Potential glacier contributions to the 2024 La B&eacute;rarde flood

Ogier, Christophe; Werder, Mauro A.; Gagliardini, Olivier; Santin, Ilaria; Moser, Raphael; Hugonnet, Romain; Blanc, Antoine; Farinotti, Daniel

doi:10.5194/egusphere-2026-466

Preprints

https://doi.org/10.5194/egusphere-2026-466

Preprints

09 Feb 2026

| 09 Feb 2026

Potential glacier contributions to the 2024 La Bérarde flood

Christophe Ogier, Mauro A. Werder, Olivier Gagliardini, Ilaria Santin, Raphael Moser, Romain Hugonnet, Antoine Blanc, and Daniel Farinotti

Abstract. On 20–21 June 2024, an unprecedented flood of the Etançons river caused important damage to the village of La Bérarde (Écrins, France). An analysis of the event showed that the flood was partially caused by the combination of an intense rain-on-snow event at high altitude and the drainage of a supraglacial lake from Glacier de Bonne Pierre. In this study, we quantify the water volume that could have also been trapped beneath the glacier in local minima of the hydraulic head, i.e., in locations that could host so-called glacier water pockets impounded by hydraulic barriers. In the absence of direct observations of water pockets, we use a numerical, steady-state approach that computes the subglacial hydraulic head from surface and bedrock topography of Glacier de Bonne Pierre. As of June 2024, hydraulic barriers at Glacier de Bonne Pierre could, in theory, have impounded water volumes on the order of 10⁵ m³, with the largest modeled water pocket beneath a surface depression that temporarily hosted a supraglacial lake. These results provide a first-order estimate of the potential subglacial water storage capacity prior to the June 2024 flood. We propagate uncertainties in surface elevation, bedrock elevation, and flotation fraction (the ratio of basal water pressure to ice overburden pressure) through a stochastic framework and show that spatial variability in the flotation fraction dominates the uncertainty in the resultant water pocket volumes. This highlights the strong sensitivity of subglacial water-routing results to poorly constrained basal water pressure conditions. While acknowledging that the actual presence and contribution of such water pockets cannot be confirmed from available observations, our study highlights the glacial flood potential of debris-covered glaciers with pronounced surface topographic depressions, which can promote both supraglacial and subglacial water storage.

Received: 26 Jan 2026 – Discussion started: 09 Feb 2026

Publisher's note: Copernicus Publications remains neutral with regard to jurisdictional claims made in the text, published maps, institutional affiliations, or any other geographical representation in this paper. While Copernicus Publications makes every effort to include appropriate place names, the final responsibility lies with the authors. Views expressed in the text are those of the authors and do not necessarily reflect the views of the publisher.

Download & links

Christophe Ogier, Mauro A. Werder, Olivier Gagliardini, Ilaria Santin, Raphael Moser, Romain Hugonnet, Antoine Blanc, and Daniel Farinotti

Status: final response (author comments only)

RC1: 'Comment on egusphere-2026-466', Anonymous Referee #1, 12 Feb 2026

The article analyzes the water pocket outburst flood (WOP) in the Glacier de Bonne Pierre and its potential impact on the 2024 La Bérarde flood.
The study focuses on the significant research interest in flooding related to climate change, particularly in comparison to glacier lake outburst floods (GLOFs). The article is well-written and easily comprehensible. The discussion analyzes different scenarios in which GLOFs could potentially contribute to flooding in 2024, as well as the uncertainties associated with this analysis.
My opinion is that the article fits the scope and standards of NHESS and should be published.
Minor issue.
Fig. 2: In the axis there aren't X and Y, there should be East and North.

Citation: https://doi.org/10.5194/egusphere-2026-466-RC1
RC2:
'Comment on egusphere-2026-466', Anonymous Referee #2, 27 Feb 2026
This study explores the critical role of Glacier de Bonne Pierre as a potential secondary water source that may have further exacerbated the catastrophic June 2024 flood which devastated the French village of Bérarde. The authors model the potential subglacial water storage capacity to investigate whether a hidden “water pocket” outburst could have contributed to the flood. Overall, this is a well written manuscript offering an innovative perspective on a complex disaster. The authors’ transparent approach to accepting the limitations of their study – such as the lack of direct subglacial observations and the use of an arbitrary correlation length of 100 m – is highly appreciated. Their findings suggest the disaster was likely intensified by a glacial outburst, either through the rupture of a large internal cavity or a sudden reorganization of the subglacial drainage network. The manuscript effectively draws attention to invisible glacial reservoirs in mountain catchments and the adverse role they may play under extreme meteorological conditions. Thus, I am in favor of accepting the manuscript with minor revisions. Its subject area falls squarely within the scope of NHESS, and I believe it will make a valuable contribution to the journal once these suggested improvements are addressed.
From my perspective, the primary areas for improvement revolve around contextualizing the study’s limitations earlier and addressing potential discrepancies in the timeline of the data:
Pre-flood vs post-flood topography: The study relies on post-flood data to model pre-flood states, specifically using a June 28 DEM for surface elevation and an October 28 DEM to derive bedrock elevation. This approach assumes that the surface and bedrock conditions did not change enough between June and October to invalidate their hydraulic barrier models. However, a massive flood – including the drainage of an estimated 100,000 m³supraglacial lake – occurred in the interim. If the topography indeed remained largely changed (which would be rather curious), this justification should be explicitly stated in the text. If they indeed did change due to the event, that physical alteration should be accounted for. I would expect to see an estimate or discussion of this in the manuscript.

Limitations of the steady state model: The reliance on a steady-state model to simulate a highly dynamic event is a major limitation, which the authors also recognize. While a steady-state approach provides an effective way to estimate maximum potential storage geometry, it fundamentally falls short of simulating the dynamic, time-dependent processes of the actual flood (e.g. cavity opening, ice creep, or a rapid reorganization of the drainage system). A detailed explanation of this model selection – including its specific pros and cons – earlier in the manuscript (e.g., in the Methods or Introduction sections), rather than being reserved primarily for conclusion (or as a potential future step if I understood it correctly), would be beneficial to the manuscript.

The hydrological budget: Because critical river discharge estimates became highly uncertain during the intense sediment transport of the flood, the study cannot successfully close the hydrological budget. This context is vital, as it renders the 160,000 m³ calculated subglacial volume strictly a theoretical estimate of potential capacity, rather than a measurable “missing link” definitely proven to be in the flood waters. Ensuring that this distinction remains sharply in focus throughout the text will strengthen the paper’s scientific rigor.
Citation: https://doi.org/10.5194/egusphere-2026-466-RC2
RC3: 'Comment on egusphere-2026-466', Anonymous Referee #3, 12 Mar 2026

The study overall discusses the reasons for a flood event in France. It mainly focuses on the possibility of glacier impact on that flood event. The quantification of the glacier water volume is based on the computational modelling. The results generally indicate the flood-generating potential of the glacier. The following corrections are suggested in order to contribute to the improvement of this study:
Introduction
Particular studies are referred in Introduction, mainly concentrating on this specific glacier. However, other studies exist in the literature, which have focused on the potential contributions of other glaciers in the world for the flood-generation. Their modelling or observation studies should also have been referred to serve as background of the necessity of this study.
Methods
Please refer to the literature for all the formulation of the calculation.
Results and Discussion
Referencing is also valid for the Results. However, they should be discussed within the Discussion section, since it is another separate section. On the other hand, discussion was conducted such as newly introducing the study. In fact, event-based other studies should also be referenced and discussed for sake of comparison with this study results. Otherwise, the study already stands alone as being a solely extreme event.

Citation: https://doi.org/10.5194/egusphere-2026-466-RC3
RC4: 'Comment on egusphere-2026-466', Anonymous Referee #4, 16 Mar 2026

The comment was uploaded in the form of a supplement: https://egusphere.copernicus.org/preprints/2026/egusphere-2026-466/egusphere-2026-466-RC4-supplement.pdf

Citation: https://doi.org/10.5194/egusphere-2026-466-RC4
RC5: 'Comment on egusphere-2026-466', Martin Lüthi, 23 Mar 2026

Dear Collegues,
A few questions, especially related to the assumption of a temperate

glacier, merit evaluation and discussion:
- A large lake forming on a temperate glacier is a quite unusual

occurrence. Usually, in temperate ice, channels and cracks do no

heal (refreeze) easily, and once a channel has been enlarged it will

likely persist for an extended period of time, especially at the low

ice thicknesses of ca 100 m.
Could it be, from similarity with Tete Rousse etc, that also this

glacier features cold ice, at least in some areas?
- It would be interesting to assess how the supraglacial lake changed

the hydraulic potential landscape. Ponding lake water could easily

create an upstream dam/barrier (in the potential) within which a

subglacial lake could form. Drainage of the surface lake would then

simultaneously lead to a change of the potential barrier, and let

the subglacial lake drain.
- If I understand correctly, large parts of the glacier were mapped

with Lidar twice in 2024. Did you look at the differences of the

DEMs? Are there any traces of subglacial collapse features, or

lowering of the surface that cannot be explained with a melt model?
- It would be interesting and instructive to show the river discharge

curves, as well as results from the hydraulic modeling.
- The conclusions should be considerably shortened, and just highlight

the essential new conclusions. Now it reads like an extended

abstract. Be to the point, and highlight the essence of the paper,

the conclusions!
Minor comments
- The units of water volumes are given as 130 x 10^3 m^3. I see the

advantage of using something the is easy to understand, but this has

not been consistently used (Fig 5 shows 10^5). I propose "Olympic

Swimming Pools". Joking aside, settle on one unit and use it

consistently. Personally I would prefer 1.3*10^5, but 10^3 units are

fine, too. But please adopt the mathematical writing with 1.3\cdot 10^5.

The \times cross is ugly and implies an operator.
- Check the proper plural forms with a native speaker (e.g. "water

pockets volume" vs. "water pocket volume", etc.). English has

grammar rules that are quite different from French, and yours sound

often wrong to me. But maybe it is just my gut feeling that is wrong.
l 5-8 Abstract sounds somewhat repetitive and could be streamlined and

condensed.
l 25 Bérarde (consistent with or without accent aigu)
l 28, 94 \sim 10^5 m3
l 39 Déline
l 40ff This sentence and enumeration should either be a proper

enumeration (if allowed by the journal), or rather be split in

four sentences.
l 49 "a water pocket"
l 56 "favourable for"
l 61 "allows us"
l 85 3*10^5 m3
l 85 exported -> eroded (?)
l 90 fails
l 117 one obviously also would be interested in differences of the

DEMs, and their relation to melt or subglacial processes.
l 126 "vertically": why vertically, electromagnetic waves do not feel

gravity. Better would be perpendicular to the surface
l 135 why not cite a more recent, in-depth analysis of radar signal

speed on alpine type glaciers, especially by comparing with

borehole depths?
l 139 \citet{Grab}
l 154 the value of the crucial quantity \rho_i should be given. For

most alpine glaciers it is likley 900 kg/m3 (not 917).
l 157 each pixel
l 160 just choose one or two relevant sources (not 10), and use \citep
l 166 "dampened" -> "smoothed / distributed"
l 170 shorten title
Fig 2 a) it would be helpful to indicate the glacier flow direction by

a blue arrow. Also make the green lake really green, and maybe

add a white border, such that it is visible
Fig 2 c) make the water lines wider, they are barely discernible from black
Fig 2 d) use a color map where not everything is the same red (e.g.

"jet")
Fig 2 all panels: use reasonable numbers for the color bars (not 67/101..)

or the random elevations in panel c
Fig 2 all panels: you could simplify the coordinates wrt. to an

offset, then the panels would be larger.

1. all in km

2. subtract an offset in x and y

3. call the coordinates "Easting" and "Northing" instead of x,y
Fig 2 make all panels exactly the same!
Fig 2 legend: streamline the repetitive parts,

remove the double plural (lines intervals)

l 203 "is" -> "was"
l 208 In GPR, water is often specular, i.e. the phase of the reflected

signal changes (e.g. Matsuoka). You should see this in the bed

echos, whether you get an inverse first reflection.
l 219 "water pocket height"
l 220 distribution ... is shown
l 222 / 227 etc 1.6*10^5, 1.48*10^5

Fig 3 Make the figure better as proposed for Fig 2:

- simple, short coordinate labels (km/offset)
Fig 3 use the same color for water reflections as in Fig 2! (maybe

light blue or pink?). Not black!
Fig 3 write "water pocket height" everywhere (discard plural 's')
Fig 4 Much like Fig 3 and Fig 2, also use reasoable color bar values

(not 21)
254 shorten title (leave away Glacier...., this is the topic of the whole paper)
258 But maybe it would also form another hydraulic dam that blocks the

subglacial water flow upstream of the lake. Such a subglacial lake

might be even bigger?
273 But should be very visible after lake drainage.
285 These arguments would be easier to follow if they were illustrated

with some sketches.
347 But before you invoked the smoothing effect of ice stress

transfer, which certainly should also somehow enter this

assessment here. Maybe just the elastic field, or some viscous

relaxation?
350 And also the detailed stress field at the base. The surface

geometry and boundary conditions lead to a non-hydrostatic stress

field at the base, which differs substantially from the simple

\rho*g*H assumption.
353 But then, debris has a void space of 30-50%, such that density is

not that different from compact ice.
358 "water pocket volume" (also 359) (be careful with plural in

English! there are lots of these errors everywhere)
360 Water pockets alone are not important, but only the fraction that

can drain easily.
372 "produces"
Fig 5: Why are 10^5 units shown here, whereas throughout the text

there are 10^3 units. the labels could be considerably shortened

by leaving away ".0" and also 10^5 (could be added to the axis label).
Fig 5 caption: leave away "noted" (you mean denoted, but leave it away)
Fig 5 caption: "crosses mark" -> "crosses"
Fig 5 caption: shorten and streamline the text
Acknowledgments: I think you should add the full SNF project number

(starting with 200020_ or similar), not just a part of it.

Citation: https://doi.org/10.5194/egusphere-2026-466-RC5
RC6:
'Comment on egusphere-2026-466', Anonymous Referee #6, 25 Mar 2026
This is an interesting and worthwhile study that uses traditional techniques to estimate the volume of water stored in subglacial water pockets using hydraulic potential calculations in service of better understanding a multifaceted flood event in France in June 2024. In addition to its interesting application to a specific hazard event, what differentiates this study from others is its statistical estimation of uncertainty in water-pocket volume arising from surface/bed DEM uncertainty and assumed subglacial flotation fraction uncertainty. The paper is clearly written and structured and includes an appropriate level of detail in the methods and results. With minor revisions, and perhaps a few tweaks to the uncertainty analysis, I think it will make a nice contribution to the glaciological/hydrological hazards literature.
General queries:
Spatial smoothing of high-resolution surface DEM. The high-resolution LiDAR DEM has to be smoothed for the purpose of estimating bed elevation and hydraulic potential at an appropriate scale. Given the assumptions underpinning subglacial hydraulic potential analysis, I would imagine one should smooth the ice-surface DEM using a kernel large enough to eliminate features that do not influence the hydrostatic stress at the bed. I assume such a kernel would equal or exceed the ice thickness, whereas 10% of the local ice thickness was used here. A more thorough justification would be appreciated here, but even better, a demonstration of the sensitivity of the results (if any) to this kernel size (i.e., include this as an uncertain parameter in the larger uncertainty analysis, if interesting). This could be interesting and useful for other studies, now that high-resolution DEMs are increasingly available, and surface features (e.g. crevasses, rocks, moraines) that do and do not geometrically influence subglacial water flow must be distinguished.

Range of f and l_f in uncertainty evaluation. Since flotation fraction f and correlation lengthscale l_f impart much greater uncertainty in water-pocket volume than the bedrock and surface elevations as examined, I am curious about the choices of ranges of each. What are the results for, e.g., f=[0.65,1.05] instead of f=[0.9,1.1]? The former seems more realistic to me than the narrow range (symmetric about 1) tested. Similarly, what are the results for l_f=10m as a minimum instead of 50m? Does this choice matter? 10m seems like a more reasonable lower bound from borehole studies, as was cited in the paper. Related, I found some of the statements on page 15, lines 361-369, surprising: “f=0.9 reflects relatively low water pressure…” and “f=1.1 represents a pressurized drainage system…”. The latter is certainly true, but this would only, to my knowledge, occur in isolated pockets, rather than characterize a reasonable assumption about the drainage system itself.

Details (page.line):
2.35-40. I find it bizarre that an outburst from a subglacial water body formed from high geothermal flux is a GLOF, but an outburst from a subglacial water body formed without high geothermal flux is called a WPOF. Am I reading this right?
4.87-91. Would be nice to state here that it was not possible to close the water budget (so sad!) as is stated later in the paper (not til 13.292). At this point the reader may wonder why the water budget, or at least the total water volume of the event, is not used as a clue.
5.121. Rather than, or in addition to, reporting 50 MHz as GPR antenna frequency, report wavelength in ice (which will be different than in air). It would be nice for the reader to have this and basic antenna characteristics reported here (antenna type, dipole lengths if applicable) even though the system is described elsewhere.
6.164. Moving average filter on surface elevation (see also query above). What is the kernel shape: boxcar, Gaussian? What is the spatial scale that defines the “local” ice thickness here?
7.175-179. \psi_S = \psi? Not sure what subscript S means here.
8.Fig 2. Hard to see the difference between blue and black lines in (c). Please change/enhance attribute difference between the two.
9.206-208. Two closely spaced reflectors. Here it would be relevant to know the GPR wavelength in ice and the typical separation of these reflectors. I would also suggest giving these features a descriptive rather than genetic name (e.g. “double/multiple basal reflectors” vs “subglacial water reflection”) given the difficulty of interpreting such features and the expectation that the bed is fully temperate (i.e. likely to have some amount of water everywhere).
9.214 “This is because the modeled… derived from hydraulic head”. I think the narrow nature of the suggested pathways is also a consequence of the D8 algorithm (steepest gradient) chosen. Alternatives that partition flow between all down-gradient cells (e.g. D-Inf) would produce less concentrated/narrow pathways.
12.255 I appreciated the presentation of the two scenarios. That was thought provoking. It made me wonder about the observations of supraglacial lake drainage in Greenland (Stevens et al., 2015) and whether nonlocal drainage (e.g. through a crevasse that suddenly connects to the bed) could induce supraglacial lake drainage via the flexure caused by the antecedent drainage producing crevasses in/near the lake basin.
13.310. “zones of low hydraulic gradient” AND low hydraulic potential (local potential highs can have low hydraulic gradients).
13.314. “linked cavity drainage system”. I think a connected drainage system is all that is required for this argument. Is there some reason the morphology of the system must be linked cavities instead of whatever other inefficient drainage system?
20.FigA.1 Nice semivariogram.
References:
Stevens, L.A., Behn, M.D., McGuire, J.J., Das, S.B., Joughin, I., Herring, T., Shean, D.E. and King, M.A., 2015. Greenland supraglacial lake drainages triggered by hydrologically induced basal slip. Nature, 522(7554), pp.73-76.
Citation: https://doi.org/10.5194/egusphere-2026-466-RC6

Christophe Ogier, Mauro A. Werder, Olivier Gagliardini, Ilaria Santin, Raphael Moser, Romain Hugonnet, Antoine Blanc, and Daniel Farinotti

Viewed

Total article views: 706 (including HTML, PDF, and XML)

HTML	PDF	XML	Total	BibTeX	EndNote
469	211	26	706	44	67

HTML: 469
PDF: 211
XML: 26
Total: 706
BibTeX: 44
EndNote: 67

Views and downloads (calculated since 09 Feb 2026)

Month	HTML	PDF	XML	Total
Feb 2026	313	144	15	472
Mar 2026	143	52	10	205
Apr 2026	13	15	1	29

Cumulative views and downloads (calculated since 09 Feb 2026)

Month	HTML	PDF	XML	Total
Feb 2026	313	144	15	472
Mar 2026	143	52	10	205
Apr 2026	13	15	1	29

Viewed (geographical distribution)

Total article views: 688 (including HTML, PDF, and XML) Thereof 688 with geography defined and 0 with unknown origin.

Country	#	Views	%

Latest update: 12 Apr 2026

Short summary

In June 2024, a destructive flood impacted the village of La Bérarde in the French Alps. Rain, snowmelt, and the drainage of a surface lake on a glacier cannot fully explain the flood magnitude. We used glacier topography to estimate how much water could also have been stored beneath the glacier before the event. Our results show that large volumes of hidden water may have existed and could have amplified the flood, highlighting an overlooked hazard in debris-covered mountain glaciers.


Total:	0
HTML:	0
PDF:	0
XML:	0