Initial comments

This study uses a time-series of high resolution DSMs to investigate the interplay between grounded ice, ice shelf, floating tongue and basal channels at Thwaites Glacier. This is a topic of great interest in the cryosphere and the research is very well conceived, investigated and presented. The methodology is mostly made clear (see comments below), substantial interesting findings are presented, and the whole paper may serve as a model for the use of precisely calibrated DEMs for investigating changes in the “Hydrostatic Boundary” at ice shelf-ocean interfaces.

General Comments

It is understandable that the (very) recent paper by Eric Rignot (Widespread seawater intrusions…) is not mentioned in this study, probably because it was in review as this paper was being submitted. I recommend that the authors include this paper in their review not simply because it is relevant, but because it could serve to clarify the relationship between the transition zone between grounded and floating ice as detected by InSAR and the Hydrostatic Boundary as measured by DSM analysis. Professor Rignot’s paper finds evidence of seawater-induced vertical motion inland beyond the HBs in this paper and the discussion could be quite informative. The adoption of informal names of some sub-glacial features may also be appropriate. From the present high quality of argumentation and discussion I doubt it will take long to add this potentially valuable element.

Mostly, the remaining Thwaites ice shelf is referred in this paper to as “TGIS” (Thwaite Glacier Ice Shelf), but I detected some “TEIS” references (Thwaites Eastern Ice Shelf) which is my own preference because it acknowledges the former existence of a western ice shelf. Consistency is obviously required and you (and maybe the Editor) should decide which to use.

Specific comments

Line 112: which “annual velocity map”?

Line 114: how can you have a “median of two”, and how do you define “summer quarters”?

Line 137: please expand what you mean by “each independent continuous grounding line”

Line 148: I admire that you have used ‘inclusive’ colour scales. It is best not to refer to the (subjectively received) colours in the main text but allow the figures to speak for themselves

Figure 1: The IPY GL and 2011 GL are apparently in the same colour and the former is probably obscured by the HB sequence. Some adjustments (or removals) are reauired here.

Line 208: “Remaining artefacts .. are filtered out”. Please elucidate.

Line 236: “Several”. Why not be precise here?

Line 261: “TEIS” and “TWIT”. I would say these have gained enough currency for general adoption. But then I would.

Lines 279 and 286: seven, then eight basal channels?

Line 303: “by the end of the study period”. You could help the reader here by giving precise time boundaries.

Line 305: I couldn’t see how Figure 7 could be used as evidence here.

Line 405: “was extended .. arbitrarily”. Please explain more precisely what you mean.

Line 427: “or an error in the manual delineation”. This alerted me to the fact that I had missed that a manual step is involved in the method - I had assumed that the process was automated. Perhaps you could expand the methods section to explain this in a nit more detail and discuss the potential errors. Errors in manual steps are rather different from uncertainties in automated processing. I think this sentence needs some more nuance.

Line 442: extra brackets.

Line 520: “volume of basal melt .. 3.5Gt”. Please give a time period as well as an area. To claim this as “ample” requires some more data or argumentation.

Line 529: Figure S8a would need some more annotation to support this statement.

Great work.

Adrian Luckman, 22^nd May 2024

This study present new observation of grounding line migration and basal melting under the floating section of the Thwaites glacier of the West Antarctica Ice Sheet.  The study focuses particularly on the link between the presence of basal channels, basal melting, and grounding line migration. To achieve this, the authors analyse an extensive dataset of high-resolution digital elevation models, providing a unique lens into the intricate processes taking place at grounding lines of a rapidly melting sector of Antarctica.

The research is very topical and the findings provide novel insights into the potential role of channels in modulating both the melting under ice shelves as well as grounding retreat. The manuscript is very well written and illustrated, the method is very well documented, the results and discussion are informative and provide a balanced view of findings and limitations.

Specific comments:

Line 35-38: It is unclear here what the relative role of subglacial channelization and meltwater plumes is in creating these ice shelves basal channels.

L83: SAR-In -> SARIn

L122: Could you say more about the 07-09 GL datasets? Is this the 2011 grounding line in https://agupubs.onlinelibrary.wiley.com/doi/full/10.1002/2014GL060140 ? Which would then tie in nicely with the start of your HB record in 2011.

L203: Should it say “the velocity divergence is computed at each time step prior to the DSM being flow–shifted, …”?

L206: What are the implications of having the SMB data only up to 2016 when the melt rate is calculated up to 2023? What would the impact of large anomalies e.g. https://www.nature.com/articles/s41467-023-36990-3 be?

Figure 3: (g) and (h) labels are only partially visible.

L255: Not sure what you mean here? What other than melting would cause basal mass change? Please clarify.

L256: I suggest caution in how you present “thinning” over ice shelves, this is Lagrangian change in elevation – “thinning” might create confusion with the reader. I would suggest using a different term. This applies to many sections of the manuscript, where elevation and thickness change over floating ice is mentioned.

L263: Meaning that apparent refreezing is an artefact of the floating assumption? Or does the transient grounding leads to real refreezing somehow? Please clarify.

L276: Is it “everywhere”? In several sectors (cavities 6, 7, 8 &9), IPY GL appear to be inland of the HB position in ~ 2011 and 2012. Given the importance of this sector it is probably worth discussing and providing potential explanation for this.

L450: I am curious whether you observe the pinning point evolution at TWIT described in: https://doi.org/10.1029/2023GL103088 from the unfiltered figure 5 it appears so but it would be worth a mention, and why this pinning point may or may not be more robust than the other unfiltered HB features.

L466: It would be good for the discussion to reflect on the implication of the findings in light of the recent publications (e.g. https://www.pnas.org/doi/full/10.1073/pnas.2404766121 but also others) on ocean intrusion within the grounding zone, possibly by expanding some of the related discussion in section 5.2.  Your comments on the absence of elevation thinning in these sectors for example seem particularly relevant.

L475: “Rapid thinning” in a Lagrangian sense has a different meaning than the general understanding of ice shelf thinning. I would urge caution. Somewhere in the manuscript it would be good to articulate what can lead to ice shelf thinning in a Lagrangian reference frame, and in the discussion to address the plausibility of various processes.

L553: Could you discuss the implication of this magnitude of error on the smooth annual HB? Could this explain some of the discrepancy with the IPY GL in the TWIT sector (fig. 5)?