the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 29 Feb 2024
Dependence Models for Multi-Hazard-Events
Abstract. In recent years, the focus of research about natural hazards has turned from single-hazard studies to multi-hazard ones. While single hazards (like earthquakes, floods, droughts, etc.) have been extensively studied in the past and many quantitative statements about intensities and severities are available, quantitative studies about multi-hazards and dependencies are still rare. This paper introduces new statistical models for the dependencies of cat-event processes of different hazard types based on Poisson-type event processes. Moreover, the models are applied to data for several natural hazard events from the Danube area in Europe. The analysis should help to bridge the gap between the more conceptual contributions to this discussion by providing empirical evidence on interactions on a large-scale region.
- Preprint
(1202 KB) - Metadata XML
- BibTeX
- EndNote
Status: final response (author comments only)
-
RC1: 'Comment on nhess-2023-194', Anonymous Referee #1, 14 Mar 2024
Dear authors,
from what I can see, your model mainly determines interdependencies between types of hazards, which
seem to be quite obvious, while those which are less obvious are also not predicted by your model.
Also it would be helpful to see the location of events (at least for a sub-set/region) as their spatial relationships are also important.
Therefore, as such I do not see any particular interest of the study. But this could be related to the fact
that I am more used to work on physical relationships between hazard events rather than on pure 'mathematical-statistical' ones.
Anyway, the paper should also be able to raise interest among physical hazard modellers.
Finally, multi-hazard .. after reading - as relationships with earthquakes are barely indicated and those with landslides poorly,
the title should specify 'multiple climatic hazards'.
yours
reviewer H
Citation: https://doi.org/10.5194/nhess-2023-194-RC1 -
AC1: 'Reply on RC1', Georg Pflug, 10 Apr 2024
Dear colleague,
I do not agree on what you write. We estimate in a quantitative way the effect of one event to the intensity of another. This is "obvious" for you?
The model allows to make new quantitative analyses about these triggering and cascading effects in multi-hzard situations- this is 100% innovative. If you send me some data about such event series I would be more than happy to estimate the degree of intensity raise (or drop?) as a causal effect of the triggering event to events of some other type.
Unfortunately, you are hidden by anonymousity, I am not: You can read my homepage, find out my citations, my h-Index etc. But I am of course interested in a discussion about the merits of such a quantitative approach in a world where qualitative statements are frequent, but sound quantitative ones are rare, especially in this field.
Unfortunately, the lack of reliable data limits the extension of the scope to other dependency questions. But data are getting better and better.
With best regards
O.Prof. Dr. Georg Pflug
Head of Risk Management, University of Vienna, Austria
Citation: https://doi.org/10.5194/nhess-2023-194-AC1
-
AC1: 'Reply on RC1', Georg Pflug, 10 Apr 2024
-
RC2: 'Comment on nhess-2023-194', Anonymous Referee #2, 05 Apr 2024
Dear Prof. Pflug and co-authors,
Thank you for giving me a chance to review your manuscript "Dependence models for Multi-Hazard-Events".
I think the article addresses a very relevant problematic, the quantification of interrelations between hazards. Furthermore, the method presented offers promising perspectives, allowing the quantification of the temporal dependence, causality, and also incorporating a spatial component. However, this promising approach is undermined by the poor quality of the writing, figures and applications.
The article lacks supporting literature, missing key articles, which addressed quantification of hazard interrelations but also single-hazard events. This results in a lack of clarity while using concepts revolving around multi-risk (e.g, interrelation types, spatial and temporal scales, intensity assessment). The use of the EM-DAT database, despite its fame, does not appear to me as a relevant choice.
In summary, the topic of the article deserves attention, and it is clear that the method presented are relevant in a multi-hazard context. However, the quality of the writing is not up to the standard for scientific publication. Models that are not used are presented while the two triggering models are extremely briefly introduced. Results are poorly communicated (in text and in figures). Additionally, the authors repeat the same analysis other several pairs of hazards with very little background on the relevance of the pairs chosen. The discussion is extremely short, and does not really discuss the potential concrete usage of the method neither its limitations. I would suggest major revisions and an intensive work on the quality of the writing for the article to be published. I have provided a number of general comments and some specific to different sections.
General comments:
A. Writing and structure
The article is difficult to read. This partly because figures and table are all over the place (which may be due to formatting) and not properly mentioned and discussed in text! Another reason is the general flow on the article, with a lot of complex or unclear sentences. Probably rewriting the abstract could help to point towards the main objectives and findings of the study. The introduction is too weak and basically relies on one single article. I am not discussing here the excellence of Gill and Malamud (2014), but in 2024, there are quite a lot of other articles which address precisely multi-hazard interrelations and their quantification (Tilloy et al., 2022; Zscheischler et al., 2020; Lee et al., 2024; Simmonds et al., 2022; De Angeli et al., 2022; Terzi et al., 2019; Tilloy et al., 2019, 2020; Ridder et al., 2020). The method section is probably the most readable. However, it would deserve a sort of summary table at the end to clarify which model can be used for different cases of the following sections. Section 2.3 seems useless, as you are not using this model. I suggest displacing it to the discussion section as a perspective for further development. Section 3 should be merged with Section 2. Figures in Section 4 have too small text, unhelpful captions and are not well discussed. I do not understand the meaning of Section 5. Section 6 is excessively short and lacks references.
B. Unclear sentences, jargon and lack of context
Already in the abstract, terms are not well introduced (e.g., cat-event). In several places, the vocabulary used is very unclear or inaccurate, (intensity=timing?). The differences between hazard, risk, or disaster are ignored (l.10 to l.20). The paragraph between l.38 and l.49 is discussing a figure of Gill and Malamud (2014) as if we had it in front of our eyes, with limited added value to the introduction. In Section 2, it is not clear in which disciplines the presented methods have been used in the past. If you want to use EM-DAT, you will need to provide more information on exactly which data on which hazards you extracted, and also on the database itself (See Tilloy et al., 2022 for example).Overall I did not understand what you did in Section 4 (see detailed comments). Section 5 is just a repetition of the same workflow applied to hazard interrelations selected on unknown criteria. The dynamics and processes behind each interrelation are poorly described. Overall, significant improvements are needed on clarity.
C. Relevance of data and examples used
In the introduction you stat that you want to “give concrete examples using real data” (l.61). However, by using EM-DAT only, on such a large scale, and most importantly by using the count of events as a representation of intensity, you are not providing concrete examples. The region you are analysis is massive, composed of different climates, and with very different exposures (as you mention l.113), but you then decide to pool all events together, because there are too few of them. Indeed, EM-DAT is a global dataset and is known for being biased toward large-scale/high impact events and to have inaccurate recordings of location, duration and damages of events. I believe that other disaster datasets are available to increase your sample size and the accuracy of your hazard combinations (e.g., Paprotny et al., 2023, Claassen et al., 2023). I believe that with the spatial and temporal scales you are using, it is very likely that you find causality between unrelated hazards. I would suggest narrowing down the spatial and temporal scales considered (each interrelation has different dynamics and scales), increase the size of your event sample and focus on one or two interrelations.
Specific comments:
- Line 4 p1. What is a cat-event for you?
- Line 6 p1. The analysis aims to help to bridge…
- Line 11 p1. Do you mean continuous increase in damage? Use of insurance report could be a plus to back up the statement.
- Line 13 p1. Single events? What is it?
- Line 21-22 p1. Please provide references to support this statement
- Line 32 p2. Please avoid the word obvious.
- Line 33 p2. “In order to get an insight, which causal relationships are possible”.There is a problem with the sentence.
- Line 39 p2. please refer to your table 1.
- Line 46-49 p2. Is this from Gill and Malamud 2014? if yes please refer to the article at the end of the quote. See the nhess author guide.
- Line 52 p3. Intensity = timing? How?
- Line 74 p4. What is the @ for?
- I realize here that your line numbering is wrong
- Line 79-80 p4. Why do you mention this model if you do not wish to use it?
- Line 112-114 p6. Is there a more precise reason for the selection of the region? The region analysed deserved to be shown on a map here, with boundaries clearly shown (not like in the appendix, which is basically a google maps screenshot).
- Table 2 p7. What is the period in which the events occurred. What hazard is a storm exactly?
- Line 140 p7. There is plenty of literature on the interrelationship between drought and fire, please provide a few references (e.g., Sutanto et al., 2020; Joseph et al., 2019)
- Line 146 p9. “The increase in wildfires can (graphically) be explained by previous droughts or other influences”. I do not really see that.
- Figure 2 and Figure 3. What is the added value of these figures
- Line 157 p10. “Computing the intensity function via Maximum Likelihood gives the following estimate” which maximum likelihood?
- Line 162 p9.Please avoid using obvious…And the relationship between drought and fire is far from being obvious.
- Figure 6. I do not understand what it means and how it has been create, caption is insufficient.
- Line 199 p15. The process is not well explained, and you can find more academic references.
- Line 213 p16. The summary needs to give more information, this is just a tiny sentence.
- Line 216 p17. Lightning are not naturally triggered most of the time, please refer to literature for this as well.
- Line 244-245 p19: This is only a specific case of flood triggering, please see the literature on this topic (Berghuijs et al., 2019; Tarasova et al., 2019).
- Table 10 to 20 are repetitive and unclear.
- Figure 9. Please provide a decent map.
To conclude, it was difficult for me to understand the methods and processes developed in this paper. There is some work to do to consolidate the references, improve the writing, the figures and the general flow of the article. Introduction and conclusion have to be enhanced. Section 2 need more explanation on the models and their previous usages while Section 5 shall be removed or massively modified. I think the current way disaster data is used is not making justice to the potential of the method and can lead to misleading conclusions.
References:
Berghuijs, W. R., Harrigan, S., Molnar, P., Slater, L. J., and Kirchner, J. W.: The Relative Importance of Different Flood‐Generating Mechanisms Across Europe, Water Resources Research, 55, 4582–4593, https://doi.org/10.1029/2019WR024841, 2019.
De Angeli, S., Malamud, B. D., Rossi, L., Taylor, F. E., Trasforini, E., and Rudari, R.: A multi-hazard framework for spatial-temporal impact analysis, International Journal of Disaster Risk Reduction, 73, 102829, https://doi.org/10.1016/j.ijdrr.2022.102829, 2022.
Joseph, M. B., Rossi, M. W., Mietkiewicz, N. P., Mahood, A. L., Cattau, M. E., St. Denis, L. A., Nagy, R. C., Iglesias, V., Abatzoglou, J. T., and Balch, J. K.: Spatiotemporal prediction of wildfire size extremes with Bayesian finite sample maxima, Ecological Applications, 29, e01898, https://doi.org/10.1002/eap.1898, 2019.
Lee, R., White, C. J., Adnan, M. S. G., Douglas, J., Mahecha, M. D., O’Loughlin, F. E., Patelli, E., Ramos, A. M., Roberts, M. J., Martius, O., Tubaldi, E., Van Den Hurk, B., Ward, P. J., and Zscheischler, J.: Reclassifying historical disasters: From single to multi-hazards, Science of The Total Environment, 912, 169120, https://doi.org/10.1016/j.scitotenv.2023.169120, 2024.
Paprotny, D., Terefenko, P., and Śledziowski, J.: An improved database of flood impacts in Europe, 1870 - 2020: HANZE v2.1, Earth System Science Data Discussions, 1–37, https://doi.org/10.5194/essd-2023-321, 2023.
Ridder, N. N., Pitman, A. J., Westra, S., Ukkola, A., Hong, X. D., Bador, M., Hirsch, A. L., Evans, J. P., Di Luca, A., and Zscheischler, J.: Global hotspots for the occurrence of compound events, Nature Communications, 11, 1–10, https://doi.org/10.1038/s41467-020-19639-3, 2020.
Simmonds, R., White, C. J., Douglas, J., Sauter, C., and Brett, L.: A review of interacting natural hazards and cascading impacts in Scotland, 2022.
Sutanto, S. J., Vitolo, C., Di Napoli, C., D’Andrea, M., and Van Lanen, H. A. J.: Heatwaves, droughts, and fires: Exploring compound and cascading dry hazards at the pan-European scale, Environment International, 134, 105276, https://doi.org/10.1016/j.envint.2019.105276, 2020.
Tarasova, L., Merz, R., Kiss, A., Basso, S., Blöschl, G., Merz, B., Viglione, A., Plötner, S., Guse, B., Schumann, A., Fischer, S., Ahrens, B., Anwar, F., Bárdossy, A., Bühler, P., Haberlandt, U., Kreibich, H., Krug, A., Lun, D., Müller‐Thomy, H., Pidoto, R., Primo, C., Seidel, J., Vorogushyn, S., and Wietzke, L.: Causative classification of river flood events, WIREs Water, 6, https://doi.org/10.1002/wat2.1353, 2019.
Terzi, S., Torresan, S., Schneiderbauer, S., Critto, A., Zebisch, M., and Marcomini, A.: Multi-risk assessment in mountain regions: A review of modelling approaches for climate change adaptation, Journal of Environmental Management, 232, 759–771, https://doi.org/10.1016/j.jenvman.2018.11.100, 2019.
Tilloy, A., Malamud, B. D., Winter, H., and Joly-Laugel, A.: A review of quantification methodologies for multi-hazard interrelationships, Earth-Science Reviews, 196, 102881, https://doi.org/10.1016/j.earscirev.2019.102881, 2019.
Tilloy, A., Malamud, B. D., Winter, H., and Joly-Laugel, A.: Evaluating the efficacy of bivariate extreme modelling approaches for multi-hazard scenarios, Nat. Hazards Earth Syst. Sci., 20, 2091–2117, https://doi.org/10.5194/nhess-20-2091-2020, 2020.
Tilloy, A., Malamud, B. D., and Joly-Laugel, A.: A methodology for the spatiotemporal identification of compound hazards: wind and precipitation extremes in Great Britain (1979–2019), Earth System Dynamics, 13, 993–1020, https://doi.org/10.5194/esd-13-993-2022, 2022.
Zscheischler, J., Martius, O., Westra, S., Bevacqua, E., Raymond, C., Horton, R. M., van den Hurk, B., AghaKouchak, A., Jézéquel, A., Mahecha, M. D., Maraun, D., Ramos, A. M., Ridder, N. N., Thiery, W., and Vignotto, E.: A typology of compound weather and climate events, Nature Reviews Earth & Environment, 1, 333–347, https://doi.org/10.1038/s43017-020-0060-z, 2020.
Citation: https://doi.org/10.5194/nhess-2023-194-RC2 -
AC2: 'Reply on RC2', Georg Pflug, 10 Apr 2024
Dear colleague,
Thank you very much for your long reply. We will take these comments seriously. However, to criticize that we are not able to write scienific papers goes too far to my taste.
With a citation number of over 11000 and a h-index of 48, my scientific record is quite important. What may be the case is that this readership group wants a different style and we will try to meet these requirements. Our approach is novel, Maybe we can cite a few more papers which have done some work in a similar direction. But I am reluctant to cite work which is not closely related. Believe me, as a (past) editor of about 7 scientific journals, I know that some referees are pushing to cite their work. I do not at all say that this applies to you, since I do not know you (you are anonymous, but I am transparent to you).
Anyway, thank you again for the work you have done for us.
With best regards,
o.Prof. Dr. Georg Pflug, Head of Risk Management, University fo Vienna, Austria
Citation: https://doi.org/10.5194/nhess-2023-194-AC2
Viewed
| HTML | XML | Total | BibTeX | EndNote | |
|---|---|---|---|---|---|
| 352 | 81 | 27 | 460 | 14 | 17 |
- HTML: 352
- PDF: 81
- XML: 27
- Total: 460
- BibTeX: 14
- EndNote: 17
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1