Estimations of statistical dependence as joint return period modulator of compound events – Part 1: Storm surge and wave height

Petroliagkis, Thomas I.

doi:https://doi.org/10.5194/nhess-18-1937-2018

Articles | Volume 18, issue 7

https://doi.org/10.5194/nhess-18-1937-2018

© Author(s) 2018. This work is distributed under
the Creative Commons Attribution 3.0 License.

https://doi.org/10.5194/nhess-18-1937-2018

© Author(s) 2018. This work is distributed under
the Creative Commons Attribution 3.0 License.

Articles | Volume 18, issue 7

Research article

|

12 Jul 2018

Research article |

| 12 Jul 2018

Estimations of statistical dependence as joint return period modulator of compound events – Part 1: Storm surge and wave height

Thomas I. Petroliagkis

Download

Final revised paper (published on 12 Jul 2018)
Supplement to the final revised paper
Preprint (discussion started on 24 Jul 2017)

Interactive discussion

Status: closed

AC: Author comment | RC: Referee comment | SC: Short comment | EC: Editor comment

- Printer-friendly version

- Supplement

RC1: 'reviewer comment on manuscript nhess-2017-177', Anonymous Referee #1, 21 Aug 2017
- AC1: 'AC1 Response to Anonymous Referee 1', Thomas Petroliagkis, 19 Nov 2017
RC2: 'Interactive comment on „Estimations of statistical dependence as joint return period modulator of compound events. Part I: Storm surge and wave height” by Thomas I. Petroliagkis', Anonymous Referee #2, 09 Oct 2017
- AC2: 'AC2 Response to Anonymous Referee 2', Thomas Petroliagkis, 19 Nov 2017

Peer-review completion

AR: Author's response | RR: Referee report | ED: Editor decision

ED: Reconsider after major revisions (further review by editor and referees) (08 Dec 2017) by Uwe Ulbrich

AR by Thomas Petroliagkis on behalf of the Authors (29 Dec 2017)

ED: Referee Nomination & Report Request started (23 Jan 2018) by Uwe Ulbrich

RR by Anonymous Referee #3 (16 Feb 2018)

Suggestions for revision or reasons for rejection

The manuscript addresses an important topic and has some interesting results. Although I think the results should be published, I cannot recommend publication of this manuscript in its present form. I had already read the technical report by the author which is available online and I was therefore familiar with the key findings, which I found exciting. This paper is based on the report, where most of the figures (and long text paragraphs) can be found in the same way. I don’t think there is any harm in writing a detailed technical project report first and then also publish a peer-reviewed article from it, especially if it contains relevant scientific advancements. However, and maybe that is a personal opinion, I believe that the paper should be a concise summary of the report (which can be cited and referred to). In this case, it would likely be the other way round. The author has obviously decided to publish the report as a series of three papers (with relatively minor changes), which together would probably be much longer than the report (which has 72 pages, this first paper has already more than 30). That is also reflected in the manuscript. There is lot of text book information and long text paragraphs repeating certain aspects (e.g., the entire introduction until page 4 repeats over and over how dependence between storm surge and wave height is important from various different angles). Subsequent sections are also unnecessarily long and the manuscript only becomes interesting toward the end. I am really wondering how the other two manuscripts for parts 2 and 3 will look like, since the same (or very similar) methods must have been used. Will the first half in each manuscript be nearly the same?
I am sorry that I cannot be more supportive of the manuscript and I leave it to the editor to decide if NHESS is the right outlet for the anticipated series of three papers. If the other two papers will have a similar length I would always opt for reading the report instead of the papers, and I believe it should be the other way round. I encourage the author to reconsider and think about publishing one concise paper on interdependencies in drivers of coastal flooding across Europe. In such a paper, references can be made to the report regarding methodological details and model validations, whereas the paper itself should focus on the results and their implications.

Hide

RR by Anonymous Referee #1 (05 Mar 2018)

Suggestions for revision or reasons for rejection

Dear Editor,
I have read the answers of the author to my comments. Here are my further comments referring to the points listed in “nhess-2017-177-author_response-version2.pdf”.

I confirm that the paper is in my view interesting, the new structure has improved the readability of the article, but the new version remains affected by two shortcomings that I had already noticed in the first review: a) the reliability of the hindcasts (particularly for waves and for intense/extreme events) is not convincingly documented b) there insufficient effort to interpret the dependences that are found. On both these aspects I think that progresses are possible with a reasonable, but not huge effort.

Considering the specific answers of the author to my comment:

(1) Validation of hindcasts and its presentation

The goal of this paper is, indeed, not the validation of the hindcasts. However, the reliability of the analysis depends on the accuracy of the hindcasts, which is essential for this study. The minimal requirement is a documented validation of the two (storm surge and wave) hindcasts in previously published studies, which is what the author has partially added. However, the answer of the author is not fully clear on this respect, particularly as the published article validating the wave hindcast is concerned.

I had formerly commented that “No sufficient attention is paid to assessing whether storm surge and wave simulations are capable of reproducing extremes values. Correlation and bias are not representative in this sense”. I find in the revised version no new material answering to this request.The added information referring to Vousdoukas et al. (2016), in not relevant for intense and extreme events. Further absolute values are not representative of the actual importance of these errors. I think that values would be more informtive if they are normalized with the values of the mean observation.

Concerning waves, author refers to Phillips, B.T., Brown, J.M., Bidlot, J.-R. and Plater, A.J.: Role of Beach Morphology in Wave Overtopping Hazard Assessment. J. Mar. Sci. Eng. 5(1), 2017. I have checked this article, but I Have not found the validation of the wave hindcast that the author claim. I might have missed it. A specific indication by the author is required.

In synthesis, considering the new version I remain unconvinced that the statement “the overall performance of both surge and wave hindcasts is considered satisfactory” is documented at a convincing level. If the paper main goal is demonstration a methodology, this may be an acceptable limitation, but some rephrasing, admitting it and suggesting therefore, caution on the results that are found, is advisable

other minor points in section 3
Section 3.1, line 6 depending on whether the hindcast includes the computation of nonlinear tide-surge interaction, the “residual” can include tide surge interaction as well
Section 3.1 Line 1 I think correct English expression would be “observation mode” … anyway note that I am not considering in this review typos and

(2) Description of statistical methods

I am positive with the new version, which I think is more readable and useful to the reader.
The section “statistical dependence" has now become very short, I suggest to merge it ( as a separate subsection) with “Data and Metoodology" in a single section

(3) Spatial variations of dependence and correlation, interpretation of results

Obviously, the effect of atmospheric pressure (that in steady condition is described by the inverse barometer effect) is included in the model simulations. My comment was a suggestion that lack of dependence could be related to a large contribution of atmospheric pressure on storm surges, because waves are not affected by it. Large dependence could be related to the common action of wind on sea level and waves, depending on the characteristic of the fetch and on water depth along it. On this respect I think that the comment and the added material does not answer to my request and the manuscript still would benefit from a discussion with an interpretation of results.

I accept that “a thorough understanding of all factors leading to a compound event is above the scope of this study“, but some discussion should be added, anyway. Any indication on under which conditions dependence would be high and when low?

I suggest to split the final section “discussion and conclusion” in two sections: “discussion” and separated “conclusion”. Reading it is not clear tome whether the goals of the paper is to demonstrate a methodology or to discuss dependences between storm surge and wave events at a representative set of European RIEN points

Comments on figures

In general figures need to be improved
Names in figure 1 will likely not readable when reduced too journal size
Quality of figure 2 is poor. I expect no geographical name will be readable and dots will be hardly visible when reduced to standard journal size
Graphic quality of fig 3 is low. It contains a lot of useless details (e.g. roads and motorways labels) and unreadable (and useless in this context) names of towns and locations
In figs 4 and 6 annotations are too small. They will be unreadable in the printed journal article. This problem is present in most figures

Hide

ED: Reconsider after major revisions (further review by editor and referees) (26 Mar 2018) by Uwe Ulbrich

AR by Thomas Petroliagkis on behalf of the Authors (12 May 2018) Author's response Manuscript

ED: Referee Nomination & Report Request started (18 May 2018) by Uwe Ulbrich

RR by Anonymous Referee #1 (02 Jun 2018)

ED: Publish subject to technical corrections (15 Jun 2018) by Uwe Ulbrich

AR by Thomas Petroliagkis on behalf of the Authors (17 Jun 2018) Manuscript

Download

Article (4728 KB)
Full-text XML

Short summary

Compound events – that is, two or more events occurring simultaneously – can lead to high impacts, even if the two single events are not extreme per se, i.e. only their combination. The possibility of utilizing statistical dependence methods in coastal flood hazard calculations is investigated. The statistical analysis between storm surge and wave height is performed over 32 river ending points along European coasts. The top 80 compound events are defined, and their low-level flow is analysed.