the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 09 Feb 2023
Monitoring Glacier Calving using Underwater Sound
Abstract. Climate shifts are particularly conspicuous in the Arctic. Satellite and terrestrial observations show significant increases in the melting and breakup of Arctic tidewater glaciers and their influence on sea level rise. Increasing melt rates are creating an urgency to better understand the link between atmospheric and oceanic conditions and glacier frontal ablation through iceberg calving and melting. Elucidating this link requires a combination of short and long-time scale measurements of terminus activity. Recent work has demonstrated the potential of using underwater sound to quantify the time and scale of calving events to yield integrated estimates of ice mass loss (Glowacki and Deane, 2020). Here, we present estimates of subaerial calving flux using underwater sound recorded at Hansbreen, Svalbard in September 2013 combined with an algorithm for the automatic detection of calving events. The method is compared with ice calving volumes estimated from geodetic measurements of the movement of the glacier terminus and an analysis of satellite images. The total volume of above-water calving during the 26 days of acoustical observation is estimated to be 1.7 ± 0.7 × 107 m3, whereas the subaerial calving flux estimated by traditional methods is 7 ± 2 × 106 m3. The results suggest that passive cryoacoustics is a viable technique for long-term monitoring of mass loss from marine-terminating glaciers.
-
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
(13243 KB)
-
The requested preprint has a corresponding peer-reviewed final revised paper. You are encouraged to refer to the final revised version.
- Preprint
(13243 KB) - Metadata XML
- BibTeX
- EndNote
- Final revised paper
Journal article(s) based on this preprint
Interactive discussion
Status: closed
-
RC1: 'Comment on egusphere-2023-115', Evgeny A. Podolskiy, 14 Mar 2023
The paper demonstrates how recent published efforts by some of the co-authors can be combined into a semi-automatic retrieval of calving flux from hydro-acoustic data. In appendices, the paper also discusses how many uncertainties and calibrations are still needed for using this method.
Overall, this is a well-written, clearly motivated and conveniently supported research. The key possible criticism about possible confusion of calving with free-floating iceberg disintegration has been partly addressed by the authors. Therefore, most of my remarks are rather minor/moderate, as listed below.
Lines 23, 32
it is my personal feeling, but I would keep “urgent” out of the manuscript.
Line 45
wind and precipitation examples need references.
Line 94
sound is {analyzed} into a spectrogram
-> transformed?
Line 128
can appear to be very large
->can be confused with very large?
Line 132
I suggest to explain what is alpha-stable distribution, since not many readers may follow.
Line 137
I would just mention that remote icebergs, say at the same distance as the calving front, but behind the receiver still remain an open problem.
Lines 191-203
I feel there is some incoherency (“oranges / apples") in comparison of sound-retrieved volume to geometric changes. Specifically,
- camera data was used for vertical area estimate, without conversion to volume.
- satellite data was used for area estimate with conversion to volume.
In both cases, one has to assume some weighting for going from 2D to 3D. For satellite data, this was cliff height measured in 2015. For camera data, a similar step was not attempted, while commonly used in papers, including those by the authors. Homogeneous comparison makes more sense to me, especially than satellite analysis is inevitably pretty rough and dGPS data are not relevant to the period of acoustic measurements.
Lines 204-217
There are quite many alternative hypothesis for discrepancy between acoustic and image-derived fluxes, due to uncertainty in velocity, calibration, and etc. Perhaps reducing this number might be possible by estimating volume from photos and revisiting ice velocity?
Furthermore, I suppose by 2015-laser measurements, the glacier thinned and thus the ice cliff (i.e. cross-section) was lower than in 2013, which could lead to an underestimate of calving flux and could be checked.
Data availability: (sound, dGPS, etc)
With due respect to the authors, “upon request” is not in line with open data policy of TC. Also, from personal experience, I got no replies to my two last requests for data to authors using similar statements or saw that email addresses became obsolete.
Considering that the amount of sound data from polar regions is expected to increase, the code used for this paper might be helpful for the community and would increase citations to this work, so I encourage the authors to keep such code alive, say at github.
Appendix A1
Line 290
to me, there is insufficient amount of detail, because it is not clear what is “a difference technique” and what exactly is compared (greyscale intensity?). Please elaborate because retrieval of area from oblique images is not a trivial task and the reader has little idea how to reproduce this. Please also see my comment on area vs. volume.
I could not follow Appendix A2
Line 299
Q = U_i + U_f*L*H
= m3/d ? please check units
Lines 299, 308
. subaerial calving flux
. Subaerial …
Line 305
“precise dGPS in August 2013”
Please indicate the exact dates, because as I understand, they were outside the hydro-acoustic monitoring period and their “precise” nature was of arguable help?
Line 306
Furthermore, I was puzzled by “a cross-check” of velocity field by using December 2012 imagery, which tells us little about dynamics in the end of summer. Why December? As I understand, there are plenty of Glacier Image Velocimetry open-source tools to retrieve velocity field at good temporal resolution (e.g., https://doi.org/10.5194/tc-15-2115-2021)
Line 312
Please note that “stake measurements” were not shown and, if I am not missing something, the mean daily ice flow velocity was also not mentioned (Section 6). Considering high sensitivity of satellite-derived flux estimates to this value, this detail is important. It alone could explain the discrepancy between the acoustic and satellite derived fluxes.
Line 326
please show units for grain size
Evgeny Podolskiy
Citation: https://doi.org/10.5194/egusphere-2023-115-RC1 - AC1: 'Reply on RC1', Oskar Glowacki, 23 May 2023
-
RC2: 'Comment on egusphere-2023-115', Anonymous Referee #2, 24 Apr 2023
This paper presents results from a field campaign to monitor calving fluxes at Hansbreen, Svalbard, with particular emphasis on the use of underwater acoustic records for this purpose. The results have strong applicability well beyond the selected field site, and the case is well made for the usefulness of such records over large spatial areas in polar/glaciated regions. Of particular note is the the development and presentation here of an automated algorithm that enables detecting and quantifying subaerial calving events and their fluxes; this is a distinct advance, since it removes the previous bottleneck associated with human effort to derive these. The algorithm is, by the authors' own assessment, not yet perfect, but it is certainly useful and publication will stimulate further research effort to refine it further.
The research is well described and important, the conclusions are suitably justified, and the presentation of the manuscript is clear and professional both for written and graphical aspects. I recommend publication subject to a few minor revisions.Minor comments:-
The abstract and introduction present information that sets the scene and justifies the interest in glacial loss, relying quite heavily on sea level rise aspects for this. This is fair enough, but there are also dynamical implications from calving glaciers that add to such drivers, and which are separate from sea level rise considerations. These include ocean mixing, and things that depend on that (e.g. productivity, climate, sea ice production). Can a few sentences of text be included to explain this importance?
The abstract and Introduction discuss the Arctic and Greenland in the context of glacial change, with good justification. Some other regions also matter in this context, e.g. Antarctic Peninsula, Patagonia etc. Perhaps mentioning these would broaden the applicability even further in the readers' mind?
Line 16. RCP8.5 is looking very unlikely as a trajectory that we are likely to follow. Perhaps include numbers for e.g. RCP4.5 instead, if available?Line 25. "plunged into darkness" sounds a bit dramatic - perhaps just "subject to darkness"?
Line 63. Figure 1 is not really a schematic, but instead a map and a photograph (both of which are good).
Line 71. Salinity, as measured, is a ratio and hence is dimensionless. While one sees "PSU" quite often, it is not actually correct. Correct terminology would be "... a salinity of 30 on the practical salinity scale". Should probably quote salinity to at least one decimal place also. Perhaps include a figure of the temperature/salinity data?Line 105 and about. It would be worth some text explaining the extent to which the choice of constants and thresholds are likely specific to the location under study - this has relevance when considering how to apply this technique at other locations.
Line 150. It seems to me that a big next step, if possible, would be automated algorithm for submarine calving detection and flux calculation. What is needed for that to be developed?
Line 177. Units are here written as "decibels" and were "dB" previously - I personally dont mind which, but need to be consistent.
Line 232. This is a good idea, but how would the glaciers be selected? It is not possible to monitor all - can the most representative be chosen somehow? What would be the criteria for this?
Line 234. It is a good idea to monitor concurrent environmental drivers; surely it would be a good idea to also monitor current environmental impacts, e.g. ocean stratification etc?
Line 259. I'm sure this method has applicability elsewhere and on larger spatial scales, as stated in the text. I would presume that this method has great utility for the calving of grounded marine-terminating glaciers, but that calving of ice shelves or floating ice tongues would be different, since these would be more similar to detachment of an already-floating section of ice without necessarily the same impact on the ocean? Might be worth making this explicit if so, to avoid confusion.
Appendix. Is it worth detailing the CTD methodology, a little?
Citation: https://doi.org/10.5194/egusphere-2023-115-RC2 -
AC2: 'Reply on RC2', Oskar Glowacki, 23 May 2023
Dear Reviewer,
thank you for your suggestions for possible improvements. We will incorporate all of them in the revised version of the manuscript.
Citation: https://doi.org/10.5194/egusphere-2023-115-AC2
-
AC2: 'Reply on RC2', Oskar Glowacki, 23 May 2023
Interactive discussion
Status: closed
-
RC1: 'Comment on egusphere-2023-115', Evgeny A. Podolskiy, 14 Mar 2023
The paper demonstrates how recent published efforts by some of the co-authors can be combined into a semi-automatic retrieval of calving flux from hydro-acoustic data. In appendices, the paper also discusses how many uncertainties and calibrations are still needed for using this method.
Overall, this is a well-written, clearly motivated and conveniently supported research. The key possible criticism about possible confusion of calving with free-floating iceberg disintegration has been partly addressed by the authors. Therefore, most of my remarks are rather minor/moderate, as listed below.
Lines 23, 32
it is my personal feeling, but I would keep “urgent” out of the manuscript.
Line 45
wind and precipitation examples need references.
Line 94
sound is {analyzed} into a spectrogram
-> transformed?
Line 128
can appear to be very large
->can be confused with very large?
Line 132
I suggest to explain what is alpha-stable distribution, since not many readers may follow.
Line 137
I would just mention that remote icebergs, say at the same distance as the calving front, but behind the receiver still remain an open problem.
Lines 191-203
I feel there is some incoherency (“oranges / apples") in comparison of sound-retrieved volume to geometric changes. Specifically,
- camera data was used for vertical area estimate, without conversion to volume.
- satellite data was used for area estimate with conversion to volume.
In both cases, one has to assume some weighting for going from 2D to 3D. For satellite data, this was cliff height measured in 2015. For camera data, a similar step was not attempted, while commonly used in papers, including those by the authors. Homogeneous comparison makes more sense to me, especially than satellite analysis is inevitably pretty rough and dGPS data are not relevant to the period of acoustic measurements.
Lines 204-217
There are quite many alternative hypothesis for discrepancy between acoustic and image-derived fluxes, due to uncertainty in velocity, calibration, and etc. Perhaps reducing this number might be possible by estimating volume from photos and revisiting ice velocity?
Furthermore, I suppose by 2015-laser measurements, the glacier thinned and thus the ice cliff (i.e. cross-section) was lower than in 2013, which could lead to an underestimate of calving flux and could be checked.
Data availability: (sound, dGPS, etc)
With due respect to the authors, “upon request” is not in line with open data policy of TC. Also, from personal experience, I got no replies to my two last requests for data to authors using similar statements or saw that email addresses became obsolete.
Considering that the amount of sound data from polar regions is expected to increase, the code used for this paper might be helpful for the community and would increase citations to this work, so I encourage the authors to keep such code alive, say at github.
Appendix A1
Line 290
to me, there is insufficient amount of detail, because it is not clear what is “a difference technique” and what exactly is compared (greyscale intensity?). Please elaborate because retrieval of area from oblique images is not a trivial task and the reader has little idea how to reproduce this. Please also see my comment on area vs. volume.
I could not follow Appendix A2
Line 299
Q = U_i + U_f*L*H
= m3/d ? please check units
Lines 299, 308
. subaerial calving flux
. Subaerial …
Line 305
“precise dGPS in August 2013”
Please indicate the exact dates, because as I understand, they were outside the hydro-acoustic monitoring period and their “precise” nature was of arguable help?
Line 306
Furthermore, I was puzzled by “a cross-check” of velocity field by using December 2012 imagery, which tells us little about dynamics in the end of summer. Why December? As I understand, there are plenty of Glacier Image Velocimetry open-source tools to retrieve velocity field at good temporal resolution (e.g., https://doi.org/10.5194/tc-15-2115-2021)
Line 312
Please note that “stake measurements” were not shown and, if I am not missing something, the mean daily ice flow velocity was also not mentioned (Section 6). Considering high sensitivity of satellite-derived flux estimates to this value, this detail is important. It alone could explain the discrepancy between the acoustic and satellite derived fluxes.
Line 326
please show units for grain size
Evgeny Podolskiy
Citation: https://doi.org/10.5194/egusphere-2023-115-RC1 - AC1: 'Reply on RC1', Oskar Glowacki, 23 May 2023
-
RC2: 'Comment on egusphere-2023-115', Anonymous Referee #2, 24 Apr 2023
This paper presents results from a field campaign to monitor calving fluxes at Hansbreen, Svalbard, with particular emphasis on the use of underwater acoustic records for this purpose. The results have strong applicability well beyond the selected field site, and the case is well made for the usefulness of such records over large spatial areas in polar/glaciated regions. Of particular note is the the development and presentation here of an automated algorithm that enables detecting and quantifying subaerial calving events and their fluxes; this is a distinct advance, since it removes the previous bottleneck associated with human effort to derive these. The algorithm is, by the authors' own assessment, not yet perfect, but it is certainly useful and publication will stimulate further research effort to refine it further.
The research is well described and important, the conclusions are suitably justified, and the presentation of the manuscript is clear and professional both for written and graphical aspects. I recommend publication subject to a few minor revisions.Minor comments:-
The abstract and introduction present information that sets the scene and justifies the interest in glacial loss, relying quite heavily on sea level rise aspects for this. This is fair enough, but there are also dynamical implications from calving glaciers that add to such drivers, and which are separate from sea level rise considerations. These include ocean mixing, and things that depend on that (e.g. productivity, climate, sea ice production). Can a few sentences of text be included to explain this importance?
The abstract and Introduction discuss the Arctic and Greenland in the context of glacial change, with good justification. Some other regions also matter in this context, e.g. Antarctic Peninsula, Patagonia etc. Perhaps mentioning these would broaden the applicability even further in the readers' mind?
Line 16. RCP8.5 is looking very unlikely as a trajectory that we are likely to follow. Perhaps include numbers for e.g. RCP4.5 instead, if available?Line 25. "plunged into darkness" sounds a bit dramatic - perhaps just "subject to darkness"?
Line 63. Figure 1 is not really a schematic, but instead a map and a photograph (both of which are good).
Line 71. Salinity, as measured, is a ratio and hence is dimensionless. While one sees "PSU" quite often, it is not actually correct. Correct terminology would be "... a salinity of 30 on the practical salinity scale". Should probably quote salinity to at least one decimal place also. Perhaps include a figure of the temperature/salinity data?Line 105 and about. It would be worth some text explaining the extent to which the choice of constants and thresholds are likely specific to the location under study - this has relevance when considering how to apply this technique at other locations.
Line 150. It seems to me that a big next step, if possible, would be automated algorithm for submarine calving detection and flux calculation. What is needed for that to be developed?
Line 177. Units are here written as "decibels" and were "dB" previously - I personally dont mind which, but need to be consistent.
Line 232. This is a good idea, but how would the glaciers be selected? It is not possible to monitor all - can the most representative be chosen somehow? What would be the criteria for this?
Line 234. It is a good idea to monitor concurrent environmental drivers; surely it would be a good idea to also monitor current environmental impacts, e.g. ocean stratification etc?
Line 259. I'm sure this method has applicability elsewhere and on larger spatial scales, as stated in the text. I would presume that this method has great utility for the calving of grounded marine-terminating glaciers, but that calving of ice shelves or floating ice tongues would be different, since these would be more similar to detachment of an already-floating section of ice without necessarily the same impact on the ocean? Might be worth making this explicit if so, to avoid confusion.
Appendix. Is it worth detailing the CTD methodology, a little?
Citation: https://doi.org/10.5194/egusphere-2023-115-RC2 -
AC2: 'Reply on RC2', Oskar Glowacki, 23 May 2023
Dear Reviewer,
thank you for your suggestions for possible improvements. We will incorporate all of them in the revised version of the manuscript.
Citation: https://doi.org/10.5194/egusphere-2023-115-AC2
-
AC2: 'Reply on RC2', Oskar Glowacki, 23 May 2023
Peer review completion
Journal article(s) based on this preprint
Viewed
| HTML | XML | Total | BibTeX | EndNote | |
|---|---|---|---|---|---|
| 401 | 159 | 24 | 584 | 9 | 10 |
- HTML: 401
- PDF: 159
- XML: 24
- Total: 584
- BibTeX: 9
- EndNote: 10
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1
Jarosław Tęgowski
Michał Ciepły
Małgorzata Błaszczyk
Jacek Jania
Mateusz Moskalik
Philippe Blondel
Grant B. Deane
The requested preprint has a corresponding peer-reviewed final revised paper. You are encouraged to refer to the final revised version.
- Preprint
(13243 KB) - Metadata XML