the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 25 Nov 2024
Brief Communication: Investigating trends in European hailstorm damage using CMIP6-DAMIP climate models
Abstract. Warming seas around Europe have been driving recent upward trends in hailstorm severity. Therefore, learning more about changes to sea temperatures can give insights into hail climate too. Here, we use the DAMIP (Detection and Attribution Model Intercomparison Project) set of climate model experiments to explore how external forcings have modified Mediterranean temperatures. Climate models indicate external forcings caused most of the multidecadal changes in modern times, with anthropogenic aerosols explaining the cool period from about 1900 to the late 1970s, and greenhouse gas increases mainly responsible for the rapid 0.5 K/decade warming of the Mediterranean since then. Current trends in anthropogenic forcing are expected to continue warming seas which suggests European hailstorm risk will keep rising.
- Preprint
(879 KB) - Metadata XML
- BibTeX
- EndNote
Status: open (until 17 Jan 2025)
-
CC1: 'Comment on nhess-2024-210', Cameron Rye, 12 Dec 2024
reply
I have a comment about the trends in insured losses reported in the paper. Further details around the methodology used to adjust historical losses to present-day values is required. For example, it appears that GDV losses for Germany automobiles have been adjusted for inventory (number of cars) and inflation. However, the standard method for adjusting insured losses is to account for three factors: inflation, wealth and inventory. I suspect wealth has not been taken into account in this instance. The $ value of cars has increased significantly over time, particularly with the introduction of electric vehicles. As a result, I suspect the average pay-out for a auto hail claim will have significantly increased over time. How can the authors be sure that the trend in Figure 1a is due to increasing risk and not simply reflecting the increasing value of vehicles? The same point applies to other datasets, for example it appears that the France property data have only been adjust for the cost of reconstruction (FFB index), and not other factors.
Secondly, I suggest that the focus of the paper could be expanded to consider other perils. The med ocean is also important for e.g. medicines, cut-off lows, Vb cyclones. Hail risk is complex, and not entirely dependent on the med ocean (i.e. there is not necessarily a direct 1-to-1 link). In my view it would be a more rounded article if the focus is the med, and then talk about how this is important for a number of perils.
Citation: https://doi.org/10.5194/nhess-2024-210-CC1 -
RC1: 'Comment on nhess-2024-210', Anonymous Referee #1, 23 Dec 2024
reply
This manuscript aims to assess potential trends in hailstorm damage across Europe using climate modelling. The study aligns well with the scope of the journal, addressing a topic that is both timely and underexplored in previous studies. However, in my opinion the manuscript suffers from a notable methodological limitation: the main approach and results only loosely correspond to the stated objective of the study.
Specifically, the authors utilize modelling outcomes from the Detection and Attribution Model Intercomparison Project (DAMIP) to demonstrate that warming sea surface temperatures in the Mediterranean are predominantly attributable to anthropogenic influences. While it is justifiable that rising sea temperatures correlate with increased hailstorm activity, this connection is neither directly analyzed in the manuscript nor explored through more appropriate proxy variables, such as wet-bulb temperature or measures of convective instability. Sections 2 and 3 (Methodology and Results) bear no direct relationship to hailstorm trends, apart from the general hypothesis that continued warming will lead to an increase in hail-related damage.
While the current methodology requires refinement to align with the study’s objectives, the authors have demonstrated a significant effort in leveraging DAMIP outcomes to explore anthropogenic influences on sea surface temperatures. Thus, despite my overall concerns, I am not suggesting rejection. Instead, I encourage the authors to undertake a major revision to address the above-mentioned issues, acknowledging that this will involve significant challenges. One possible path forward would be to incorporate more targeted hail-related variables into the methodology. Alternatively, the authors could expand their literature review on hail damage, employing more rigorous analytical techniques and relocating this analysis from the Introduction to the Methodology and Results sections.
Specific & Minor Comments:
1) Lines 18 – 19: Please verify whether this is the appropriate format for referencing a website according to the journal’s guidelines.
2) Some abbreviations appear unnecessary or unconventional, such as “R21” for Raupack et al. (2021), and “the Med” for the Mediterranean Sea.
3) Figure 1. It is unclear whether this figure represents the authors’ original analysis or a direct visualization of data from other studies. If it is original, it should be moved to the Results section. If it is based on external data, the appropriate references must be included in the figure caption.
4) Lines 74 – 77. The above comment regarding clarification of data sources applies here as well.
5) Lines 106 – 107. Please specify which monthly-mean near-surface temperature diagnostics were used in this study to ensure transparency.
6) Consider expanding Table 1 to include more detailed information about the modeling simulations, such as specific parameters, assumptions, or configurations.
7) Lines 152 – 153. The information about the red tick marks should also be included in the figure caption for clarity.Citation: https://doi.org/10.5194/nhess-2024-210-RC1 -
RC2: 'Comment on nhess-2024-210', Anonymous Referee #2, 14 Jan 2025
reply
The manuscript discusses how the risk for damage from hail has changed in the past decades in Europe. It addresses a problem that is not widely studied and is well in the scope of the journal. To showcase that damage from hail has increased, the authors present results from hazard-based studies as well as studies and catalogues looking at insured losses. To study the causes behind this trend, the authors use DAMIP models from CMIP6 to study the effect of different forcing mechanisms on northern Mediterranean sea temperatures which have been shown to be linked with hail-producing storms. They find that sea temperature in the Mediterranean has had a mostly linear response to forcings. They show that from 1850 until 1970 the cooling effect from anthropogenic aerosols dominated the overall trend of Mediterranean temperatures, after which warming effect due to increased greenhouse gases took over and has caused rapid warming since, leading to an increased hail risk in Europe.
General comments:
Overall, the manuscript is well written and structured clearly .The literature review is conducted thoroughly and the synthesis in the introduction nicely shows how damage from hail has increased in recent decades in parts of Europe. The DAMIP models are cleverly used to attribute different trends in temperatures to different forcings. While this analysis is valid, it could be extended to better substantiate its link to hailstorm damage or risk. I have two main points to highlight:
- While the link between certain variables and the occurrence of hailstorms is established in previous studies, the analysis presented in the manuscript using DAMIP models does not convincingly address the trends in hailstorm damage directly or through these links. To better justify the claims made in the manuscript, the association of surface temperatures (sea or near-surface) to the variables mentioned in the introduction, such as CAPE and near-surface moisture or wet-bulb temperature, should be explained more in detail or demonstrated in the analysis. This would highlight more in accordance with previous literature how hailstorm damage trends are linked with trends of other variables.
-
The extent of the analysed area is quite small to claim that conclusions can be drawn for the whole of Europe. The mechanism with which conditions in northern Mediterranean affect hailstorm occurrence elsewhere in Europe should be highlighted more than through the mention of a reference. In addition or alternatively, some statistics from previous studies of how often a hailstorm (elsewhere) in Europe is influenced by high dew points in the Mediterranean could be included. Moreover, the claim that hailstorm damage can be assessed for whole Europe in this way, is valid only if a significant majority of damage from hail occurs in the mentioned areas in central and western Europe. If this is the case, it should be somehow shown in the manuscript. Otherwise, I suggest the title (and elsewhere in the text) to be more focused on a specific area of Europe.
Specific comments/questions:
- Line 13: Last sentence in the abstract is not factually incorrect but could be rephrased to make it not sound like it is the trends that are warming seas, rather than the forcing. For example, “Given the current trends in anthropogenic forcing, seas are expected to continue warming...”
- Line 54: During which time period does the increase of 1 to 1.5% p.a. occur?
-
Line 59: Does Figure 1a show annual losses for automobiles for combined wind and hail perils or only hail, which comprise the vast majority?
-
Line 64: I am unfortunately not very familiar with insurance terminology. How does loss cost relate to total loss? Does it increase with increasing losses?
- Line 79 and 87: Previous studies have found rising low-level humidity and wet-bulb temperature to be linked with an increased hail risk. Is the thermodynamic effect here that SST and near-surface air increase as much, which through increased evaporation retains relative humidity values? Thus wet-bulb temperature rises along with specific humidity?
-
Line 108: Although a link to CMIP documentation is provided, Table 1 could include more details about the models used, for example model resolution.
-
Line 111: What is the effect of using near-surface temperature from the model but sea surface temperature from observations? Are SSTs prescribed in the DAMIP models? Is the effect inconsequential because only anomalies are considered?
-
Line 113: What is the purpose of using data from different time periods for observed and modelled values? Is it to highlight the small anomalies at the start of the period and on the other hand the increased anomaly projected to continue after the end of modelled data? I realise that methodologically it has little influence since the anomalies are calculated for the common period.
-
Line 120: What are the boundaries of this region based on? Are the modelled near-surface temperatures considered only over the shaded region in Figure 2, i.e. over land? This could be specified in the figure caption or text.
-
Line 139: What is the cause of the reduced amplitude of the multidecadal oscillation? Is due to smoothing from using a multi-model and ensemble mean?
-
Line 143: Does this refer to statistical significance in the amplitude of the multidecadal oscillation in Hist? If so, based on which test?
-
Line 169: Are the more references to substantiate the claim about anthropogenic aerosols driving the multidecadal changes such as the profound peak in European windstorm damages? I would be interested to see the given reference but it is yet to be published.
-
Line 183: Given the relatively good agreement between models, can you comment on why especially in the most disagreeing models (4 and 6) it looks like the deviation from the multi-model mean increases with time in Figure 3b? Is this an effect of the multi-model mean diverging from the climatological mean, i.e. larger forcing introducing more spread between the models?
-
Line 199: Although the overall trend between 1850 and 1970 is a cooling one, to my ear the way this sentence is phrased makes it sound like there was continuous cooling throughout the period.
Technical comments/typos:
- Line 45: I believe here and elsewhere “Radler” in citation should be “Rädler”.
-
Line 142: Symbol for standard deviation does not show properly.
-
Line 158: The word “then” is repeated in the sentence.
Citation: https://doi.org/10.5194/nhess-2024-210-RC2 -
RC3: 'Comment on nhess-2024-210', Anonymous Referee #3, 16 Jan 2025
reply
Please see attached review.
Viewed
| HTML | XML | Total | BibTeX | EndNote | |
|---|---|---|---|---|---|
| 309 | 92 | 8 | 409 | 5 | 5 |
- HTML: 309
- PDF: 92
- XML: 8
- Total: 409
- BibTeX: 5
- EndNote: 5
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1