A deep learning reconstruction of mass balance series for all glaciers in the French Alps: 1967–2015

Bolibar, Jordi; Rabatel, Antoine; Gouttevin, Isabelle; Galiez, Clovis

doi:https://doi.org/10.5194/essd-12-1973-2020

Articles | Volume 12, issue 3

https://doi.org/10.5194/essd-12-1973-2020

© Author(s) 2020. This work is distributed under
the Creative Commons Attribution 4.0 License.

https://doi.org/10.5194/essd-12-1973-2020

© Author(s) 2020. This work is distributed under
the Creative Commons Attribution 4.0 License.

Articles | Volume 12, issue 3

Data description paper

|

03 Sep 2020

Data description paper |

| 03 Sep 2020

A deep learning reconstruction of mass balance series for all glaciers in the French Alps: 1967–2015

Jordi Bolibar, Antoine Rabatel, Isabelle Gouttevin, and Clovis Galiez

Download

Final revised paper (published on 03 Sep 2020)
Supplement to the final revised paper
Preprint (discussion started on 20 Mar 2020)
Supplement to the preprint

Interactive discussion

Status: closed

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

- Printer-friendly version

- Supplement

RC1: 'Review of Bolibar et al.', Anonymous Referee #1, 25 Mar 2020
- AC1: 'Authors' response to Anonymous reviewer 1', Jordi Bolibar, 13 May 2020
RC2: 'review of Bolibar et al.', Ben Marzeion, 04 May 2020
- AC2: 'Authors' response to Ben Marzeion', Jordi Bolibar, 13 May 2020

Peer-review completion

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

AR by Jordi Bolibar on behalf of the Authors (18 May 2020) Author's response Manuscript

ED: Referee Nomination & Report Request started (12 Jun 2020) by Reinhard Drews

RR by Matthias Huss (24 Jun 2020)

Suggestions for revision or reasons for rejection

The authors have carefully revised their manuscript which has substantially improved and is almost ready for acceptance in my opinion. The additional work performed strengthens the conclusions of the article. However, some issues remain that should be addressed by the authors. Some of them have already been mentioned in my first review. I thus just copy in my former comment that I would like the authors to answer:

“A major problem of the paper as it stands now is, in my opinion, the lack of validation with independent measurements: The training data set for 32 glaciers is based on “remote-sensing” (Rabatel et al., 2016). As this lies the foundation to the entire study, more effort should be invested to describe this data set (methods, uncertainties). The training data set also contains models and assumptions – the annual SMB of these glaciers has not been directly measured. This needs to be emphasized. Measured (!) information on SMB is available from two sources: (1) the direct glaciological surveys, (2) geodetic surveys.

Although (1) is probably included in the training data set, it should explicitly be shown (e.g. figure with cumulative SMBs) how well the DL approach reproduces the observed SMB series. This would give direct evidence on the performance of the approach, independent from the training data set by Rabatel et al. (2016) that also includes model assumptions.”

The description of the basis of the principal training data set (Rabatel et al., 2016) is still very limited (see page 3, line 16). The reader just knows that it is based on “remote sensing”. However, it is clear that remote sensing alone is (yet) unable to provide annual glacier-wide mass balance. It should be described which approaches and assumptions went into the training data set and that glacier-wide mass balance is not measured, as in the case of the four GLACIOCLIM series, but is based on modelling optimally constrained with geodetic mass balances and snowline observations. I do not at all want to imply that the data set is unsuitable or inaccurate but it should be honestly stated that these are not actually “observations” as the annual mass balances are referred to several times. Related to this, I would also ask the authors to follow my advice of comparing the annual (!) series from their approach against the four GLACIOCLIM data sets. This is directly measured data and would allow investigating whether DL is able to capture these observed year-to-year variations.

Some additional comments:

- What is shown on top of page 4 is not actually Data. It rather seems to be a description of the method to incorporate them. Please try to better separate data used in the approach from the methods. I realise though that this is not trivial to achieve here.

- A general comment (sorry, for not bringing this up already in the first review): Throughout the paper mass balance is referred to as SMB (surface mass balance). But does this approach actually compute SURFACE mass balance? And not rather the total mass balance of a glacier (see Cogley et al., 2011, for an overview)? The approach is trained on the Rabatel-dataset which depends on geodetic balances for constraining the mass loss. Geodetic mass balance covers both surface mass balance and basal/internal balance. Consequently, I would consistently omit the “surface” when you refer to mass balance. Anyway, this is just terminology, as it is clear that the difference between SMB and total mass balance will be minimal for Alpine glaciers.

- It is a pity that the new validation with independent geodetic mass balances has not made it into the main paper. I think it would strongly support the credibility of the data set and is important enough to not only be presented in the Supplementary.

Hide

ED: Publish subject to minor revisions (review by editor) (26 Jun 2020) by Reinhard Drews

AR by Jordi Bolibar on behalf of the Authors (06 Jul 2020) Author's response Manuscript

ED: Publish subject to technical corrections (14 Jul 2020) by Reinhard Drews

Dear Authors,

Thank you for your revisions of ESSD-2020-35. I appreciate your response-to-reviews + manuscript edits which clearly show that you have taken the reviewer’s comments seriously. As a result, the paper has improved. I agree with both reviewers that this paper deserves publication. Congratulations.

Below you will find some more comments that came up from another re-read by myself. Please include those prior to re-submitting the next iteration. The paper will still benefit from language edits and some structural changes suggested below. Overall, this paper is a significant effort, and the reconstructed MB time series will hopefully be taken up by many other scientists.

Kind regards, Reinhard Drews

Comment 1 comparison to interannual Glacioclim data: My understanding is that model training was done using annual MB only (e.g., p4, l. 20). If so, does this mean that the models has never seen the interannual Glacioclim time series? Please state this explicitly, as this matters for the evaluation of Fig. S1. Also, state in the caption of S1 that this is an inter-annual time series. This Figure is quite isolated and poorly linked to the main text (with one sentence on p5 l 15). This was one of the major concerns during review, and although I see that it has been taken into account, the takeway message is still muddled.

Comment 2 structure/presentation: Don’t use sentence like “..can be found in the supplementary information.” Let the reader know exactly which section you refer to, so that he/she can quickly find it. During another re-read also have an eye un unnecessarily convoluted sentences (e.g. p3, l14) which are better expressed in multiple shorter sentences.

Minor edits:

Replace “$a^{-1}$” with “a$^{-1}$” to avoid that only “a” appears in italics for units. Same for other units.

P1, l 5: “The method’s validity was assessed (add: previously) ….”

P2, l 23: “Two studies include reconstructions in the European and thus the French Alps…”. This sentence sounds funny (kind of like “..it is European and therefore French”). Rephrase.

P4 l 4: Point MB is also influenced by topography and climate. This sentence as an unclear meaning, please rephrase.

P4 l 11: How does the update from temperature to snowfall work?

P4 l 15: “. .”

P4 l 18: “thorough” -> “more detailed”

P4 l 4: “…each of the 60…”. I stumbled across this, as it is formulated as if the number 60 has been mentioned previously (but it didn’t). Better state “Our ensemble contains 60 LSYGO ….”.

P5 line 15: “The ANN MB reconstructions accurately reproduce the interannual variability of glaciological observations from the GLACIOCLIM observatory …. “

P6, l9: Reference section in SI.

P9 l10 “)(“ -> “) (“

P10 l 23 “The interannual variability …… (Fig. 6)”

P12 l5 Awkward sentence. Maybe “In the following, we argue that the good fit between observations and reconstruction is meaningful and not a result of overfitting.”

P12 l8 Reference section in SI.

P13 l1 “proceed”? Do you mean “originate” ?

P13 l25 “is the cornerstone”  “aids”

Space and time somehow done differently, so where is space?

Fig. 1 what does the left ice-cliff column represent? It is ok to have it in there, but it should be explained in caption.

Fig 3c what is the black dashed line? Is it the area-weighted mean?

SI page 4: Don’t state “shown in Fig. 4”. Make it simpler by stating “X is because of Y (Fig. 4)”. This applies also elsewhere in the main text.

Hide

AR by Jordi Bolibar on behalf of the Authors (15 Jul 2020) Author's response Manuscript

Download

Article (9130 KB)
Full-text XML

Short summary

We present a dataset of annual glacier mass changes for all the 661 glaciers in the French Alps for the 1967–2015 period, reconstructed using deep learning (i.e. artificial intelligence). We estimate an average annual mass loss of –0.69 ± 0.21 m w.e., the highest being in the Chablais, Ubaye and Champsaur massifs and the lowest in the Mont Blanc, Oisans and Haute Tarentaise ranges. This dataset can be of interest to hydrology and ecology studies on glacierized catchments in the French Alps.