The past, present, and future viscous heat dissipation available for Greenland subglacial conduit formation

Mankoff, Kenneth D.; Tulaczyk, Slawek M.

doi:https://doi.org/10.5194/tc-11-303-2017

Articles | Volume 11, issue 1

https://doi.org/10.5194/tc-11-303-2017

© Author(s) 2017. This work is distributed under
the Creative Commons Attribution 3.0 License.

https://doi.org/10.5194/tc-11-303-2017

© Author(s) 2017. This work is distributed under
the Creative Commons Attribution 3.0 License.

Articles | Volume 11, issue 1

Research article

|

30 Jan 2017

Research article |

| 30 Jan 2017

The past, present, and future viscous heat dissipation available for Greenland subglacial conduit formation

Kenneth D. Mankoff and Slawek M. Tulaczyk

Download

Final revised paper (published on 30 Jan 2017)
Supplement to the final revised paper
Preprint (discussion started on 03 Jun 2016)
Supplement to the preprint

Interactive discussion

Status: closed

AC: Author comment | RC: Referee comment | SC: Short comment | EC: Editor comment

- Printer-friendly version

- Supplement

RC1: 'Referee comments', Anonymous Referee #1, 24 Jun 2016
- AC1: 'Reply to RC1 comments', Kenneth Mankoff, 29 Jun 2016
RC2: 'Referee comments', Anonymous Referee #2, 15 Jul 2016
RC3: 'Comments on “The past, present, and future viscous heat dissipation available for Greenland subglacial conduit formation” by K. Mankoff and S. Tulaczyk', Anonymous Referee #3, 27 Jul 2016

Peer-review completion

AR: Author's response | RR: Referee report | ED: Editor decision

AR by Ken Mankoff on behalf of the Authors (01 Sep 2016) Author's response Manuscript

ED: Referee Nomination & Report Request started (12 Sep 2016) by Olivier Gagliardini

RR by Anonymous Referee #3 (27 Sep 2016)

RR by Anonymous Referee #2 (27 Sep 2016)

Suggestions for revision or reasons for rejection

Aspects of this work have been greatly improved since the initial draft, notably a much simplified Methods section, and the idea behind the work remains interesting. However, other aspects of the manuscript still leave much to be desired, including the organization of the paper (which is muddled in places), the extensive comparison to Tedstone (which was unconvincing), and the majority of the Discussion text (which almost wholly lacks synthesis). Unfortunately I do not think that the manuscript meets the standard of the journal at this time.

General points

Improved Methods

The work that the authors put into this section shows well. The methods are much more condensed and clear. However, I do still find some issues.
- I remain convinced that the outflow elevation of marine glaciers would be correctly defined as sea level, as AR1 also pointed out. The authors’ treatment of this (P4 L10-13, P8 L27, and responses to reviewers) states otherwise. In my understanding, a parcel of water discharged at the base of a glacier may remain at that elevation, but it displaces a parcel of water upward, which itself displaces the next parcel upward. When these gains in potential energy are integrated, the net result is sea-level potential. However, this is only a small worry at this point.
- I am confused that the authors study only basal potential energy (rho*g*b) while discarding potential energy from the surface (rho*g*s), which is much greater in magnitude. The paragraph on supraglacial rivers suggests that the authors recognize this (P3 L16-17 especially). Perhaps my memory of the first manuscript colors this, though.
- The treatment of PTT in the calculation seems to be inconsistent (P5 L11-14). The calculation ignores PTT changes when ice thins along the basal hydraulic gradient, but includes it when ice thickens along the water path.

Organization

The manuscript suffers from poor organization. For instance:
In Results, a methods paragraph (P4 L26-31) appears.
In Discussion, a results sentence appears for the first and only time in the paper (P7 L8).
The Discussion is repetitive of the Introduction (P7 L21-25 especially, with 4 of 5 references repeated for the same facts).
There may be others. These basic offenses contributed to the difficulties I had in following the logic of this work.

Tedstone

In the revision, the authors added a substantial component to their analysis by attempting to parallel the Tedstone et al. (2015) study. To their original Historical / Present comparison (which I thought was good), the authors added a Reference / Baseline pair as well (which I found hard to keep track of, and which I think added little), based on Tedstone. I could not find that any of the reviewers asked for this addition.

The authors bluntly state that one of the Tedstone conclusions is “likely” not true (P8 L9-10) and provide an alternative hypothesis. I found the support for this new idea (section 5.2.1) entirely unconvincing and difficult to even understand the authors’ path of reasoning. The authors say that if the increase in VHD in the Tedstone basin caused the observed slowdown there (P8 L12), then this is incompatible with the ice-sheet-wide decrease in VHD observed over that period having caused the generally observed ice-sheet-wide speedup (P8 L14) — although it seems to me that these ideas do not contradict one another at all. (I am putting aside obvious issues with comparing velocity changes observed over slightly different time periods, and extrapolating land-terminating ice in western Greenland to the whole ice sheet.) But a few paragraphs later, the authors contradict themselves by extrapolating the Tedstone observations over the whole ice sheet after all (P8 L23-24). Such inconsistencies greatly undermined my confidence in this work.

Discussion

The immaturity of the Discussion and the failure of this study to integrate its results with past work are major issues. While the authors include relevant citations, they do so largely in the form of a literature review. They do not demonstrate how their results build on this literature, except in the broadest strokes (i.e., VHD enhances conduit formation, which slows ice near the margin but not far from the margin) which were already made clear in the introduction. In short, the authors leave unclear what their work adds to the community’s understanding.

Upon a closer read, I did find a few sentences where the authors introduce a nice (and perhaps novel) line of reasoning / speculation about the competing effects of increasing runoff and increasing VHD (P7 L14-17). However, that idea is buried deep within the text and not revisited. The rest of the subsection (3 paragraphs), which would normally be used to build on the idea, consists only of a rambling literature review. While I remain interested in this topic and would enjoy continuation along this line of thought, I am at a loss for specific suggestions at this point.

Alternately, it may still be possible to frame the paper differently from runoff / speedup — for example, around GHF vs VHD, which is presently an interesting section, and one that perhaps could be expanded toward greater ideas. (In fact, the abstract suggests that this is the main point of the paper, but this is not borne out.) But as it is framed now around ice sheet mass balance, the paper makes only a vague contribution. At this point, I agree with AR1 who said that the study merely converts future runoff rates into future VHD rates, which was not particularly illuminating.

Some of the section headings (4.1 Volume of subglacial hydrology; 4.3 Flow-routed spatial distribution of VHD) are inscrutable.

Figures

Figure 2 caption — how could water flow both uphill AND toward ice that thickens along flow?

Figures 4/5 — Tedstone basin should be indicated

Figure 7 — it is hard to compare the small blue dots with the large black bars. Lightening the background relief map might help, as well as plotting the dots on the top layer and using colors that provide better contrast than blue/black.

Hide

RR by Anonymous Referee #1 (04 Oct 2016)

Suggestions for revision or reasons for rejection

This is a review of the revised manuscript.

The revised paper is much improved from the original. I still find the whole calculation rather simplistic, and am not convinced that it provides new understanding, but I can accept that framing discussion in a different language may be helpful to some readers. The presentation of the revised manuscript is more coherent than the original, with more focus on results (figures), and I think this now makes it a helpful contribution, despite my reservations mentioned above and in the previous review.

I would now recommend publication, but I have some minor comments that should be considered.

In equation (1), make clear the partitioning into phi_z and phi_p - I don't think these are explicitly defined anywhere.

It may be preferable to describe the viscous heat dissipation in Watts, rather than J/year, as that would remove some of the ambiguity about increases over time (as rather confusingly raised in section 4.4). I doubt that units of PJ/year are any more informative to most readers than GW or MW.

Last paragraph of section 4.4, the sentences don't seem to agree with each other. One says that your results match the results from Tedstone et al; and then the last says there is a difference (10-20% increase rather than 50% increase). This might make sense if you mean that the runoff increase matches Tedstone et al, but the VHD increase is different; but then the explanation for the discrepancy does not really make sense, since this explanation seems to imply that the simulated runoff is incorrect. In any case, I am slightly confused why you treat the different increases in runoff and VHD as a discrepancy that needs explaining - there is no reason why they should be the same, since the VHD depends on the potential (effectively elevation) of the runoff as well as the magnitude.

The flowline in figure 6 is presumably an ice flowline, since you seem to have regions of negative potential gradient which suggest it is not possible for the water to be flowing along this line. Would it make more sense to follow a path taken by the water (ie. one of the flow paths obtained from the routing algorithm)?

In a couple of places (line 2, p8; line 5, p 11) it is suggested that viscous heat dissipation can have the effect of both accelerating and decelerating flow. Is this really what is meant? I think that is true when when talking about increased runoff, because runoff has the effect of increasing pressures as well as increasing VHD (which then opens drainage space and lowers pressures). But do you have in mind a mechanism by which VHD itself causes an increase in ice velocity (rather than the runoff that is causing the VHD)?

In section 5.2.1 (line 9, p8) it is suggested that it is 'likely' that there is a threshold change in VHD that leads to ice velocity decrease. This seems speculative to me, and I think should be phrased 'We suggest', or 'We speculate', rather than 'It is likely'. It seems well established that the effect of VHD might be to facilitate a decrease in velocity, but I do not see the evidence here for a threshold. It is also not clear what the 'baseline' for this change is (i.e. when should I measure the change from?).

Hide

ED: Reconsider after major revisions (11 Oct 2016) by Olivier Gagliardini

AR by Ken Mankoff on behalf of the Authors (28 Nov 2016) Manuscript

ED: Reconsider after major revisions (07 Dec 2016) by Olivier Gagliardini

Dear Kenneth,

I have now studied this new version of your paper based on this second round of reviews. I have still some issues with some points that need to be corrected or clarified before your paper can be accepted in TC.

My first issue is regarding two important assumptions on your approach that are not enough discussed in the current version of the paper. These two assumptions are that:
1/ the area concerned by runoff input is constant over the period of simulation. At the front, it is bounded by the current geometry (see my second point) and upstream the limitation is based on the 2000 m contour level which is considered as the maximal altitude for crevasses opening from a work by Poinar et al. (2015). For me, it looks highly speculative that this elevation will not change in the future especially if more melt is expected at higher elevations. The fact that the area at the base is constant certainly induce higher value of energy per m2 than if it was allowed to evolve with time. This assumption should be discussed and high values for future presented in W/m2 should be discussed more carefully.
2/ the upper surface elevation is not updated in the future, and from your equation (1) it is obvious that the gradient of the hydropotential is much more sensitive to the surface slopes than the bed slopes. If you melt the glacier fronts, upstream slopes will increase, changing the routes of water at the glacier base. I agree that accounting for change in surface elevation would require an ice flow model which is certainly out of the scope of your paper, but anyway you should mention and discuss this strong assumption in your model results. As far as I can see, the gradient of the hydropotential is constant in your prognostic simulation? This should be mentioned. Also, as a consequence of a fixed surface elevation, the retreat of the margins is not accounted for (see my first point).

One other important point is that, as mentioned by referee #1, it would be less confusing to express the heat dissipation in W rather than in J/a. The fact that the energy is not released uniformly in time is not an argument as anyway the melt seasonality is not discussed in your paper.

I have some other minor points that should also be corrected (page number refer to the coloured version of the paper):
- everywhere in the manuscript, "year" for units should be "y" or preferentially "a"
- abstract, line 9 and 10: I would suggest to write in EJ per year (or each year).
- page 2, line 29: is it an ice flow line or a water flow line as stated in the response to referee #2?
- page 4, line 15: changing phase change temperature?
- page 5, line 29: discuss the hypothesis of the 2000m limit
- page 6, line 6: the wording thinning or thickening is a bit confusing as you are not evolving the surface elevation. I would suggest to use "for ice thickness decreasing downstream, ..." instead?
- page 7, line 4: W would be better and would avoid the sentence in bracket which is confusing.
- page 7, line 20: not a reanalysis?
- page 7, line 24: specify what is a flow line (ice or water)?
- page 7, line 25: increase in water volume? not flow?
- page 10, line 19: problem in the sentence?
- page 10, line 21: again, would be clearer if in W
- end page 10: the case of land-terminated glaciers is not discussed here
- page 12, line 3: east-ouest vs south-north more than horizontal vs vertical?
- page 13, line 20: you approach would not be useful for an ice flow model which need the water pressure or effective presssure (and which is fixed in your approach).

Olivier Gagliardini, 07/12/2016

Hide

AR by Ken Mankoff on behalf of the Authors (10 Jan 2017) Author's response Manuscript

ED: Publish as is (12 Jan 2017) by Olivier Gagliardini

AR by Ken Mankoff on behalf of the Authors (12 Jan 2017)

Download

The requested paper has a corresponding corrigendum published. Please read the corrigendum first before downloading the article.

Article (7101 KB)
Full-text XML

Short summary

There may be a ~ 7-fold increases in heat at the bed of Greenland by the end of the century due to increased runoff. The impact this will have on the ice is uncertain, but recent results indicate more heat may reduced glacier velocity near the margin, and accelerate it in the interior. We used existing model output of Greenland surface melt, ice sheet surface, and basal topography. All code needed to recreate the results, using free software, is included.