the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 29 Jan 2024
Geomorphological and geomorphometrical characterization of subglacial channels on Devon Island, Nunavut, Canada
Abstract. Subglacial channels are morphologically and morphometrically distinct in comparison to fluvial channels, yet their identification from remote sensing data is still problematic. To contribute to the current set of criteria used to identify such channels, we performed detailed field observations of two subglacial channel networks on Devon Island, Nunavut, Canada. In planform, these channels are isolated, finger-like networks that drain into a main stem and have distinct cross-sectional and longitudinal profiles. Cross-sections are flat-bottomed with steep walls and longitudinal profiles are convex and exhibit undulations, typical of pressurized water flow (i.e., subglacial flow). To facilitate remote sensing identification, we interrogated how well-known scaling relationships capturing hydraulics and mass balance dynamics of fluvial systems differ in subglacial channels. Scaling relationships typically used to discern connections between discharge and channel and catchment size in fluvial systems were applied to both networks, yielding trends distinct from the fluvial literature. We suggest that the weakly correlated relationship we found between channel discharge and the size of the drainage area indicates a discrete point or line source of water, such as a moulin or crevasse.
- Preprint
(1414 KB) - Metadata XML
- BibTeX
- EndNote
Status: closed
-
RC1: 'Comment on egusphere-2024-164', Rob Storrar, 26 Feb 2024
This paper presents the findings of a morphometric study into a two sets of meltwater channels on Devon Island. The morphometry of the channels strongly supports their subglacial origin. Metrics are applied to compare the channels with fluvial systems and to gain some process insights into the conditions surrounding the formation of the channels. The conclusion is that the most likely source of meltwater is seasonal drainage of surface meltwater into the subglacial system, specifically in the late melt season.
This is a neat piece of work and the manuscript is well-written and supported by useful figures. The work provides a useful advance on our understanding of the processes that lead to the formation and character of subglacial channels, and connects this with the contemporary glacial hydrology literature a way that provides a holistic link between modern and palaeo ice sheet meltwater drainage, which I think is very important.
Since the channels are abandoned, proxies are needed to estimate discharge and drainage area size. Whilst these proxies may be considered speculative, they make conceptual sense and in the absence of other metrics I think that they are reasonable assumptions to make.
The only moderate-minor issue I would raise is the duration needed to form the channels, which is under-explored in the manuscript, but which is probably quite important for the interpretation. I think the seasonal drainage model explanation works, but there are questions around how long this configuration would need to persist for to generate the network of channels observed. Definitively answering this question is beyond the scope of the paper, but I feel it warrants some discussion. I think there are two key issues:
1. Synchronous formation or time-transgressive? The analysis rests on the assumption that the network of eroded channels described reflects the network of channels that were hydrologically active at a single time (or period of stability). An alternative interpretation may be that fewer channels were active at one time, and that a smaller network of channels gradually evolved under a retreating ice sheet. I think the synchronous model works in this case, but time-transgressive formation of anabranches may need consideration.
2. If the channels reflect a stable synchronous(ish) network, which I think is likely, why would this have existed? The final part of the discussion alludes to this by mentioning a period of different climatic conditions. I think something that could be made clearer is why there is a clear two-state phenomenon (i.e. state 1 is cold-based with no subglacial channels, state 2 is a stable wet-based configuration). The question in my mind is why would a wet-based configuration not change over time? I would think possible answers would be that (1) the system switched to polythermal from cold-based for a relatively short period; (2) somehow polythermal conditions are only possible under certain ice sheet configurations.
Minor comments:
Line 28: till may be more appropriate than tillite (i.e. it may not be consolidated into rock everywhere).
Line 160: Is anastomosing the correct term? My understanding is that this is a specific term for rivers that anabranch due to consolidated banks. I think anabranching is a more appropriate term to describe diverging and converging channels, although I could be mistaken. If so, please see other uses of anastomosing elsewhere.
Lines 180-181. Not sure what is meant by a LiDAR image? Are the LiDAR data not collected as a point cloud? Does this mean that point clouds were collected from 1017 and 499 locations? Please clarify.
Lines 200-202. Not sure this is a novel metric? Is this not the same as a braiding index? See for example:
Bridge, J. S. (1993). The interaction between channel geometry, water flow, sediment transport and deposition in braided rivers. Geological Society of London, Special Publications, 75, 13-71.
Friend, P. F. & Sinha, R. (1993). Braiding and meandering parameters. Geological Society of London, Special Publications, 75, 105-111.
Lines 284-286: The wording here is a bit confusing, and I would suggest rewriting it to make clearer. I think you mean that you didn’t observe any clear topographic controls on subglacial channel drainage area, and so use channel length as a proxy?
Lines 287-288: Given that both of these values are proxies with no way to demonstrate how reliable they are, might it be better to simply state that channel cross-sectional area is not correlated with channel length? This is definitely true, whereas it is difficult to prove that discharge was not related to area. Once key unknown that is likely to have influenced this is the duration of channel formation.
Line 414-427: I think the late melt season drainage event explanation makes sense. A key question in my mind is how long the channels took to erode to their final form, and whether they did so under a stable ice sheet configuration. Could the anabranching reflect longer term subtle changes in ice surface slopes/meltwater discharge that caused the planform to shift over time? Alternatively, perhaps the network does indeed reflect a long-term stable ice sheet configuration that included seasonal subglacial meltwater drainage.
Citation: https://doi.org/10.5194/egusphere-2024-164-RC1 -
RC2: 'Comment on egusphere-2024-164', Rod Smith, 05 Mar 2024
My recommendation is that this manuscript be rejected.
Without any doubt, the channels and channel systems the authors describe here are nested lateral meltwater channels, incised along retreating cold-based/polythermal glacier margins, and not as they propose, related to subglacial channelized flow.
The area of Devon Island that the authors were working in is literally riven with lateral meltwater channels, at an abundance and configuration that immediately strikes the observer (see attached Figure). Indeed, receiving and reading this paper, my first action was to download the ArcticDEM data files, and then dig out the stereo air photographs for this region and have a detailed look at their morphology and landscape configuration. It is the latter that perhaps is the most striking omission by the authors in their consideration and discussion of these features, and one that may have led them to pursue a sub-glacial, rather than glacier marginal formative environment. The channels are clearly nested, and are configured obliquely downward along hill slopes as parallel to sub-parallel bedrock erosional features (see attached Figure). There is no conceivable mechanism by which channelized (pressurized) subglacial drainage (i.e. Nye channels) would create this array of incisions, obliquely and step-wise down a slope. Further, the configuration of subparallel meltwater channels on the one side of the topographic valley they are situated in - is mirrored by the same style, form, and configuration of channels on what would have been the opposing/opposite glacial margin – clearly defining a former glacial lobe extent – this is most readily apparent with their SG2 site (see attached Figure)…but exists throughout their field area (and the SG1 site) in a truly remarkable density of lateral meltwater channels. Consider also that these channels cross interfluves, and can be followed along valley walls/former ice margins – cutting across slope-perpendicular drainage channels/valleys. The only way this is possible would have been along a cold-based glacial margin, in which meltwater (supraglacial, subglacial draining to the lateral margins (i.e. Iken, 1974), and terrestrial) was channeled along an ice margin that infilled the topography, preventing drainage down the natural valleys on the hillslope. Were this a pressurized subglacial drainage system, then there is no logic as to why flow would be channeled obliquely across slopes (in a step-wise fashion), but not deflected downward along topographic depressions where subglacial basal hydrostatic pressure would focus (barring some unimaginable arrangement of cold and warm-based configurations).
Regarding Lines 138-140 “Very little work has been done on Devon Island to map and characterize glacial landforms and their distribution in the landscape. Dyke (1999) produced the only published glacial geological map of Devon Island in an attempt to resolve the debate surrounding the existence of the IIS.”
It is unfortunate that the authors are unaware of Dyke’s 5 published detailed maps of the surficial geology of Devon Island – had they seen these, perhaps his extensive documentation of networks of lateral meltwater channels in the authors’ study area, may have helped better guide their investigations.
Dyke, A.S. 2001. Surficial geology, western Devon Island, Nunavut; Geological Survey of Canada, Map 1972A, scale 1:250 000. https://ostrnrcan-dostrncan.canada.ca/entities/publication/8cc0a2f4-87fd-4349-ab4a-5417282d2115?fromSearchPage=true
I am curious as to the authors’ assertion of convex channel thalweg profiles – and am assuming there is confusion about geomorphic order of incision of channels and “connector” channels which may relate to periods where water flowing along a cold-based margin suddenly diverted subglacially where the margins became temporarily/seasonally warm-based. I would need to see exact thalweg traces used in order to evaluate this evidence.
There is nothing distinctive about the channels in the two sites the authors have focussed on compared to the 100s of channel networks found in their immediate area. Failure to recognize an intrinsic geomorphic character of this region and its glacial history has led them to pursue an untenable, and unnecessary subglacial explanation.
-
RC3: 'Comment on egusphere-2024-164', Rob Storrar, 07 Mar 2024
Having read Rod Smith's review, I have looked at ArcticDEM and satellite imagery of Devon Island, and similar features to these are indeed widespread on the island, with orientations consistent with being flights of (recessional) lateral meltwater channels. For example, on the southern side of the island, directly south of the study area, there are many channels aligned obliquely along the sides of valleys which are strongly indicative of lateral meltwater channel geometries. This is consistent with the contemporary cold based geomorphological processes. I think this wider context is critical for interpretation of these features, and has prompted me to take a more critical look at the proposed channel origin. It would be useful to see a map of all meltwater channels in the area to get a sense of their patterns relative to palaeo-ice sheet configurations.
Given this context, the simplest explanation for these channels is that they are lateral meltwater channels. I think in order to conclusively demonstrate that these are subglacial channels they would need to have at least some portions of clear uphill slopes on the thalwegs. Whilst profile C-C' does contain some apparent uphill geometries, these are very low magnitude (a few metres) and over short lengths: overall the profile, if smoothed, would be a simple downhill slope. As such, I wonder if they may be an artifact of the digitised thalweg not corresponding perfectly to the true thalweg of the DEM - if it missed by a small amount the effect in the profile would be for it to appear to go uphill. A more reliable way of achieving thalweg profiles would be to use simple hydrological tools in GIS to generate the true thalweg from the DEM. Elevation profiles along this would be a better representation of true undulations in thalwegs.
With this additional context in mind, I think an appraisal of the origin of the channels should take into consideration the wider landscape evidence and should very carefully address the relative likelihood of these channels being lateral or subglacial, since this profoundly affects the interpretation.
Citation: https://doi.org/10.5194/egusphere-2024-164-RC3
Status: closed
-
RC1: 'Comment on egusphere-2024-164', Rob Storrar, 26 Feb 2024
This paper presents the findings of a morphometric study into a two sets of meltwater channels on Devon Island. The morphometry of the channels strongly supports their subglacial origin. Metrics are applied to compare the channels with fluvial systems and to gain some process insights into the conditions surrounding the formation of the channels. The conclusion is that the most likely source of meltwater is seasonal drainage of surface meltwater into the subglacial system, specifically in the late melt season.
This is a neat piece of work and the manuscript is well-written and supported by useful figures. The work provides a useful advance on our understanding of the processes that lead to the formation and character of subglacial channels, and connects this with the contemporary glacial hydrology literature a way that provides a holistic link between modern and palaeo ice sheet meltwater drainage, which I think is very important.
Since the channels are abandoned, proxies are needed to estimate discharge and drainage area size. Whilst these proxies may be considered speculative, they make conceptual sense and in the absence of other metrics I think that they are reasonable assumptions to make.
The only moderate-minor issue I would raise is the duration needed to form the channels, which is under-explored in the manuscript, but which is probably quite important for the interpretation. I think the seasonal drainage model explanation works, but there are questions around how long this configuration would need to persist for to generate the network of channels observed. Definitively answering this question is beyond the scope of the paper, but I feel it warrants some discussion. I think there are two key issues:
1. Synchronous formation or time-transgressive? The analysis rests on the assumption that the network of eroded channels described reflects the network of channels that were hydrologically active at a single time (or period of stability). An alternative interpretation may be that fewer channels were active at one time, and that a smaller network of channels gradually evolved under a retreating ice sheet. I think the synchronous model works in this case, but time-transgressive formation of anabranches may need consideration.
2. If the channels reflect a stable synchronous(ish) network, which I think is likely, why would this have existed? The final part of the discussion alludes to this by mentioning a period of different climatic conditions. I think something that could be made clearer is why there is a clear two-state phenomenon (i.e. state 1 is cold-based with no subglacial channels, state 2 is a stable wet-based configuration). The question in my mind is why would a wet-based configuration not change over time? I would think possible answers would be that (1) the system switched to polythermal from cold-based for a relatively short period; (2) somehow polythermal conditions are only possible under certain ice sheet configurations.
Minor comments:
Line 28: till may be more appropriate than tillite (i.e. it may not be consolidated into rock everywhere).
Line 160: Is anastomosing the correct term? My understanding is that this is a specific term for rivers that anabranch due to consolidated banks. I think anabranching is a more appropriate term to describe diverging and converging channels, although I could be mistaken. If so, please see other uses of anastomosing elsewhere.
Lines 180-181. Not sure what is meant by a LiDAR image? Are the LiDAR data not collected as a point cloud? Does this mean that point clouds were collected from 1017 and 499 locations? Please clarify.
Lines 200-202. Not sure this is a novel metric? Is this not the same as a braiding index? See for example:
Bridge, J. S. (1993). The interaction between channel geometry, water flow, sediment transport and deposition in braided rivers. Geological Society of London, Special Publications, 75, 13-71.
Friend, P. F. & Sinha, R. (1993). Braiding and meandering parameters. Geological Society of London, Special Publications, 75, 105-111.
Lines 284-286: The wording here is a bit confusing, and I would suggest rewriting it to make clearer. I think you mean that you didn’t observe any clear topographic controls on subglacial channel drainage area, and so use channel length as a proxy?
Lines 287-288: Given that both of these values are proxies with no way to demonstrate how reliable they are, might it be better to simply state that channel cross-sectional area is not correlated with channel length? This is definitely true, whereas it is difficult to prove that discharge was not related to area. Once key unknown that is likely to have influenced this is the duration of channel formation.
Line 414-427: I think the late melt season drainage event explanation makes sense. A key question in my mind is how long the channels took to erode to their final form, and whether they did so under a stable ice sheet configuration. Could the anabranching reflect longer term subtle changes in ice surface slopes/meltwater discharge that caused the planform to shift over time? Alternatively, perhaps the network does indeed reflect a long-term stable ice sheet configuration that included seasonal subglacial meltwater drainage.
Citation: https://doi.org/10.5194/egusphere-2024-164-RC1 -
RC2: 'Comment on egusphere-2024-164', Rod Smith, 05 Mar 2024
My recommendation is that this manuscript be rejected.
Without any doubt, the channels and channel systems the authors describe here are nested lateral meltwater channels, incised along retreating cold-based/polythermal glacier margins, and not as they propose, related to subglacial channelized flow.
The area of Devon Island that the authors were working in is literally riven with lateral meltwater channels, at an abundance and configuration that immediately strikes the observer (see attached Figure). Indeed, receiving and reading this paper, my first action was to download the ArcticDEM data files, and then dig out the stereo air photographs for this region and have a detailed look at their morphology and landscape configuration. It is the latter that perhaps is the most striking omission by the authors in their consideration and discussion of these features, and one that may have led them to pursue a sub-glacial, rather than glacier marginal formative environment. The channels are clearly nested, and are configured obliquely downward along hill slopes as parallel to sub-parallel bedrock erosional features (see attached Figure). There is no conceivable mechanism by which channelized (pressurized) subglacial drainage (i.e. Nye channels) would create this array of incisions, obliquely and step-wise down a slope. Further, the configuration of subparallel meltwater channels on the one side of the topographic valley they are situated in - is mirrored by the same style, form, and configuration of channels on what would have been the opposing/opposite glacial margin – clearly defining a former glacial lobe extent – this is most readily apparent with their SG2 site (see attached Figure)…but exists throughout their field area (and the SG1 site) in a truly remarkable density of lateral meltwater channels. Consider also that these channels cross interfluves, and can be followed along valley walls/former ice margins – cutting across slope-perpendicular drainage channels/valleys. The only way this is possible would have been along a cold-based glacial margin, in which meltwater (supraglacial, subglacial draining to the lateral margins (i.e. Iken, 1974), and terrestrial) was channeled along an ice margin that infilled the topography, preventing drainage down the natural valleys on the hillslope. Were this a pressurized subglacial drainage system, then there is no logic as to why flow would be channeled obliquely across slopes (in a step-wise fashion), but not deflected downward along topographic depressions where subglacial basal hydrostatic pressure would focus (barring some unimaginable arrangement of cold and warm-based configurations).
Regarding Lines 138-140 “Very little work has been done on Devon Island to map and characterize glacial landforms and their distribution in the landscape. Dyke (1999) produced the only published glacial geological map of Devon Island in an attempt to resolve the debate surrounding the existence of the IIS.”
It is unfortunate that the authors are unaware of Dyke’s 5 published detailed maps of the surficial geology of Devon Island – had they seen these, perhaps his extensive documentation of networks of lateral meltwater channels in the authors’ study area, may have helped better guide their investigations.
Dyke, A.S. 2001. Surficial geology, western Devon Island, Nunavut; Geological Survey of Canada, Map 1972A, scale 1:250 000. https://ostrnrcan-dostrncan.canada.ca/entities/publication/8cc0a2f4-87fd-4349-ab4a-5417282d2115?fromSearchPage=true
I am curious as to the authors’ assertion of convex channel thalweg profiles – and am assuming there is confusion about geomorphic order of incision of channels and “connector” channels which may relate to periods where water flowing along a cold-based margin suddenly diverted subglacially where the margins became temporarily/seasonally warm-based. I would need to see exact thalweg traces used in order to evaluate this evidence.
There is nothing distinctive about the channels in the two sites the authors have focussed on compared to the 100s of channel networks found in their immediate area. Failure to recognize an intrinsic geomorphic character of this region and its glacial history has led them to pursue an untenable, and unnecessary subglacial explanation.
-
RC3: 'Comment on egusphere-2024-164', Rob Storrar, 07 Mar 2024
Having read Rod Smith's review, I have looked at ArcticDEM and satellite imagery of Devon Island, and similar features to these are indeed widespread on the island, with orientations consistent with being flights of (recessional) lateral meltwater channels. For example, on the southern side of the island, directly south of the study area, there are many channels aligned obliquely along the sides of valleys which are strongly indicative of lateral meltwater channel geometries. This is consistent with the contemporary cold based geomorphological processes. I think this wider context is critical for interpretation of these features, and has prompted me to take a more critical look at the proposed channel origin. It would be useful to see a map of all meltwater channels in the area to get a sense of their patterns relative to palaeo-ice sheet configurations.
Given this context, the simplest explanation for these channels is that they are lateral meltwater channels. I think in order to conclusively demonstrate that these are subglacial channels they would need to have at least some portions of clear uphill slopes on the thalwegs. Whilst profile C-C' does contain some apparent uphill geometries, these are very low magnitude (a few metres) and over short lengths: overall the profile, if smoothed, would be a simple downhill slope. As such, I wonder if they may be an artifact of the digitised thalweg not corresponding perfectly to the true thalweg of the DEM - if it missed by a small amount the effect in the profile would be for it to appear to go uphill. A more reliable way of achieving thalweg profiles would be to use simple hydrological tools in GIS to generate the true thalweg from the DEM. Elevation profiles along this would be a better representation of true undulations in thalwegs.
With this additional context in mind, I think an appraisal of the origin of the channels should take into consideration the wider landscape evidence and should very carefully address the relative likelihood of these channels being lateral or subglacial, since this profoundly affects the interpretation.
Citation: https://doi.org/10.5194/egusphere-2024-164-RC3
Viewed
| HTML | XML | Total | BibTeX | EndNote | |
|---|---|---|---|---|---|
| 241 | 69 | 15 | 325 | 5 | 8 |
- HTML: 241
- PDF: 69
- XML: 15
- Total: 325
- BibTeX: 5
- EndNote: 8
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1