Deep uncertainties in shoreline change projections: an extra-probabilistic approach applied to sandy beaches

Thiéblemont, Rémi; Le Cozannet, Gonéri; Rohmer, Jérémy; Toimil, Alexandra; Álvarez-Cuesta, Moisés; Losada, Iñigo J.

doi:https://doi.org/10.5194/nhess-21-2257-2021

Articles | Volume 21, issue 7

https://doi.org/10.5194/nhess-21-2257-2021

© Author(s) 2021. This work is distributed under
the Creative Commons Attribution 4.0 License.

https://doi.org/10.5194/nhess-21-2257-2021

© Author(s) 2021. This work is distributed under
the Creative Commons Attribution 4.0 License.

Articles | Volume 21, issue 7

Research article

|

30 Jul 2021

Research article |

| 30 Jul 2021

Deep uncertainties in shoreline change projections: an extra-probabilistic approach applied to sandy beaches

Rémi Thiéblemont, Gonéri Le Cozannet, Jérémy Rohmer, Alexandra Toimil, Moisés Álvarez-Cuesta, and Iñigo J. Losada

Download

Final revised paper (published on 30 Jul 2021)
Supplement to the final revised paper
Preprint (discussion started on 13 Jan 2021)

Interactive discussion

Status: closed

RC1:
'Comment on nhess-2020-412', Anonymous Referee #1, 02 Feb 2021

This paper examines the extent of deep uncertainty in studies on projections of shoreline changes.

First of all, I find elegant the definition of deep uncertainty based based on the concept of "possibility" in an "extra-statistics" context.

This could be indeed an interesting tool to assess more correctly the variability of predictions, when we have a reasonable range for the epistemic uncertainty.

The paper is well written and well structured. The explanation of what possibility is and of the model(s) used for shoreline predictions and their parameters is clear.

I believe a few lines would be useful, summerizing how the montecarlo over the "possibility", possibly with references. Is a possibility distribution treated mathematically like a probability distribution? Or is it used to select a CDF?

Apart from this request of clarification, I have only minor comments.

- formula 1: the angle is missing after tan. I would suggest to call the angle "beta" as "alpha" can be confused with the confidence level used earlier, and beta this is the symbol generally used in the Bruun formula.

- line 156, a brief explanation of what nx, Tx and lvar are should be given here, mentioning that the terms will be explained later in more detail.

- line 157, substitute "in the following" with "below".

- figure 2: I would appreciate a brief summery on how the hybrid montecarlo works.

- line 173-174: is lvar an uncertainty term, with and unknown sign? It would look like this, as it is computed as a standard deviation. This should be made more clear, for example writing +/- lvar in the formula.

- line 277: the acronym GNSS should be defined.

- line 308: the acronym SONEL shoudl be defined.

- table 1: I would suggest to indicate also the possibilistic choice of the model here.

- line 240 and elsewhere: check what sign you use to indicate erosion/accretion. Here and in figure 4, a positive change is erosion. But looking at figure 5, it looks that the reverse convetion is used (a negative shoreline change where you have erosion).

- figure 6: I would suggest to add lines to identify the extent/position of the ambiguity and of the high/low ends.

- figure 7: Is it possible that fixing some param values the ambiguity increases? Maybe a line explaining this would be useful.

- figure 8: the figure is truncated, the x axis is missing.

Citation: https://doi.org/10.5194/nhess-2020-412-RC1
- AC1: 'Reply on RC1', Rémi Thiéblemont, 08 Apr 2021
  
  The comment was uploaded in the form of a supplement: https://nhess.copernicus.org/preprints/nhess-2020-412/nhess-2020-412-AC1-supplement.pdf
  
  Citation: https://doi.org/10.5194/nhess-2020-412-AC1
RC2:
'Comment on nhess-2020-412', Anonymous Referee #2, 07 Feb 2021

The study 'Deep uncertainties in shoreline change projections: an extra-probabilistic approach applied to sandy beaches' explores an elegant method to deal with a combination of a aleatoric (intrinsic) and epistemic (deep) uncertainties. The methods are well explained and seem easy to apply and very helpful to better understand different kind of uncertainties. And so, the manuscript convincingly illustates the methods' attractiveness.

Given the attractiveness of the method and well-written manuscript (as discussed in the last section), I am a little surprised that some major sources of (deep) uncertainties are not included. Especially, the choice to only use one set of sea-level projections that do not include high-end scenarios seems a little odd. There are multiple recent sea-level projections that explicitly included high-end contributions of the W-AIS (e.g. Le Bars et al. 2016, Wong et al., 2017).

# Minor:

- the applied sea-level projections need a little more explanation. Do I understand well that 21 minus 2 CMIP projections were used? Could you explain in one sentence why two runs were judged 'unrealistic' with respect to sterodynamic behaviour (and others not). And could you explain how sources of uncertainty other than sterodynamical were included (or were they excluded)?

- In figure 4, could you explain how the linear fit was made? In 4b, the fist black dot (yr=1989, shore line change ~ -491) seems to be excluded. Otherwise, I would expect a very different R2.

References:

Dewi Le Bars et al 2017 Environ. Res. Lett. 12 044013

Wong, T.E., Bakker, A.M.R. & Keller, K. Impacts of Antarctic fast dynamics on sea-level projections and coastal flood defense. Climatic Change 144, 347–364 (2017). https://doi.org/10.1007/s10584-017-2039-4

Citation: https://doi.org/10.5194/nhess-2020-412-RC2
- AC2: 'Reply on RC2', Rémi Thiéblemont, 08 Apr 2021
  
  The comment was uploaded in the form of a supplement: https://nhess.copernicus.org/preprints/nhess-2020-412/nhess-2020-412-AC2-supplement.pdf
  
  Citation: https://doi.org/10.5194/nhess-2020-412-AC2

Peer review completion

AR: Author's response | RR: Referee report | ED: Editor decision | EF: Editorial file upload

ED: Publish subject to minor revisions (review by editor) (20 Apr 2021) by Ira Didenkulova

AR by Rémi Thiéblemont on behalf of the Authors (07 May 2021) Author's response Author's tracked changes Manuscript

ED: Publish as is (21 May 2021) by Ira Didenkulova

AR by Rémi Thiéblemont on behalf of the Authors (02 Jun 2021)

Download

Article (4897 KB)
Full-text XML

Short summary

Sea level rise and its acceleration are projected to aggravate coastal erosion over the 21st century. Resulting shoreline projections are deeply uncertain, however, which constitutes a major challenge for coastal planning and management. Our work presents a new extra-probabilistic framework to develop future shoreline projections and shows that deep uncertainties could be drastically reduced by better constraining sea level projections and improving coastal impact models.