the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 19 Nov 2024
Failure of Marmolada Glacier (Dolomites, Italy) in 2022: Data-based back analysis of possible collapse mechanisms as related to recent morpho-climatic evolution and possible trigger factors
Abstract. A small, isolated portion of the Marmolada glacier partially broke off on July 3, 2022. The detached ice mass had an estimated volume of 70,400 m3 and slid down the slope killing 11 mountaineers after having travelled for approximately 2.3 km along the northern slope. This event is considered among the deadliest ice avalanches historically recorded in the Alps. The cause of the collapse is most likely to be closely related to anomalously high air temperatures in late spring and early summer of that year and also to progressive warming of basal ice layers and subglacial permafrost. Increased melting resulted in significant amounts of water that could not be discharged but remained trapped in crevasses of the isolated, cold ice body. The usually high permeability of karst systems in limestone are likely to have been blocked by subglacial and surrounding permafrost rocks, thereby increasing basal water pressure. Morpho-climatic data, satellite and aerial images and geophysics were analysed to better understand role and interaction of the controlling factors. This enabled the implementation of a conceptual model that was further investigated through numerical simulations using the Limit Equilibrium Method. The outcome from the back-analysis indicates that a combination of driving forces was required to finally trigger the collapse. The seasonal alternation of prolonged heating and refreezing meltwater of subglacial permafrost (i.e. the active layer) could have caused the plasticization of the basal ice resulting in the presence of a “weak bottom layer” where the available shear strength progressively dropped, finally triggering the failure.
- Preprint
(3277 KB) - Metadata XML
-
Supplement
(2288 KB) - BibTeX
- EndNote
Status: final response (author comments only)
-
RC1: 'Comment on nhess-2024-212', Martin Truffer, 31 Dec 2024
Review of Francese et al.
This paper has a lot of interesting information about the catastrophic glacier failure of a piece of the Marmolada Glacier that caused several fatalities and must be rated as one of the bigger glacier related natural disasters in Europe. With the rapid change of glaciers in the current climate, it is important to better understand the mechanisms that lead to such an event.
The paper does a good job in assembling a variety of data sets that allow for some conclusions. However, the paper would benefit from a significant reorganization and more detailed analysis. Parts of the paper go into great detail about things that may only be peripheral (parts of the permafrost discussion), while being really sparse on things that matter significantly (e.g. how the LEM model works).
This review is rather long, because I think the paper needs major work. However, a clear paper on this topic is an important contribution to the literature and I hope it can be put into publishable form.
The permafrost discussion is confusing. It appears that you argue that this glacier used to be temperate with a firn layer. Then it lost its firn layer and became cold-based. Then it started warming at the base again and weaken the ice, partially due to warming air temperatures and also heat infiltration through bed rock from the south facing side. While all of these are plausible at some level, this could be presented much more consistently. The Boeckli et al map is useful for context, but ultimately, the map and the nearby borehole will not allow much of an assessment of the thermal state. Neither do the thermal surface measurements. Those were acquired at a time when it could simply be cold surface temperatures that say nothing about the thermal state of the ground a few meters down. The most useful measurement is the borehole in the ice. This needs to be emphasized more. It clearly shows a cold ice base, although the base is barely below the level of where seasonal fluctuations would affect temperature.
The influence of the southern slope on the other side of the mountain is an interesting hypothesis, but it is very difficult to see how that heat flux would have allowed the glacier base to first freeze and then later thaw again. There is a lot of potential for some simple thermal model here, but even in the absence of that, the various influencing factors and hypotheses need to be stated much more clearly.The model discussion also needs to be made clearer and many questions remain. First, little detail is given about the LEM model. Second, it is not clear to me whether the parameters from Huang et al are applicable here. That paper finds peak strength in shear tests that are conducted under very high constant strain rates. Applying such high strain rates leads to stresses that are very high and to a quick ductile/brittle transformation. However, in the situation of this glacier, the geometry imposes certain stresses that would lead to deformation of ice, and, in this case, failure. The high stresses obtained in the Huang paper probably explain why it is so difficult to obtain failure in the model.
Some of the mechanisms in the model are confusing. What exactly does plasticization of the basal layer mean? First, it is not clear what the mechanism for this is (see the permafrost discussion above). Second, a 'plasticization' would lead to increased deformation rates, which should be observable. Hydraulic jacking is also a strange mechanism as applied here, because it is inconsistent with a frozen basal condition and a plugged drainage system. How does water access the entire base of the ice under these conditions?It is fair to assess how the LEM model does, since that is used in practice. But I think a first step is to use the fact that the glacier failed to estimate what stresses actually led to failure. You know that nature and magnitude of the failed surface, and you know all the relevant stresses (gravitational stress from the weight of the failed ice body and hydrostatic stresses from the water-filled crevasse. This allows you to calculate an average stress on the failed surface, which is a good place to start.
It seems overall, the extremely warm temperatures and the fact that a crevasse is filled with water to the top is probably the best indicator of how failure occurred. The excess pressure from the water-filled crevasse would have led to hydraulic jacking that would have progressively increased from the crevasse downward. In that area, shear strength would be lost entirely and the stress on the remaining intact surface would increase until failure.
The abstract is full of speculation and conjecture. A warming of subglacial permafrost is not really documented in this paper. Neither is a 'plasticization of basal ice' or the presence of a subglacial active layer.
The paper should also be carefully edited for grammar.
Detailed list of comments:
The title is a bit of a mouthful; consider simplifying
l.14: delete 'partially' (isolated portion and partially are redundant)
l.22: ... understand THE role ...
l.36/37: ice shelves are not collapsing due to acceleration and thinning of feeder glaciers, it's the other way around.
l.48: neo-formation of -> formation of new ..
l.48: is there really evidence for more subglacial lakes? I didn't look through the references carefully, but didn't see anything on a cursory look.
l.65-70: a lot of this is unclear and speculative ('probably played a primary role', 'presumably', ...). How does active layer thickness affect glacier stability? Or are you referring to a subglacial active layer?
l.92: 'could be outlined' -> 'is'
l.94 ; -> .
l.144: 'could be' -> 'was'
l.161: delete 'a'
l.164: The hypothesis about changing plasticity appears out of thin air here. Perhaps it should be stated as a hypothesis to explore; same for active layer.
Sec. 3.3: Satellite imagery and seismology is a bit of an odd combination for the same subsection
l.194: define 'GSD'
l.208: lately -> later
l.215: .. using a 0.05 m ...
l.216: define RES (or just use GPR throughout)
l.221: functional -> used
l.233: by -> from
l.234: using values from sea ice studies seems odd; it has quite different strength
l.237: 'could be probably considered': This kind of statement is very vague. Just state what exactly you assumed. Your borehole data provide justification for polythermal conditions.
l.238: Similar to what?
l.239/40: I am not convinced that PermaNet temperatures are directly indicative of polythermal glaciers. You actually explain this later; how firn refreeze processes can create warming conditions on glaciers that are not present on ice free ground and that are not incorporated in permafrost models.
l.247: A bit more detail on the model is warranted here. The Supplementary Materials don't help much. Say a bit more about the model, how it works, and what the differences are between the different versions.
Fig. 3 caption: .. for each OF the different ...
Fig. 4 and elsewhere: what is the difference between 'area' and 'surface'? Is one the map area and the other one actual surface area? If so, I would stick to map area. The actual surface area of a rough surface is actually not well defined and is scale dependent.
l.286: precipitations -> precipitation
Fig. 8: the grey shaded area for Marmolada seems large, given the altitude range of the glacier
l.337: after the mid of June -> after mid-June
l.339-345: several points are not clear here. You show temperature profiles for 3 July each year. These do not show the depth of the active layer. The active layer is the layer that changes from a frozen to a thawed state during the year. In mid-summer there is still part of the winter cold wave in the ground. I don't think a single profile can be used to determine depth of the active layer. Overall, this profile does not contribute much of relevance to the paper. It shows that at a lower elevation and different exposure there maybe be some permafrost, but how transferable is this to the Marmolada Glacier with a layer of glacial ice?
l.358/59: I'm having trouble identifying the 'traversel bediere' in the figure. This seems like a real oddity: how could a stream of water run in a transverse direction on a steep glacier?
l.381: 'Very little water evidences are' -> 'Very little evidence for water is'
l.382: delete 'somehow'
Fig. 11: I'm a bit skeptical about radar interpretations of bedrock under ice. These radar profiles are most likely influenced by out-of-plane reflections due to the shape of the glacieret. The v-shaped troughs indicate that they should probably be migrated. The interpreted lines in the bedrock look quite a bit like the tails of hyperbolae created by point reflectors. Migration would shed some light on that.
Fig. 12: what's the colorscale?
l.416: correct m3
l.418: could be -> was
l.421-24: can you conclude anything from these temperature measurements? This was imagery in the early morning and mostly reflects 'skin temperature', which would be very influenced by air temperature over the night.
l.425-427: the surface temperature cannot be used to say something about permafrost conditions.
Fig. 14: The seismic record is interesting. How well is the timing of the avalanche known? For example it looks like the seismic event records the initial failure (mostly in the horizontal components) and then the impact of the falling ice with both vertical and horizontal components about 20 sec later? Is this a reasonable interpretation?
l.460-464: I fail to see the relevance of this paragraph and Fig. 14B. First, the figure shows no obvious similarity between the two events (for example the Everest event had a much larger vertical component). A thorough analysis would require looking at spectra and perhaps some force-momentum modeling.
Fig 15 caption: what does 'with evidenced the lower transversal crevasses' mean?
There is a lot of overlap in Figs 15-18; these could be consolidated.
l.510: what do you mean by 'uneven combination'?
l.518-520: Here is where you state your main hypothesis for the thermal state, but this is only weakly backed up (see discussion on permafrost above).
l.526: snow precipitation is mentioned here, but it is not well motivated.
l.529-30: this is very much a hypothesis and conjecture and would require some thermal modeling to make a more definitive statement.
l.544: second tens of -> middle of
l.556/7: Clarify what you mean here when you talk of 'thermal inertia in the exposed ice body'.
l.570: pervious -> previous
l.577-85: Again, there is quite a bit of speculation here
l.629: Altel -> Altels
l.634: Finding temperate ice with RES is a fraught subject with often questionable conclusions.
l.664/5: Again, very speculative
l.673/4: ditto
l.711/2: this is an interesting observation and it needs some elaboration. It is not clear how a crevasse could penetrate to near the base in a cold glacier without the help of water. If the glacier is frozen to the bed, there are no longitudinal strain rates there, and failure under tension is not possible
l.740: something is missing here.The data statement reads odd. Every effort should be made to make data available publicly. That may not be possible with some of the proprietary imagery. But statements like 'data could be made available upon request' do definitely not meet modern open data requirements of most journals.
Martin Truffer
Citation: https://doi.org/10.5194/nhess-2024-212-RC1 -
RC2: 'Comment on nhess-2024-212', Anonymous Referee #2, 12 Jan 2025
First, I would like to commend the authors for compiling extensive data related to the tragic collapse of the Marmolada Glacier. They have gathered information from multiple sources to evaluate the factors that triggered this catastrophic event. This manuscript is undoubtedly valuable and has the potential for high citation and recognition within the scientific community.
However, before publication, the manuscript requires major revisions. This includes shortening and restructuring the content, as well as thoroughly checking for numerous redundancies. Additionally, the discussion of certain processes related to the glacier's thermal regime and permafrost needs re-assessment, as also addressed by M. Truffer in Review 1. While I started reading with eagerness to learn more about the event, I was somewhat disappointed because of obvious shortages. In my opinion, the manuscript is overly lengthy, the figures are not cited in order, and some are redundant and difficult to interpret. Below, I provide major and detailed comments. I have read M. Truffer's comments and fully support the points made, especailly regarding the glacier's thermal regime and permafrost as possible triggers for the landslide.
Abstract: Move the sentence starting at line 17 to the end, after listing the methods used, as it is repeated there. Also, clarify whether the active layer in the permafrost is below the glacier, we do not have any “active layers” below ice.
Introduction: This section is too long. Remove the initial list of collapses, including the ice shelf, as it is not relevant. Delete everything from line 65 onward related to “overall structure,” as these discussions belong later. The introduction should end by stating the objectives of the paper and mentioning key methods (e.g., climate observations, thermal regime, seismicity) used to achieve these goals.
General Setting: The field area is not adequately introduced. Provide information about the overall geological and structural setting, climate (temperature and precipitation), and permafrost limits in the area. In Figure 1, the text is too small, there's no scale in (A), and the 3D plot's scale is not readable. It would help to include a regular image of the area and images from the collapse.
Data, Methods, etc.: Remove the initial paragraph listing methods, as these should be detailed with manufacturer information, resolutions, and other relevant details. Line 159 contains an interpretation that repeats throughout the manuscript; avoid such redundancies. The calculation starting from line 169 is unclear; specify if GPR was used and the strategy for filling in missing meteorological data. Figures are cited out of order (e.g., Fig. 9 on line 189), so ensure they follow a sequence. Important installations could be visualized in Fig. 1 or 2; consider using a standard map layout instead of complex 3D plots. I found it challenging to follow the glacier stability and back analysis. The discussion of parameters is confusing and should be reserved for the discussion section—just present the parameters used along with references and justification. The ice temperature of -4°C seems low; is this value justified?
Results: What is meant by "overall evolution"? Call it "development" or just "evolution." It is not interesting here who conducted the measurement (l. 251), just provide the results. Increase the font size in Fig. 3, as it is hardly readable. Fig. 4c is difficult to understand, and I cannot read the figure. What is meant by the "centroid of the glacier front"? Low readability also applies to Fig. 5—text within the figure is not clear. Ensure figures are readable on paper. I gave up on Fig. 6 because, while it looks nice, it is not understandable.
Figure 6c presents a curve done for two cement blocks, not for the ice-bedrock interface. There's no certainty that the same relation is valid here; consider removing it. Furthermore, what is meant by "guessed thermal conditions"?
Line 346 and following contain interpretation, not results; this should be moved. Chapter 4.3. is difficult to follow. Figs. 9 and 10 are very challenging; consider using color and clear color ranges. These are certainly important figures, but difficult to follow. Fig. 11 is similar (especially parts A and D). What is RES? What does "very short wavelength" mean in line 400? Fig. 12 is very difficult to read. Line 425 should be revised, as surface temperature cannot be used to infer permafrost under the conditions mentioned.
This chapter about seismology looks good, including the figures.
Slope Stability Back Analysis: I found this section difficult to follow—perhaps it is my fault—but the figures (e.g., Fig. 15) are not user-friendly.
Discussion: There is much redundancy here. The first paragraph provides a conclusion rather than a discussion. Section 5.1. is repetitive (line 574). How do you know the thermal regime of the LIA glacier (line 565)? Some observations have been mentioned before and seem redundant. The section on "triggering factors" is very lengthy; consider creating subheadings like "seismic factors," "thermal factors," etc.
Conclusions: The conclusions contain much speculation and a few typos around line 740. I recommend using short conclusions with clear statements in bullet points rather than long text. You can omit the last paragraph, as it is not informative.
Citation: https://doi.org/10.5194/nhess-2024-212-RC2
Viewed
| HTML | XML | Total | Supplement | BibTeX | EndNote | |
|---|---|---|---|---|---|---|
| 285 | 71 | 14 | 370 | 35 | 8 | 8 |
- HTML: 285
- PDF: 71
- XML: 14
- Total: 370
- Supplement: 35
- BibTeX: 8
- EndNote: 8
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1