Ocean–Ice Sheet Coupling in the Totten Glacier Area, East Antarctica: Analysis of the Feedbacks and Their Response to a Sudden Ocean Warming

Round 1

Reviewer 1 Report

Summary

This is modelling study of the ice flow and regional ocean for a specific drainage basin in East Antarctica thought to be vulnerable to ocean warming and potentially a source of significant global mean sea level rise.  Compared to some previous work, the novelty in this study lies in the sophistication of the coupling of the dynamic ice flow with the ocean, with higher spatial resolutions and more explicit accounting of some processes. The paper has a tight focus on the impact of including the dynamic response of the ice in their system, compared to the ice shelf melt rates found in their ocean-only simulations. This allows for a useful analysis of what the coupling in the system is really doing, although it comes somewhat at the expense of treating the simulations more as technical exercises than as representations of reality.  The analysis presented is generally thorough and clearly written, with plenty of figures. I haven't done a proper language check, but my impression was that there were quite a few places with minor grammatical issues (eg. number and tense issues with verbs) that could do with a qualified proof read. I have some minor comments and suggestions for improvements but on the whole I think it's worth publishing largely as it is.


General comments

There are two areas where I feel the paper could be improved into a more rounded contribution to the literature. It's possible that the authors have deliberately not written in this way in order to keep their focus purely on the impact of the coupling on their simulated melt rates, but I still think it would be worth doing something in these directions.

Firstly, although the impact of the coupling on melt rates in their base-state simulation is analysed in depth, little is shown to persuade the reader that their simulations are realistic, past the comment on line 276 that the average melt rates  - although not the interannual variability(?) - are similar to those derived by Rignot. You could argue that if you are only interested in the difference caused by including the coupling then the details of that base state don't matter but I would personally like to see a more serious evaluation here.  For example, having noted in the introduction that actual grounding line retreat has been detected during the same time period that they are simulating in the REF3mcpl simulation, some comparative comment would be welcome in the text about why they don't simulate that behavior.  Is the variability they see in line with that observed and modelled (eg Gwyther et al 2018)? I'm aware that a previous paper by the authors /does/ go into some depth on the ocean side of things, but that isn't really referred to in this paper. There's also no illustration of the output from the ice sheet model - you could argue that it hasn't had time to evolve significantly away from the initial condition it was inverted to, but since the ice velocity field is crucial to the differences in shelf thickness caused by the coupling it would be nice to see what it is in this configuration - and perhaps compare any changes to those recently observed for the real shelves -  even if it doesn't evolve much.

The second area is that while the introduction does note some of the pre-existing literature, the results of the simulations are not really put into any of that context. The resolution and domain and coupling are, on the face of it, really not so different from Pelle et al.'21, and although Pelle's 21st century scenarios are not quite the same as the warm anomaly applied here, Pelle et al claim that they have a very similar goal to the present authors in looking at the generic response of the system to ocean warming, not actually doing projections. Just noting that the authors simulate some grounding line retreat in the same area as Pelle et al feels pretty thin when there are quite a few parallels with that study and in the Introduction they have gone to the trouble of noting various improvements in their modelling system compared to Pelle's. Can they identify what in their simulation, if anything, is improved as a result of their more sophisticated model? Tides and eddies are mentioned as improvements in their system in the Introduction but I don't think are mentioned beyond that in the analysis? Are the changes in melt rate they attribute to the coupling (or the warming) significant compared to variability?  Again, such a comparison perhaps doesn't speak to the primary focus of the pure impact of coupling in a very tightly controlled system, but the Discussion in this paper really adds very little to the recap of the main results, and it would be interesting to have their overall conclusions analysed or put in /some/ wider context or other.

As a final note: I appreciated the many figures provided to illustrate the results, but found the monochrome colour scales without contours (figs 3,4,9,10,13,14) not the easiest to read in terms of making out the actual values being plotted for any location, or for differentiating between closely plotted solid and dashed lines. More varied colour maps would help with this, even if breaks with the authors apparent preferred paradigm of "more melt=more red". The scale labels on figs 4,5,6b and 14c are also overlapping


Specific comments

Introduction - There is a literature on modelling melt rates for Totten Glacier that isn't really noted here, eg Gwyther et al 2014, Roberts et al 2018 and references therein, including some with coupling. A wider view of the literature could be taken

Figure 1. I think readers would appreciate some larger context for where on Antarctica the study area is, and also what the domain of the ice sheet model is, not just the ocean.

Figure 3. The impact of coupling on shelf melt rates is not constant in time, nor entirely simultaneous between the shelves. Can you comment? The variability seems to be several times higher than quoted from Rignot et al - that seems worth a comment too.

Figure 8b. The implied dependence on the change in velocity is quite striking. The resolution and vertical coordinate of the ocean model is not sufficient to be capturing any plume dynamics properly, even if it does show some steepening of the draft. Some comment could be made here on how robust the strength of the simulated effect is and what the impact of a more complete turbulent model might show - or even on the simpler coupling or melt models used for some ice sheet studies that only really consider temperature in the cavity?

Figure 15b. has the resolution of the ice sheet model been tested for convergence in terms of grounding line behavior for this system?

Gwyther et al 18. www.nature.com/articles/s41467-018-05618-2
Gwyther et al 2014 https://doi.org/10.5194/os-10-267-2014
Roberts et al '2018 https://doi.org/10.1144/SP461.6

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Reviewer 2 Report

The article describes the results that are understandable and easily explained from a physical point of view. But the authors pay little attention to this. As an external observer, I can say that the rate of melting of the glacier is related to the external flow of heat, which can be brought from the atmosphere, their sea or from the marine crust. Therefore, the melting rate should be linearly related to changes in atmospheric temperature, as well as to the flow of heat from the sea, which directly depends on the magnitude of the flow velocity. In lines 476-483, the authors write: «The negative feedback associated with the ice sheet coupling, which reduces the effect of the ocean warming on the basal melt rate, can be explained by the changes in basal melt rate present in the center of the cavity. In this zone, where high melt rates are observed in WARM, melt rates are decreased in WARM3mcpl. As illustrated in Figure A8 (in the appendix), at those locations, the ice shelf starts to thin at the beginning of WARM3mcpl (from 800 to 500 m of ice) but stabilises after 3 years. The thinner ice shelf rises the in situ freezing temperature at the ice–ocean interface, which, at some point, becomes larger than the ocean temperature and decreases the basal melt rates.» . The authors associate the feedback with the thickness of the glacier, which is completely wrong. The increase in freezing temperature is related to the amount of fresh water and has nothing to do with the thickness of the ice. An increase in the ratio (fresh water/salt water) leads to a decrease in salinity, i.e. to an increase in the freezing point. And this can be at any thickness of ice.

Completely wrong phrase: The changes in basal melt rate are less related to the changes in ocean temperature. It should depend on physical considerations.

Again, there is no influence of biota on these processes. Let him bring his thoughts.

The conclusion needs to be rewritten, involving physical patterns in the explanation.

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Round 2

Reviewer 2 Report

The article can be published.