphere-2022-1453', Anonymous Referee #2, 09 Feb 2023
According to the title and the abstract, the present work “Thunderstorm Types in Europe” by Deborah Morgenstern, Isabell Stucke, Georg J. Mayr, Achim Zeileis, and Thorsten Simon shall provide a possibility to distinguish thunderstorms independent of the season and fixed thresholds (like CAPE). The work is mainly motivated by the fact that thunderstorms also occur in winter and for high infrastructure even no significant seasonal correlation can be identified. For this purpose, model parameters of ERA5 reanalysis data and EUCLID lightning data are statistically analyzed and it has been possible to describe significant differences between at least two situations in which thunderstorms occur. The results suggest that other atmospheric processes are important for the development of winter thunderstorms compared to thunderstorms in summer. The results allow to compare thunderstorm types for different regions independent of the season.
The results presented are relevant on this point, showing that obviously different atmospheric processes enable thunderstorms. However, there are also important limitations. For example, the analyzed data are only coarsely resolved and accordingly do not allow to resolve the thunderstorms. At best, they reflect the large-scale environment and parameterization of thunderstorms. Moreover, the parameter selection is also arbitrary (line 79 and following is not well-founded enough in this context) and accordingly not necessarily suitable to characterize thunderstorm situations. The results are accordingly only of limited use for further research, since, contrary to what is suggested in the abstract and summary, by no means only fixed thresholds or seasonal parameters are used for thunderstorm forecasting, especially not when using high-resolution weather models.
This is accompanied by the most important points of criticism of the submitted manuscript. Both the title and the abstract are misleading, since they give the impression that different types of thunderstorms are actually treated, and moreover, even the differences in lightning physics are discussed. However, to differentiate between thunderstorm types, one would examine smaller scale characteristics, such as subdivision into mesoscale systems or isolated thunderstorm cells. Also, the vulnerability of high infrastructure as a motivation and the dissimilarity of winter and summer thunderstorms in terms of their lightning characteristics is not discussed further, except that it is mentioned prominently and is dismissed as "beyond the scope of this paper" in the course. This short-coming is already noticeable in the introduction, when there is a strong break in the content between general thunderstorm climatology and the forced-seeming motivation for this paper in lines 40-44. In this respect, more depth may be added to the text.
In this sense, the paper can disappoint the reader. Consideration should at least be given to changing the title, abstract, and also the summary accordingly to match the results of the paper. These are essentially described above and can be understood as an indication that thunderstorms are not exclusively caused by strong seasonal heating of the subsurface (solid or liquid), but also independent dynamic factors, called "wind-field" in this paper, play a role.
Despite these short-comings, the overall quality of the paper is good.
Citation: https://doi.org/10.5194/egusphere-2022-1453-RC2 -
EC1: 'Comment on egusphere-2022-1453', Johannes Dahl, 17 Feb 2023
Dear authors,
I would like to add a comment: The nomenclature of "wind-field" thunderstorms vs. "mass field" thunderstorms is potentially misleading I think. To some readers, it might suggest that wind-field storms are somehow driven by the wind field, rather than by CAPE (because you contrast them with "mass-field" thunderstorms). Really, what you are saying is that in some regimes, CAPE is a poor predictor for whether or not lightning occurs. But this appears to be a result of CAPE being generally small in these regimes (and hence, it has only small variability). Maybe you could add some clarification in the revised manuscript.Citation: https://doi.org/10.5194/egusphere-2022-1453-EC1 - AC1: 'Comment on egusphere-2022-1453', Deborah Morgenstern, 03 Mar 2023
Peer review completion
Journal article(s) based on this preprint
Deborah Morgenstern et al.
Model code and software
Supplementary material: Thunderstorm Types in Europe Deborah Morgenstern, Isabell Stucke, Thorsten Simon, Georg J. Mayr, Achim Zeileis https://doi.org/10.5281/zenodo.7436761
Deborah Morgenstern et al.
Viewed
| HTML | XML | Total | BibTeX | EndNote | |
|---|---|---|---|---|---|
| 368 | 162 | 14 | 544 | 5 | 3 |
- HTML: 368
- PDF: 162
- XML: 14
- Total: 544
- BibTeX: 5
- EndNote: 3
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1
The requested preprint has a corresponding peer-reviewed final revised paper. You are encouraged to refer to the final revised version.
- Preprint
(1321 KB) - Metadata XML