Deriving the 100-Year Total Water Level around the Coast of Corsica by Combining Trivariate Extreme Value Analysis and Coastal Hydrodynamic Models

Round 1

Reviewer 1 Report

The reviewed manuscript is example of well-though and in-depth, high-tech paper, which addresses an issue important to both theoreticians and practitioners. This issue attracts attention of the international research community, and it is linked to the global environmental change. This manuscript is also rich in methodological ideas. It is generally well-written, well-illustrated, and well-referenced. Some minor changes are required – my suggestions are listed below starting from the most principal and ending with some technical.

• From methodological section, I'd prefer to learn more about software used for your analysis.
• Section 4: Results MUST be separated from Discussion, as required in all major international journals. This is easy to do in fact.
• In Discussion, it would be reasonable to state some policy implications, as well as to consider possible socio-economical effects of the anticipated coastal flooding. These discussions can be linked broadly to what is known from the other parts of the Mediterranean and Europe. For instance, this work can help: https://www.sciencedirect.com/science/article/pii/S0964569120304191
• From this work, I'd be curious to learn about the tectonic factor, i.e., coastal subsidence/rise, which can be significant on the scale of decades.
• Line 145: something is missed in ()?
• Please, control that all abbreviations (even well-known) are explained in the first place of their use.
• Please, do not start sentences with []. Indicate the authors in such cases and put [] after their names.
• Please, enlarge the figures to make all details well visible.
• Small linguistic polishing is necessary.

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Reviewer 2 Report

Dear authors,  

Thank you very much for an interesting paper that focuses on estimating the 100-year total water level around Corsica using a combination of statistical and numerical methods. The topic is highly relevant since communities across the globe struggle with accurately estimating coastal hazards in the current and future climate. The paper is clearly written however I have several major comments and therefore I would reconsider this manuscript after major revisions.

Main points.

1. Definition of total water level. Total water level (TWL) at the shoreline is the combination of tides, surge, and wave runup. Authors only compute wave setup in SWAN and a wind setup in SWASH. Wave-driven swash or runup is not taken into account. SWASH can be (more appropriately) be used to compute individual and runup.
2. Numerical modeling details. Provide more detail on the calibration and validation of the modeling train. For example, what settings were used (besides default), which calibration steps were performed, how well is the method doing compared to several historical storms and what are the errors involved?
3. Statistical analysis: the paper shows some of the intermediate steps for NWW3 560 but doesn’t mention or show this for other locations. Is the same threshold applied for the GPD applied everywhere? That seems problematic. 
4. Statistical uncertainty: the paper doesn’t explicitly compute uncertainty. What are the uncertainties of the statistical method and how does that compare to SLR projections. The latter is completely missed: SLR is not a static value but a range of possible values in the future given different RCP scenarios and possible outcomes of global models. It doesn’t suffice to use a French regulation value in an academic paper.
5. Quality figures: several figures are hard to read. Please improve the resolution and clarity.
6. General scientific relevance: the paper is completely written around the 2m/NGF threshold. While I can completely understand that when writing a report for a client, this is a scientific manuscript for the global community. One could focus on the efficiency of the method compared to brute force computations or the importance of wave-driven processes compared to SWL, etc.

Minor points

L11: recommend adding ‘often’ Since coastal flooding ‘often’ occurs…

L23: not clear what the 2m NGF is
L24: reference level of 1m along the eastern coast

L32: the introduction is a copy-paste of the abstract. Please revise.

L87: give more explanation of 2m NGF. Reference to MSL etc.

Figure 2a: what is water level /OD.

Figure 2B: hard to read the names
L201: seems incorrect. You stated before 60 cm uncertainty.

L323: please comment on the implications of this assumption

L372: suggest rephrasing the ‘Whatever’

L379: please explain the arbitrary margin of 25cm

 

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Reviewer 3 Report

The study entitled "Deriving the 100-year Total Water Level around the coast of Corsica by combining Trivariate Extreme Value Analysis and Coastal Hydrodynamic Models" considers a semiparametric multivariate extreme value analysis to derive joint probabilities of Hs, U and SWL, together with their covariates Tp, Dp and Du, to define extreme scenarios acting as input for coastal hydrodynamic models to compute the 100y-TWL along the coasts of Corsica. Results of the study indicate that the uniquely defined value of 2m relative to the French national topographic reference used in present studies overestimates flooding exposure of the coastal areas of Corsica to extreme marine current and future conditions. The work presented is interesting, the methodological approach fοllοwed includes innovative aspects and its conclusions are sound and informative for coastal flooding vulnerability studies. Some general and some minor comments/corrections are given below:

 

1) In Section 2.1 the authors should include some indicative results on data correction and validation, or at least provide existing references on these results, if any. In the present form of the work, this information is only shortly presented. Some indicative results, at least from the validation process, could significantly contribute to the completeness of the work.

 

2) It is not clear why the authors did not use only wave heights (and wind speeds) of directions affecting the coastal areas under study in their multivariate analyses. The authors could have selected to work only with wave events of directions affecting the coastal area of each nested box studied around Corsica Island (remove all events heading offshore). This choice could facilitate the multivariate analysis, eliminating the Dp (and associated Du) covariate and repealing the assumption of imposing the dominant direction leading to extreme Hs to extract extreme scenarios to drive coastal hydrodynamic models. The authors should comment further on this and defend their choice of working with the directional covariates.

 

3) The wave storm event duration is a crucial factor that can significantly affect flooding results. It is not clear if the authors tried to consider the duration of wave storm events in their analysis, at least in the first step of their methodology dealing with selection of events and data set constitution. In Section 3.1. it is mentioned that a block of a 3-days period with 1.5 days between each Hs peak is used, however it is not clear if each Hs peak is defined from wave storm events of different duration. In Section 4 it is emphasized that the approach followed is static, the duration of an event is not taken into account and therefore the volume of water is only limited by the capacity of the low-lying zone to fill up to an altitude corresponding to the TWL at the shoreline. The authors could refer to the work of Callaghan et al. (2008), among others, for considering event duration in the joint probability analysis conducted.

Callaghan, D. P., Nielsen, P., Short, A., & Ranasinghe, R. W. M. R. J. B. (2008). Statistical simulation of wave climate and extreme beach erosion. Coastal Engineering, 55(5), 375-390.

 

4) In Section 3.1. it is mentioned that to define the triplets of events, for each Hs peak, the SWL and U maxima were selected in a 12-hour window centered on the Hs peak. From this, it can be concluded that the wave height is defined as the primary variable for coastal flooding around the coasts of Corsica. Is this the case? What does the 12-hour duration represent? Please provide more details on this selection.

 

5) In Section 3.3.2 the authors mention that 34 scenarios are selected for each location to force the hydrodynamic models. How were these scenarios selected? What were the criteria used to aid this selection? Have the authors performed some form of sensitivity analysis based on extracted results from the hydrodynamic models?

 

6) In their “2100” extreme scenario, the authors consider an increase in sea level rise equal to 0.40m. However, it should be noted that the wave climate and associated uncertainties could significantly differ from those in present conditions. The authors should comment further on this issue.

 

7) At the beginning of Section 8 the authors could include estimates of joint extremal dependence on the different regions of Corsica (e.g. western and eastern), so that the reader can better conceive the differences analysed. Some measure of joint dependence, even for given values of SWL, could be really helpful for the reader.

 

Minor comments

 

1) Line 21 – “…Total Water Level (100y-TWL) …”

2) Lines 49-51 – “… one of the difficulties … which complicates …”

3) Line 51 – “[12] developed …”

4) Line 63 – “… consisting of less …”

5) Line 68 – “… in the Mediterranean Sea …”

6) Line 80– “… used a few dozens of …”

7) Lines 98-100 – “The strategy … has also been presented …”

8) Line 102 – “… limitations …”

9) Line 103 – “Conclusions are presented in the last section.”

10) Lines 106-109 – Please check the sentence “In this zone, … coastline.” for consistency.

11) Line 134 – “…, and run hydrodynamic simulations …”

12) Line 139 – “… and gives …”

13) Lines 142 & 145 – What is the use of the parentheses ()?

14) Lines 161-162 – “… for Confidence Intervals (CI) …”

15) Line 167 – “b) between Hs and Tp using the conditional distribution of wave steepness (St) …”

16) Line 177 – “… outputs of step 6;”

17) Line 191 – “…the Sea Level Rise (SLR) …”

18) Lines 228-230 – “For SWL, …the 100 year return level is 0.43 m. The threshold is set to 5.32 m for Hs and the 100-year Hs is 8.016 m, and set to 16.92 m/s for U leading to an 100-year U of 22.55 m/s.”?

19) Line 232 – “… Chi-squared ,…”

20) Lines 234-236 – “In addition, … the method of moments (MOM), the probability weighted moments (PWM) and the maximum likelihood method (MLE).”

21) Line 236 – “ … the MOM was used …”

22) Lines 239-240 – “…the extremes are … and waves with directions …”

23) Figure 3 caption – It is better not to mention the thresholds for Hs, U, SWL and estimated 100-year return levels, that have already been mentioned in the text. The Figure caption should better read “Example …. (a)… MARS 8 provides SWL (m). Generalized Pareto Distribution (GPD) and associated 70% and 95% CI for (b) Hs, (c) SWL and (d) U. Colorbars give …”

24) Line 279 – “… access the joint exceedance isocontours …”

25) Lines 289 – “Figure 5 (b) is …”

26) Line 298 – What is the use of the parentheses ()?

27) Line 320– “ … peak directions Dp, …”?

28) Figure 7 caption – “… dominant peak direction Dp ….”

29) Line 338 – “… appear to be less …”

30) Line 242 – “… the marginal distributions ….”

31) Line 357 – “… offshore conditions’ combination tested …”

32) Line 360 – “… study area is between …”

33) Lines 373-374 – “… (2.4 m when considering future conditions) seems to be overestimated …”

34) Lines 379 – “… 0.25 m margin on the TWL …”

35) Line 381 – “… did not apply …”

36) Lines 384 – “… need to work on.”

37) Line 391 – “… is still the topography.”

38) Line 399 “… to apply …”

39) Line 413 – “… (Figure 10 a) of Table 2 lead to …”

40) Line 438 – “Indeed, … coastal sectors with altitudes below 2 m/NGF, considered ….”

41) Lines 447 – “… hydrodynamic models …”

42) Lines 452 & 454 – “… low lying zone.”

Author Response

Thank you for your detailed review. Please see the attachment.

Author Response File: Author Response.pdf

Round 2

Reviewer 3 Report

The revised version of the manuscript is significantly improved compared to its original version. The authors tried to adress all the reviewers' comments. 

Some final minor syntax and grammar corrections are given below:

1) Line 332 – “…in a semi-enclosed basin …”

2) Line 461 – “… for the western and …”

3) Line 800 – “… in implementing a …”

4) Line 807 – “… duration the same way as Hs [54].”

5) Lines 810-811 – “This could be relevant to be taken into consideration, …”

6) Line 820 – “… remains the topography.”

7) Line 875 – “…along the French coastline …”

8) Line 879 – “We can note that these reports also identify …”

9) Line 882 – “Finally, the purpose here was the assessment …”

10) Line 910 – “…values obtained …”

11) Line 924 – “… the coast’s morphology.”