Referee comment for preprint manuscript egusphere-2023-1397 “The staggered retreat of grounded ice in Ross Sea, Antarctica since the LGM” by Matthew A. Danielson and Philip J. Bart

General comments

Danielson and Bart present an interesting compilation of seismic data from the Ross Sea Embayment (RSE) shelf to 1) assess the retreat style of grounding lines during and subsequent to the Last Glacial Maximum (LGM) and to 2) estimate grounding line stillstand durations during general retreat represented by several grounding-zone wedges (GZWs) situated in different paleo-ice stream troughs. It is a well written and a concise presentation of available data that the authors try to apply to answer longstanding research questions related to the past grounding line behavior within major high-latitude glacial outlets, i.e., which stillstand durations do individual grounding-zone wedges represent and how can that knowledge be used to characterize past ice sheet retreat in large embayments such as the Ross Sea Embayment. However, in the current manuscript version too many general assumptions regarding accumulation and erosion rates are made that base on a single data set of reliable radiocarbon age constraints that made actual calculations for one grounding line stillstand event in the eastern Ross Sea Embayment possible (Bart et al., 2018; Bart and Tulaczyk, 2020). The authors themselves highlight the apparent modern asynchronicity of grounding line response within individual glacial troughs in the Ross Sea sector to external forcing, which can therefore also be expected to grounding line retreat in the past. Generalizing accumulation and erosion rates across individual glacial troughs is therefore quite a risky and speculative approach. Therefore, the author’s key assumptions need a more robust justification since quite far-reaching conclusions are drawn from this. In the following specific comments, I elaborate on this in more detail and really hope that the authors will be able to provide conclusive reply to this, i.e., discuss those issues in more detail in their manuscript, since I generally think that this work will deliver significant broader insight into past grounding-line dynamics in a large Antarctic embayment. This will not only be important to better understand past Antarctic ice sheet dynamics but will also deliver significant information for numerical modelers that aim to improve simulations for the grounding line’s future response.

Specific comments

Major

• Chapter 2.3 “Volume and duration calculation” and resulting discussion (Chapters 4.1, 4.2, and 4.3): As 2.3 is the most important paragraph of your method section, you need to justify your approach more clearly and carefully here. Transferring the accumulation and erosion rates from one age-constrained GZW to other major GZWs – particularly from different glacial troughs – is very risky and quite speculative. Consider that large outer shelf GZWs accumulated during initial post-LGM retreat, a time of rapid GL retreat along the Antarctic margin that may have involved large amounts of subglacial meltwater that elevated sediment output at the grounding zone. Further, the initial forcing on GL retreat has likely been quite different along the Ross Sea sector (Lowry et al., 2019; Sci Adv 5), resulting in heterogenous responses of GLs and GZW formation across different glacial troughs in the RSE. In lines 217–219 you specify that “the degree of glaciation could not have been a significant contributor to erosion rate differences between drainage areas” and that “there is no evidence of warmer-than-present intervals that might have significantly increased meltwater production that would have contributed to high end erosion rates” but not only the degree of glaciation affects erosion rates in a glacial trough but also different topographic settings, subglacial geology, and exposure to external climatic and oceanic forcing. The outermost parts of the continental shelf are usually characterized by very low gradients that are not only a lot more susceptible to minor forcing but also result in larger grounding zones of thinner ice that has probably been in constant motion by e.g., tidal movements and wave action. So, those environments cannot not really be compared to modern grounding zones that may well be susceptible to a warmer climate but are usually a lot further inland on higher gradient beds mostly made from more resistant bedrock. And yes, West Antarctica is probably mostly made from sedimentary rocks but for most locations – particularly beneath the ice sheet – we simply don’t know. And even if sedimentary rocks dominate WAIS’ substrate, those may be metasedimentary in places (e.g., Jordan et al., 2020, Nat Revs Earth & Env), i.e., still be of very high resistance against glacial erosion. In my opinion, this should be discussed and considered in the manuscript as it may have affected past GL retreat by quite a lot and may have been the reason for stillstands in places as it can largely affect the bed gradient and thus GL retreat across those sections. Generally, there is a strong need to discuss the high uncertainties related to your approach outlined in 2.3. Also, elaborate more on the sediment yield and erosion rate – as of now it reads if the sediment yield of the entire WDB catchment at the grounding zone equals the erosion rate, implying that nothing would accumulate there. That should be clarified.
• Chapter 4.1.3 “Stillstand durations within individual troughs” (Lines 205–209): This really depends on both sediment supply to the grounding zone and vulnerability of individual ice stream troughs to external forcing. So small scale features are not necessarily indicative of minor stillstands. Please elaborate on that more to justify your approach.
• Chapter 4.2 “Post-LGM erosion rates in Ross Sea” (Lines 215–220): I think only considering the degree of glaciation as a control factor on erosion rates is too simplified. Not only different topographies of those troughs but also subglacial geological variations along their axes may have controlled erosion rates by quite a lot, even though those troughs were exposed to the same cold climate. This should at least be discussed as it may lead to further uncertainties for your calculations.
• Chapter 4.3 “A staggered post-LGM retreat of WAIS grounding lines in Ross Sea” (Lines 248–254): As already mentioned above, also the effect of subglacial geology should be discussed in more detail. Subglacial geological variations may be expressed as topography on the past ice sheet bed but do not necessarily have to but still may affect stillstand durations quite significantly (e.g., Dowdeswell & Fugelli, 2012, GSA Bulletin; Klages et al., 2014, QSR; Klages et al., 2015, Geomorphology). Does seismic information from within those troughs in your study area reveal sections of potentially indurated sediments (e.g., bedrock) or even basement cropping out at the former ice sheet bed? If yes that should be considered as an additional factor for inducing retreat deceleration, potential stillstand, and subsequent GZW accumulation.

Minor

• Lines 10–11: Only by mapping GZWs you cannot say anything about stillstand durations. Please rephrase!
• Line 15: Please delete the “radiocarbon” in “…at least five thousand radiocarbon years.” For referring to a period, it doesn’t really matter if you refer to years, radiocarbon years, or calibrated years before present, etc. since you do not compare your age to anything. This is only necessary if you refer to an age and to elaborate more on the radiocarbon age and the calibrated age in comparison.
• Figure 1: Please provide geographic context to both the Antarctic overview map (e.g., EAIS, WAIS, APIS, ocean basins, South Pole, etc.) but also your RSE map (names of coasts/lands, ocean basins, etc.). Also the depth scale needs to be more precise – two annotations at -40 and -1,500 mbsl is not enough for referring to the map. Cite the paper related to the IBCSO map.
• Lines 91–92: Define sediment yield here. Yield at the grounding zone?
• Figure 2: Same as for Fig. 1 here – please label geographic features for proper orientation and reference. Label the different catchments also within the figure for quicker reference. Lighter and darker shades for highlighting present and paleo drainage areas are really hard to differentiate – please change. And either change the depth scale as suggested for Fig. 1 or take out entirely because you cannot see much shelf bathymetry anyway.
• Line 119: Please clarify here. The seafloor beneath the Ross Ice Shelf could not be investigated in the scope of your study but generally it could be studied with e.g., vibroseismic methods.
• Lines 126–127: The trough actually extends all the way to the modern GL and beyond towards the interior. Therefore, rather say “…three GZWs on the inner continental shelf proximal to the modern ice shelf edge.”
• Especially line 140 but also 191 and throughout: Reduce the number of abbreviations/acronyms. It’s really confusing and distracting for the reader. Abbreviating “middle” and “outer continental shelf” for example is not necessary and also not common at all as it is referring to a general geomorphological feature.
• Figure 3: Organize the three panels of the figure next to each other. As already mentioned for the other figures – please label geographic features for orientation and reference. It’s also really hard for me to make out differences from the three plots. Maybe zoom in to specific locations to make that clearer. Same issues regarding the depth scale as for the other figures – needs to be more detailed.
• Figure 4: Labels a–q need to be larger in the figure. Again: add general geographic names. Thickness scale needs to be more detailed. Depth scale for bathymetry is missing. Prominent troughs could be labelled in figure. Add reference to IBCSO.
• Lines 262–263: This comment is a little bit in contrast to what you said earlier about initial retreat, i.e., a rather uniform response of the ice margin. Please clarify!

Technical corrections

• Title: I’m a little tripped up by the term “staggered” as it is not commonly used within the community. Rather use “episodic” or “stepwise”, etc. You can also be more specific about your study area, i.e., “across the Ross Sea Embayment shelf” or similar. Suggestion: “Episodic post-LGM grounding line retreat across the Ross Sea Embayment shelf, Antarctica”.
• Line 8: Maybe avoid the term “greater” and write something like “The inland retreat of the WAIS in the Ross Sea sector was more significant than…”.
• Line 15: Avoid “progressed” and maybe write “…shorter stillstands throughout the deglacial”. But as I said in my major comments…can we really be sure that they were shorter? You don’t have age constraints and just by the size of a wedge it’s risky to infer stillstand durations. Maybe the sediment supply was just a lot smaller.
• Line 16: “Over the course of the post-LGM” sounds weird. Suggestion: “…subsequent to the LGM” or “…throughout the deglacial”.
• Line 24: Define the acronym here and provide duration of the LGM in Antarctica in “cal. ka BP”. You can delete “Antarctica in that sentence. And specify the location – “western Ross Sea” is too imprecise since the Ross Sea extends well into the deep ocean. Say “western Ross Sea Embayment” or similar. Same for referring to the eastern part.
• Line 30: Avoid mentioning “trough” twice and write “…foredeepened Drygalski Trough, JOIDES Basin, and Pennell Trough…”.
• Lines 33–34: Rather write “During post-LGM ice sheet retreat, the GL retreat paused within the outer part of the GCB…”.
• Line 51: Get rid of the second “regional” in that sentence.
• Line 56: Rather write “The seismic profiles were interpreted…”.
• Line 67: This sounds like that your seismic interpretation is of LGM age. Therefore, rather clarify and write e.g., “Seismic interpretation and isopach mapping of (post)-LGM grounding zone wedges”.
• Line 89: Add “as a basic parameter” between “used” and “to”.
• Line 91: Replace “concerning” with “to calculate”. Add “for every ice stream” after “paleo-drainage area”.
• Line 116: Change to “Seismically-resolvable GZWs on the Ross Sea Embayment shelf”.
• Line 121: What do you mean by “GZWs have seafloor exposures”? That they were mapped and identified by multibeam bathymetry surveys? Please clarify!
• Line 123: What do you mean by “inner reaches of the middle continental shelf”? The transition from the inner to the mid shelf? Generally, try to avoid “inner reaches” here and throughout and change to “inner continental shelf” or similar.
• Line 178: Again, to me “Ross Sea” is too general. Specify to “Ross Sea Embayment shelf” or similar.
• Line 180: Depending on what you mean get either rid of the “s” in “GZWs” or in “suggests”.
• Line 181: MCS as acronym was already introduced further up. But as I suggested above, it’s kind of weird and not necessary to abbreviate general features such as shelf sections.
• Lines 181–183: Please change to “GZWs on the RSE shelf are generally larger than those on other Antarctic continental shelves.”
• Line 184: Change to “sedimentary bedrock”.
• Line 188: Change chapter title to “Variable stillstand durations between troughs”.
• Line 189: Specify to “Our data suggest that GL stillstands on the RSE shelf were of millennial to centennial durations.”
• Line 192: Change to “three millennia”.
• Line 194: Delete “the” before “millennial”.
• Line 195: Rephrase “would have had to have”.
• Table 5: What do you mean by “stratigraphic superpositions” here. Specify in caption or main text.
• Line 202: Here you write “super-position” but “superposition” in Table 5. Decide for either one. And what do you exactly mean by “superposition”? That if you consider the seismic stratigraphy of the entire shelf that the outer shelf GZWs must be the oldest? Please clarify.
• Line 204: As above – clarify if you mean transition from inner to mid shelf. Avoid acronym.
• Line 205: Please also cite more recent literature here, e.g., using data constraints or modeling evidence.
• Line 206: Change to “RSE troughs”.
• Line 210: Delete “on the” in the chapter title.
• Lines 214–215: Change to “…regional climate and associated precipitation, and the presence/absence of meltwater.”
• Line 217: Replace “100%” with “entirely”.
• Line 224: Replace “from trough to trough” with “in between troughs”.
• Chapter 4.3: Are all those ages really radiocarbon ages or were they calibrated? If calibrated, write “cal. ka BP”. If some of them are radiocarbon ages and some were calibrated, you cannot relate them to each other here.
• Lines 245–246: Change to “…in an unsteady episodic retreat style within individual troughs”.
• Line 256–257: Replace “from trough to trough” with “between troughs”.
• Line 282: Same here.

This manuscript explores the possible durations of grounding zone stillstands across the Ross Sea during post-LGM. To do this, the authors calculate the volumes of major grounding zone wedges using seismic data, and divide by sediment flux, which is derived from the product of the drainage area of the catchment and a value for sediment yield that was pulled from another Ross Sea publication. The results of the manuscript bolster the Antarctic research community’s ever-increasing agreement that the Ross Sea’s ice streams retreated asynchronously after the Last Glacial Maximum, first from the deep troughs, with grounded ice persisting on shallow banks for some time after.

The concept of this paper is interesting and the results could be valuable to the ice sheet modeling community; however, there are some issues in the execution of the study that should be considered, mostly related to the need for further exploration of boundary conditions. The authors have, understandably, had to make some assumptions for their calculations. It is fine to do this, but for a study that is so heavily dependent on assumed numerical boundary conditions, there should be more direct exploration of the range of possible values. After some additional analyses that showcase the breadth of possible grounding durations, these findings will be a valuable contribution to Antarctic science across disciplines.

Major comments

Firstly, the value for sediment yield that was chosen for the calculations was taken from a study of the age-constrained grounding zone wedge complex in Whales Deep Basin in the eastern Ross Sea. The authors point out that the catchment for the ice stream that once flowed through this trough has a largely sedimentary bed. The authors also point out that the catchment area for the ice flowing from the East Antarctic Ice Sheet into the western Ross Sea is floored by mostly crystalline bedrock, which can generate a 30% lower sediment yield than a sedimentary bed. This is a significant amount, yet it is dismissed. I would recommend the authors conduct another set of calculations using a 30% lower sediment yield, if not for the entire Ross Sea, then at least the western Ross Sea. They can then provide readers with a range of possible durations of grounding zone stillstands at each study location.

Another issue needing clarification is the approach for choosing particular catchment areas for the calculations. For instance, it isn’t clear how the catchment area is divided between JOIDES and Pennell basins. Is the modern-day “brown” drainage area essentially halved for each, or is there some more nuanced reasoning on how the area Is partitioned into the two distributaries? It would be good to insert the values for each of the colored catchment areas as annotations on Figure 2, if appropriate, along with some sort of visual representation of how the JOIDES and Pennell areas were split.

On a related note, there is an underlying assumption here that there was no reorganization of flow during retreat. Yet, Greenwood et al., 2018 (Nat. Comm.) show clear geomorphologic evidence of ice retreating from what appears to be the Danielson and Bart “JOIDES inner reach” GZWs toward Mawson Glacier, which is located north of Ross Island and flows through the Transantarctic Mountains. As shown in Greenwood et al., the catchment area could be quite different than what is used for calculations in this paper, as it would involve a majority of the flow coming from the pink drainage area in Fig. 2, not the brown. Perhaps the JOIDES inner reach calculations should be supplemented with another test using a drainage area that corresponds to this reorganization.

Minor comments

As seismic interpretation is so crucial to the outcome of these calculations, it might be prudent to include a figure demonstrating a before and after interpretation of a seismic line in the primary manuscript, so that readers can see how a GZW might be recognized in seismic.

Supplemental figure 2 is far too small. It is impossible to read the writing when printed on standard-sized paper.

If I am being honest, the three maps in Fig. 3 all look identical. I have to squint really hard to spot a couple of contour lines that might be slightly different, and the colored depth scale bar in the legend is so condensed that I’m not gaining any information from it either. Unless there is a way to make the differences between the maps very clear, then I suggest removing this figure. I think the manuscript would be fine without it.

Be careful not to exclude the EAIS from the narrative, as it is a significant contributor to the Ross Sea and delivered grounded ice to a large percentage of the Ross seafloor during more glaciated times, just like the WAIS. I noticed the EAIS was left out of the abstract and a few other places where it would be relevant.

Stylistic comments

Some sections of this manuscript read very robotically, so just be on the lookout for areas where you could add transition words, vary sentence structure, and write in the active voice instead of passive. The abstract, in particular, should be revised with this in mind.

The Introduction needs some work in order to establish a narrative and show the broader research community why your work could be useful to them, so they will keep reading. As it currently stands, the Introduction launches immediately into Background information with no lead-in. The Introduction would benefit from a more deliberate structure. A suggested format: first paragraph that introduces the “Big Picture”—the reasons why Antarctica, the Ross Sea, and their geologic history are important. The second paragraph (a shortened version of your Lines 1 – 38) could give more niche background information that sets up the specific problem you investigate. The specific problem and previous attempts to solve it would be outlined in the third paragraph—perhaps Lines 45 – 47, followed by 38 – 43 and some explanation of what remains to be learned. A final, fourth paragraph could summarize what you do to solve it and could include Lines 43 – 45.

Consider expanding the Conclusion section (it is currently a little too perfunctory) to give space for suggestions of future work or returning to any “big picture” ideas from the introduction. Talk about the value of this study to the broader scientific community. Some transition words between thoughts could also improve the narrative.

It is a little jarring to read Ross Sea without “the” in front of it. I wouldn’t directly refer to the Caspian Sea or the Atlantic Ocean without the “the” in front of them, unless I’m treating them as adjectives (e.g., “We analyzed Atlantic Ocean sediment”), so it is strange to treat the Ross Sea this way. Do consider modifying the title, abstract, and other instances of this wording.

Line edits

Line 8: “greater” is an odd word choice—maybe say “more significant” and change “for” to “in.”

Line 24 – 25: change to “outermost” and add “the” before “western” and “eastern”. Add a comma after (wRS) and another after “Antarctica”

Line 25: Remove “In other words”

Line 26: Remove “At that time”

Line 27: Add a comma after “broad” and remove “and” so that it says, “broad, foredeepened troughs”

Line 31: change to “backfilled”

Line 35: Add a comma after (LAB)

Line 44: add “the” before Ross Sea

Line 45 – 46: Remove “the Ross Sea-wide” and add “in the Ross Sea” after “retreat”

Line 55: add a comma after “ages”

Line 56: remove “the” before Petrel

Line 70 – 74: consolidate these sentences to remove redundancies around the seafloor reflection and the unconformities associated with the GZWs.

Line 77: change “the Petrel software” to “Petrel” and add a comma after “meters” and remove the one after “primary input”

Line 81 – 82: join these sentences so it says “plotted in Petrel and interpolated to create depth and velocity maps”

Line 83 – 84: add “s” after “millisecond” and “meter”

Line 88: remove “the” before “QGIS”

Line 90: change “produce” to “product”

Line 98: For this and other similar instances, I recommend simply writing out “outer continental shelf.” Too many acronyms complicates this paper more than it needs to be, and it isn’t customary to do that for unofficial place names anyway. I found myself wondering what bank or trough was abbreviated as “OCS” because I forgot what it meant.

Line 102: remove “the” before “David” and do capitalize “Glacier”

Line 125: change “two” to “a” and add “two” after “and”

Line 127: “(Table 2; Fig. 4)”

Line 128: what is meant by “define part of the banks”?

Line 174: specify in Table 4 caption that the durations are measured in years

Line 180: edit to “The calculated durations suggest that the largest GZWs in the Ross Sea had stillstands lasting up to a few millennia”

Line 181: give a little more detail about these radiocarbon dates—what range of ages, what material, what facies?

Line 192: change “millennium” to “millennia”

Line 194: remove “the” before “millennial” and add “have” before “been outside”

Line 203: “data” is always plural, so change “suggests” to “suggest”

Line 205: “predicts”

Line 212: “ice streams”

Line 217: remove “100%” and add a more qualitative word like “completely”. Also add a comma after “grounded ice”

Line 226: change “Antarctica” to “Antarctic”

Line 234 – 236:  are these ages calibrated? They need the correct notation. If uncalibrated, they shouldn’t be compared to the calibrated ages. Also, for consistency, change the 8715 cal yr BP mentioned later in this paragraph to 8.7 cal kyr BP.

Line 248: change “small” to “low”

Line 259: if this is the first mention of MWPs, it should be spelled out

Line 262: change “current” to “modern” or “contemporary”

Line 263: remove “style”

Line 283 – 284: Perhaps say something similar to “…multiple episodes of small amplitude sea-level rise rather than a rapid increase in sea level from synchronous retreat”

Line 284: change “ground” to “grounding”