the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 07 May 2024
Spectral characteristics of seismic ambient vibrations reveal subglacial hydraulic changes beneath Glacier de la Plaine Morte, Switzerland
Abstract. Glaciers have a complex hydraulic and dynamic behavior that needs to be investigated to improve our understanding of changes in the cryosphere. To tackle this issue, we employ various passive seismic analysis methods on continuous measurements from a temporary seismic array deployed on Glacier de la Plaine Morte in Switzerland. First, we asses the reliability of ambient noise Horizontal-to-Vertical spectral ratio (HVSR) measurements to the glacier's dynamic environment. The spatiotemporal variations in HVSR curves are predominantly attributed to changing nearby noise conditions influenced by hydraulic, drainage-related tremors, moulin resonances and anthropogenic sources. A careful analysis of the local noise source variations related to glacier dynamic behaviour in order to distinguish between source and medium changes reflected in the HVSR measurements. Only a few hours of HVSR measurements may lead to biases in the interpretation of the HVSR curve. Despite the influence of these external factors, with long time series of the HVSR measurements, we successfully detect a spatiotemporal trend in HVSR curves. Notably, an HVSR trough emerges following the drainage of Lac des Faverges, an ice-marginal lake that rapidly drains, causing water to flow through a channelized system beneath the glacier. This HVSR trough is indicative of a low-seismic velocity layer at the ice-bed interface. Seismic velocity changes derived by interferometry support the presence of a low-velocity layer at the ice-bedrock interface. Inversion and forward modelling reveal a probable thickness of this low-velocity layer of 10–30 m and a change in S-wave velocity up to 40 %. This layer has a local extend covering an estimated 4.5 to 27 % of the glacier, as indicated by the spatial variations in HVSR trough throughout the array and an independent water volume estimate. The changing seismic velocities are thus a manifestation of temporal water storage at the glacier bed in response to sudden injection of lake water. Our results highlight the value of long time series of HVSR measurements which show variations in the peak/trough structure that reflect hydraulic changes beneath the ice surface.
-
Notice on discussion status
The requested preprint has a corresponding peer-reviewed final revised paper. You are encouraged to refer to the final revised version.
-
Preprint
(5471 KB)
-
The requested preprint has a corresponding peer-reviewed final revised paper. You are encouraged to refer to the final revised version.
- Preprint
(5471 KB) - Metadata XML
- BibTeX
- EndNote
- Final revised paper
Journal article(s) based on this preprint
Interactive discussion
Status: closed
-
RC1: 'Comment on egusphere-2024-646', Andreas Köhler, 21 Jun 2024
In this study the authors use a passive seismic method to investigate spatial-temporal changes at the bed of an Alpine glacier. While the H/V method is well-established for passive seismic near-surface investigation in geological settings with soft sediments over bedrock, applications in the cryosphere are still rare and challenging due to the nature of the noise-wavefield. Several studies have shown the method’s potential for glacier, ice sheet and permafrost applications. I appreciate further studies on this topic and therefore believe that the current manuscript is an excellent contribution to establish H/V in cryoseismology. I acknowledge from my own experience that noise conditions make the interpretation of H/V observations in the cryosphere challenging, especially when the objective is to resolve temporal variations in the sub-surface structure from H/V. What make this study particularly valuable is the combination with other seismological analysis taking advantage of seismic array measurements. I agree absolutely with the authors findings and recommendations that it is important and beneficial to use array records on glaciers which allow performing noise interferometry as well as beamforming for noise source characterisation and surface wave dispersion analysis. Overall, from a seismological perspective, there is nothing significant to revise in my opinion. I have only a few technical corrections and minor questions to be addressed. I’m not a glaciologist, so the validity of the interpretations regarding the glacier hydraulic system are a bit out of my expertise.
Line 7-9: something is missing here. “... is done in order ... “?
Line 47: “significant contribution of Rayleigh waves” This is most likely true for glaciers with mostly natural sources I guess but note that Love waves have been shown to have a significant contribution in urban areas with anthropogenic sources. Also, body waves may affect HVSRs at lower frequencies. I suggest to briefly mention noise wavefield composition and its effect on HVSR here.
Line 57: one “active” too much
Line 93-94: Just for clarification: Do you have an idea what controls the hydraulic tremors indicated in Fig 2c, day 237-240 before the drainage?
Line 94: and (?) between references
Line 125: It is of course very appropriate to use this well-known formula for a depth estimate. But if I'm not mistaken, this originates from theoretical considerations if SH wave resonance is responsible for the H/V peak. There are studies which showed that the peak frequencies from Rayleigh wave elliptticity peak and SH wave resonance are very close, so this is no issue as such. I mention this because of the wavefield composition mentioned above (Rayleigh waves).
Figure 5: It seems the through starts appearing already around day 228. Not very clear though, but could this be due to melt water presence?
236 ff: The second inversion seems to result indeed in a slightly better fit, but I'm not sure if I would call this a clear improvement visually. Can you quantify the improved misfit?
Line 261ff: I agree that inversion and modelling results support the low-velocity hypothesis. However, the fits in Fig 9 are not very good. Can you speculate a bit what could be the reason why a more complex model is required in your glaciological setting? Additional layers? 2D/3D effects?
Line 296-298: Could then pressurised sediments be an explanation for the misfit? See my comment above.
Line 376: Remove «For»
Line 421: remove brackets around references
Citation: https://doi.org/10.5194/egusphere-2024-646-RC1 -
AC2: 'Reply on RC1', Janneke van Ginkel, 26 Jul 2024
Dear Andreas Köhler,
Thank you for your positive remarks and insightful comments on the paper. We appreciate the time and effort that you have dedicated to our manuscript. We have discussed your main technical suggestions and the supplement to this response summarizes the outcome. The small editorial suggestions will be also incorporated in the revised manuscript, which will be uploaded at a later stage.
We hope we cover your comments and are willing to respond to any further questions and suggestions you may have.
Sincerely,
Janneke van Ginkel, Fabian Walter, Fabian Lindner, Miroslav Hallo, Matthias Huss and Donat Fäh
-
AC2: 'Reply on RC1', Janneke van Ginkel, 26 Jul 2024
-
RC2: 'Comment on egusphere-2024-646', Florent Gimbert, 24 Jun 2024
Review of « Spectral characteristics of seismic ambient vibrations reveal subglacial hydraulic changes beneath Glacier de la Plaine Morte, Switzerland” by Janneke van Ginkel et al., submitted to the Cryosphere.
SUMMARY
The authors present a comprehensive study of the effect lake drainage on glacier hydrology observed from seismic data and in particular H/V analysis on Glacier de la Plaine Morte, Switzerland. They report a change in H/V spectral characteristics from before to after lake drainage, specifically the appearance of a trough around 4-5 Hz in the H/V spectra. They attribute this spectral change to water being stored at the ice bed interface. They support this interpretation based on several other lines of evidence, such that a change in surface wave velocity at these frequencies and an estimate of the stored volume based on an observed shift in the melt versus outlet discharge relationship. They suggest the H/V ratio may be used as a simple single-station method to identify hydrology changes beneath glaciers.
MAIN COMMENTS
I find the authors contribution to be sound and promising for future applications. The application of H/V analysis has been applied in the glaciology literature, but to my knowledge mostly to estimate glacier thicknesses. Although it was known based on numerous studies in other contexts (e.g. sedimentary basins) that more information could in principle be retrieved based on H/V spectra, the present study provides the first evidence that H/V can show drastic changes through time in response to temporal changes in glacier hydrology. The observations is particularly convincing, especially in light of the complementary analysis the authors conduct to check that the temporal changes in H/V spectra indeed reflect internal changes, verifying the contribution of heterogenous noise sources, calculating and showing associated changes in surface wave dispersion curves, and conducting inversions to attribute the changes to structural changes. The H/V technique being simple to conduct and applicable on single stations, I do find that it has strong power for future applications in a range of array designs. I also particularly appreciate the efforts the authors have been doing at checking their observations with glaciological variables such as discharges, melt rate estimates, etc… Finally, the paper is overall well written, the structure logical and the figures appropriate.
For all the above reasons, I recommend this paper for publication. However, I do have one major point that I would like the authors to address before considering it further for publication, as well as few other minor points.
MAJOR POINT
Despite the comprehensive effort the authors put in attributing the observed H/V and dispersion curve changes to subsurface hydrology changes, it appears as still unclear to me what these changes correspond to. The hydrology suggestion is very sound based on all the different lines of evidence, but none of these lines clearly point towards an identified storage source. The joint H/V – surface wave dispersion curve inversions indicate the low velocity layer might be englacial (between 100-150 m deep), but confidence is low and inversions with or without low velocity layer actually both reproduce the saddle in H/V spectra. The forward modelling done with a water layer at the ice-bed interface convincingly reproduces the saddle in the H/V spectra, but does not reproduce the dispersion curves properly. Finally, the water volume estimates convincingly show water is missing at the outlet, but this could also come from water being routed elsewhere (such as in the ground) rather than stored englacially or subglacially.
Although the authors acknowledge this in the text, they do claim in several instances, such as, most importantly, in the abstract (line 16-17), that the H/V changes reflect changes in water storage at the ice bed interface. I don’t think this claim is convincingly enough supported by their observations. Either the authors provide more convincing evidence that this is the case, or they need to soften their statements throughout, e.g. by saying hydrological changes below the glacier surface, either englacial, subglacial or underground. Actually, even the title is a bit misleading, because it implies that changes occur below the glacier (implicitly at the ice bed interface), while results suggest it might as well be englacial or in the ground.
MINOR POINTS:
- Lines 7-9: I don’t understand the sentence.
- Lines 102: seems like a lot of past H/V analysis involve removing impulsive events such as icequakes from the catalog. The authors do no do this nor mention it. Could you clarify why ?
- Line 114: statement about tremors is hard to connect to Fig 4. Needs more explanation.
- Figure 3: would be nice to show the frequencies outlines in Table 1 in the figure. Also, Table 1 should have a caption.
- Figure 4: would be nice to show temperature timeries on this figure, as a panel on top.
- Line 158: the trough seems to be existing, although of course less pronounced, before the lake drainage. In fact it seems to gradually increase before the lake drainage, and I wonder whether that is significant or not, and related to surface melt ? It would be nice to comment more on this.
- Line 192: scattered waves ? really ? these waves are probably not much scattered since observed at times quite close to the direct wave.
- Line 199: do you use the stretching method ? please specify.
- Line 241: the low velocity layer is hard to see. Maybe you could highlight it by a circle on the figure ? It is actually quite low confidence, thus you might want to specify this in the text.
- Line 243: I don’t see in which ways the dispersion curve predicted with the low velocity layer shows a better fit than without the low velocity layer. The fits appear as equivalent to me, which makes the argument weak. Either the authors explain this better or remove and acknowledge the poorly constrained nature of inversions.
- Figure 8: the inversion find a Poissons ratio of 0.5. Is that realistic ? I would rather think 0.3 is a realistic value, while 0.5 would really correspond only to water ? Isn’t there a problem there ? The grey shaded area is also not specified. Does that correspond to bedrock ? known from what, radar ?
- Line 240 : improve wording, we don’t fit a day.
- Line 265: Convincing from the forward modelling, not as much from the inversion. The authors should acknowledge this, as well as clearly highlighting that the inversion can reproduce the saddle without low velocity layer. They should also explain what makes their inversion able to explain the saddle without accounting for a low velocity layer.
- Figure 10: how about the water being routed underground ? I seem to remember Plaine Morte being a karstic environment, favorable for groundwater drainage that would drain water from the lake elsewhere than the main outlet ?
Citation: https://doi.org/10.5194/egusphere-2024-646-RC2 -
AC1: 'Reply on RC2', Janneke van Ginkel, 26 Jul 2024
Dear Florent Gimbert
Thank you for your positive remarks and insightful comments on the paper. We appreciate the time and effort that you have dedicated to our manuscript. We have discussed your main technical suggestions and the supplement to this response summarizes the outcome. The small editorial suggestions will be also incorporated in the revised manuscript, which will be uploaded at a later stage.
We hope we cover your comments and are willing to respond to any further questions and suggestions you may have.
Sincerely,
Janneke van Ginkel, Fabian Walter, Fabian Lindner, Miroslav Hallo, Matthias Huss and Donat Fäh
Sincerely,
Interactive discussion
Status: closed
-
RC1: 'Comment on egusphere-2024-646', Andreas Köhler, 21 Jun 2024
In this study the authors use a passive seismic method to investigate spatial-temporal changes at the bed of an Alpine glacier. While the H/V method is well-established for passive seismic near-surface investigation in geological settings with soft sediments over bedrock, applications in the cryosphere are still rare and challenging due to the nature of the noise-wavefield. Several studies have shown the method’s potential for glacier, ice sheet and permafrost applications. I appreciate further studies on this topic and therefore believe that the current manuscript is an excellent contribution to establish H/V in cryoseismology. I acknowledge from my own experience that noise conditions make the interpretation of H/V observations in the cryosphere challenging, especially when the objective is to resolve temporal variations in the sub-surface structure from H/V. What make this study particularly valuable is the combination with other seismological analysis taking advantage of seismic array measurements. I agree absolutely with the authors findings and recommendations that it is important and beneficial to use array records on glaciers which allow performing noise interferometry as well as beamforming for noise source characterisation and surface wave dispersion analysis. Overall, from a seismological perspective, there is nothing significant to revise in my opinion. I have only a few technical corrections and minor questions to be addressed. I’m not a glaciologist, so the validity of the interpretations regarding the glacier hydraulic system are a bit out of my expertise.
Line 7-9: something is missing here. “... is done in order ... “?
Line 47: “significant contribution of Rayleigh waves” This is most likely true for glaciers with mostly natural sources I guess but note that Love waves have been shown to have a significant contribution in urban areas with anthropogenic sources. Also, body waves may affect HVSRs at lower frequencies. I suggest to briefly mention noise wavefield composition and its effect on HVSR here.
Line 57: one “active” too much
Line 93-94: Just for clarification: Do you have an idea what controls the hydraulic tremors indicated in Fig 2c, day 237-240 before the drainage?
Line 94: and (?) between references
Line 125: It is of course very appropriate to use this well-known formula for a depth estimate. But if I'm not mistaken, this originates from theoretical considerations if SH wave resonance is responsible for the H/V peak. There are studies which showed that the peak frequencies from Rayleigh wave elliptticity peak and SH wave resonance are very close, so this is no issue as such. I mention this because of the wavefield composition mentioned above (Rayleigh waves).
Figure 5: It seems the through starts appearing already around day 228. Not very clear though, but could this be due to melt water presence?
236 ff: The second inversion seems to result indeed in a slightly better fit, but I'm not sure if I would call this a clear improvement visually. Can you quantify the improved misfit?
Line 261ff: I agree that inversion and modelling results support the low-velocity hypothesis. However, the fits in Fig 9 are not very good. Can you speculate a bit what could be the reason why a more complex model is required in your glaciological setting? Additional layers? 2D/3D effects?
Line 296-298: Could then pressurised sediments be an explanation for the misfit? See my comment above.
Line 376: Remove «For»
Line 421: remove brackets around references
Citation: https://doi.org/10.5194/egusphere-2024-646-RC1 -
AC2: 'Reply on RC1', Janneke van Ginkel, 26 Jul 2024
Dear Andreas Köhler,
Thank you for your positive remarks and insightful comments on the paper. We appreciate the time and effort that you have dedicated to our manuscript. We have discussed your main technical suggestions and the supplement to this response summarizes the outcome. The small editorial suggestions will be also incorporated in the revised manuscript, which will be uploaded at a later stage.
We hope we cover your comments and are willing to respond to any further questions and suggestions you may have.
Sincerely,
Janneke van Ginkel, Fabian Walter, Fabian Lindner, Miroslav Hallo, Matthias Huss and Donat Fäh
-
AC2: 'Reply on RC1', Janneke van Ginkel, 26 Jul 2024
-
RC2: 'Comment on egusphere-2024-646', Florent Gimbert, 24 Jun 2024
Review of « Spectral characteristics of seismic ambient vibrations reveal subglacial hydraulic changes beneath Glacier de la Plaine Morte, Switzerland” by Janneke van Ginkel et al., submitted to the Cryosphere.
SUMMARY
The authors present a comprehensive study of the effect lake drainage on glacier hydrology observed from seismic data and in particular H/V analysis on Glacier de la Plaine Morte, Switzerland. They report a change in H/V spectral characteristics from before to after lake drainage, specifically the appearance of a trough around 4-5 Hz in the H/V spectra. They attribute this spectral change to water being stored at the ice bed interface. They support this interpretation based on several other lines of evidence, such that a change in surface wave velocity at these frequencies and an estimate of the stored volume based on an observed shift in the melt versus outlet discharge relationship. They suggest the H/V ratio may be used as a simple single-station method to identify hydrology changes beneath glaciers.
MAIN COMMENTS
I find the authors contribution to be sound and promising for future applications. The application of H/V analysis has been applied in the glaciology literature, but to my knowledge mostly to estimate glacier thicknesses. Although it was known based on numerous studies in other contexts (e.g. sedimentary basins) that more information could in principle be retrieved based on H/V spectra, the present study provides the first evidence that H/V can show drastic changes through time in response to temporal changes in glacier hydrology. The observations is particularly convincing, especially in light of the complementary analysis the authors conduct to check that the temporal changes in H/V spectra indeed reflect internal changes, verifying the contribution of heterogenous noise sources, calculating and showing associated changes in surface wave dispersion curves, and conducting inversions to attribute the changes to structural changes. The H/V technique being simple to conduct and applicable on single stations, I do find that it has strong power for future applications in a range of array designs. I also particularly appreciate the efforts the authors have been doing at checking their observations with glaciological variables such as discharges, melt rate estimates, etc… Finally, the paper is overall well written, the structure logical and the figures appropriate.
For all the above reasons, I recommend this paper for publication. However, I do have one major point that I would like the authors to address before considering it further for publication, as well as few other minor points.
MAJOR POINT
Despite the comprehensive effort the authors put in attributing the observed H/V and dispersion curve changes to subsurface hydrology changes, it appears as still unclear to me what these changes correspond to. The hydrology suggestion is very sound based on all the different lines of evidence, but none of these lines clearly point towards an identified storage source. The joint H/V – surface wave dispersion curve inversions indicate the low velocity layer might be englacial (between 100-150 m deep), but confidence is low and inversions with or without low velocity layer actually both reproduce the saddle in H/V spectra. The forward modelling done with a water layer at the ice-bed interface convincingly reproduces the saddle in the H/V spectra, but does not reproduce the dispersion curves properly. Finally, the water volume estimates convincingly show water is missing at the outlet, but this could also come from water being routed elsewhere (such as in the ground) rather than stored englacially or subglacially.
Although the authors acknowledge this in the text, they do claim in several instances, such as, most importantly, in the abstract (line 16-17), that the H/V changes reflect changes in water storage at the ice bed interface. I don’t think this claim is convincingly enough supported by their observations. Either the authors provide more convincing evidence that this is the case, or they need to soften their statements throughout, e.g. by saying hydrological changes below the glacier surface, either englacial, subglacial or underground. Actually, even the title is a bit misleading, because it implies that changes occur below the glacier (implicitly at the ice bed interface), while results suggest it might as well be englacial or in the ground.
MINOR POINTS:
- Lines 7-9: I don’t understand the sentence.
- Lines 102: seems like a lot of past H/V analysis involve removing impulsive events such as icequakes from the catalog. The authors do no do this nor mention it. Could you clarify why ?
- Line 114: statement about tremors is hard to connect to Fig 4. Needs more explanation.
- Figure 3: would be nice to show the frequencies outlines in Table 1 in the figure. Also, Table 1 should have a caption.
- Figure 4: would be nice to show temperature timeries on this figure, as a panel on top.
- Line 158: the trough seems to be existing, although of course less pronounced, before the lake drainage. In fact it seems to gradually increase before the lake drainage, and I wonder whether that is significant or not, and related to surface melt ? It would be nice to comment more on this.
- Line 192: scattered waves ? really ? these waves are probably not much scattered since observed at times quite close to the direct wave.
- Line 199: do you use the stretching method ? please specify.
- Line 241: the low velocity layer is hard to see. Maybe you could highlight it by a circle on the figure ? It is actually quite low confidence, thus you might want to specify this in the text.
- Line 243: I don’t see in which ways the dispersion curve predicted with the low velocity layer shows a better fit than without the low velocity layer. The fits appear as equivalent to me, which makes the argument weak. Either the authors explain this better or remove and acknowledge the poorly constrained nature of inversions.
- Figure 8: the inversion find a Poissons ratio of 0.5. Is that realistic ? I would rather think 0.3 is a realistic value, while 0.5 would really correspond only to water ? Isn’t there a problem there ? The grey shaded area is also not specified. Does that correspond to bedrock ? known from what, radar ?
- Line 240 : improve wording, we don’t fit a day.
- Line 265: Convincing from the forward modelling, not as much from the inversion. The authors should acknowledge this, as well as clearly highlighting that the inversion can reproduce the saddle without low velocity layer. They should also explain what makes their inversion able to explain the saddle without accounting for a low velocity layer.
- Figure 10: how about the water being routed underground ? I seem to remember Plaine Morte being a karstic environment, favorable for groundwater drainage that would drain water from the lake elsewhere than the main outlet ?
Citation: https://doi.org/10.5194/egusphere-2024-646-RC2 -
AC1: 'Reply on RC2', Janneke van Ginkel, 26 Jul 2024
Dear Florent Gimbert
Thank you for your positive remarks and insightful comments on the paper. We appreciate the time and effort that you have dedicated to our manuscript. We have discussed your main technical suggestions and the supplement to this response summarizes the outcome. The small editorial suggestions will be also incorporated in the revised manuscript, which will be uploaded at a later stage.
We hope we cover your comments and are willing to respond to any further questions and suggestions you may have.
Sincerely,
Janneke van Ginkel, Fabian Walter, Fabian Lindner, Miroslav Hallo, Matthias Huss and Donat Fäh
Sincerely,
Peer review completion
Journal article(s) based on this preprint
Viewed
| HTML | XML | Total | BibTeX | EndNote | |
|---|---|---|---|---|---|
| 386 | 214 | 35 | 635 | 26 | 37 |
- HTML: 386
- PDF: 214
- XML: 35
- Total: 635
- BibTeX: 26
- EndNote: 37
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1
Janneke van Ginkel
Fabian Walter
Fabian Lindner
Miroslav Hallo
Matthias Huss
Donat Fäh
The requested preprint has a corresponding peer-reviewed final revised paper. You are encouraged to refer to the final revised version.
- Preprint
(5471 KB) - Metadata XML