In this manuscript, B. Stoker and others present a flowset model and cross-comparison with geomorphic data across the deglaciated northernwestern LIS to provide a vast spatial and temporal analysis of ice streaming and retreat over the Bølling–Allerød and Younger Dryas periods. This work presents compelling evidence supporting collapse of the CIS-LIS ice saddle contributing to increased ice streaming and ice output reorganization during the Bølling–Allerød and varied styles of ice retreat following this collapse. I believe this work fills a knowledge gap in incorporating glacial geomorphic data with geochronology and ice modeling outputs for this region. However, I do think this work could be improved through further development of the interpretation and discussion sections with added supplemental materials to maintain focus in the final sections of the paper. There are some pieces in this paper that I think have merit but may not need to be included in this work specifically. Please see my structural and figure notes and thorough line comments to address areas in which this manuscript could be improved.

Specific structural comments:

Please provide more information about the ages used to constrain the flowsets and associated glacial features to specific timeframes. Including a table with this information or providing a section in the introduction introducing dates and context for geochronological situated retreat (Stoker et al., 2022) would greatly clarify a lot of questions I had about how some of your interpretations were tied to temporal constraints.

In general, watch the use of ambiguous identifiers. Often the use of “that” “this” and “they” can get lost in a line of logic, so make sure you’re being very explicit when making statements.

Section 3.2 should be split into two separate sections. One that describes ice streaming and the inferences made to topography and expand on the deglacial dynamics section.

Section 3.3 could be renamed to “Ice margin during retreat” to clarify that you are describing the ice margin boundary you developed and then the one you used for temporal constraints.

Section 5.1: for this first section to be an overview of ice flow interpretation, I don’t think the evidence from your results is connected enough. You make statements about ice behavior referencing your figures without tying in the significance of cut-through dynamics or the different types of flow sets you used to make these inferences. I see you do this throughout section 5.2 and am not completely convinced the overview in 5.1 adds to the readability and competes with the merit showcased in 5.2. I would suggest either switching the order of sections 5.1 and 5.2 or simplifying/removing 5.1 and combining it with section 6, which essentially restates a lot of the overview but with better supported evidence.

I would consider renaming your section 5.2 headers. I understand the interest in continuity across sections, but I think readability and interest may increase if you give headings related to why the time slices are split the way they are. For example: “5.2.2 Local LGM (17.5 to 17.0 ka)” or “5.2.3 Ice stream network reorganization (16.5 to 16.0 ka)”.

Section 6.4: I feel like this section does not incorporate many novel results that you have not already discussed in this paper. I am fine with the 6.3 sections on ice retreat variation but feel like a one to two paragraph summary of this work could be incorporate prior to section 6.5 Additionally, in the 6.4.2 section on the Canadian Shield, this ends up being a discussion of local responses and climate responses to the Younger Dryas and no evidence is presented from the Canadian Shield. If you keep this section, I encourage including information about basal conditions on hard bed surfaces that would impact a slow in ice movement. You also state ice stream slowing is not a result of geology changes, but that directly contradicts what you say in the introduction to section 6.4 and is counterintuitive to the 6.4.2 section header. You address this well in the first paragraph of 6.4.3 so I suggest you remove 6.4.1 and 6.4.2 and focus on results from 6.4 intro and 6.4.3.

The conclusion could be much more concise—highlight the main findings without naming the specific evidence.

Figure and table comments:

Figure 1E: If you’re going to create a hypothetical flowline model with example flowsets, I would either make a schematic of factors used to determine flowset styles in parts A-D or develop a hypothetical model that includes the same description for all flowset types (maybe with some extra text or drawn geomorphic features on the figure to explain differences). This may also help clarify the differences between event and ice stream flow sets which I believe needs a bit more explanation.

Table 2: “Ice marginal position” should be renamed to “Deposition process” as you are describing how these features are developed. In the terminal moraine section of “Ice marginal position” I recommend you refine the statement “deposition” to “deposition by bulldozing or transit to the margin” because the process of “deposition” is very broad. Can you give some examples of what you mean by “a combination of processes” even if it’s just to mention plucking, meltwater redistribution, etc.

Figure 3. Can you increase the contrast of the DEM? It is hard to identify the individual features you have mapped. Please also clarify the definitions between esker complex and ridges (i.e., individual eskers could not be identified and rather there is a complex network of meltwater features) and your identification system for major vs. minor moraine crests. Perhaps you could zoom in on some of these features or provide specific DEM examples in Table 2.

Figure 4: it is difficult to see the difference between colors of deglacial and inferred deglacial flowsets. I understand the draw of using two near-similar colors because of the similarity in flowset formation, but especially for colorblind readers, I can imagine this difference would be too minuscule to be able to visualize.

Figure 6: I appreciate what I’m sure was a considerable amount of time and effort in developing Figure 6 – that is an incredible amount of data to visually represent across such a vast spatial area. My only suggestion would be to possibly increase contrast between flowset and ice extent colors – sometimes the flowsets are difficult to see over the intense blue of the ice and the dark DEM underneath. I also have trouble reading the flowset numbers at times, I would consider creating a close-up map of some highly congested areas (section I) in a supplemental figure.

Figure 7: This is great – I appreciate the visual representation here. I would consider making this figure supplemental, though. It is not entirely central to the argument you make in the section it is discussed.

Figure 8 seems to be non-essential to the arguments made in this work. I would recommend combining Figure 8 and Figure 5 to relate concepts of ice-margin interactions with glacial lakes or move this figure to a supplemental file.

Figure 9 caption: Mackenzie is spelled incorrectly in (B). Clarify what you mean by “topography”.

Figure 10D: what is a “zig-zag esker”?

Line comments:

Line 14: Suggest change from “it” to “this retreat” or something similar.

Line 32: clarify calving for marine-terminating ice systems, the first part of the statement could refer to terrestrial and marine ice streaming.

Line 41: suggest change from “it” to “Laurentide Ice Sheet” could read as ambiguously referring to the North American Ice Sheet complex.

Line 48: clarify the ablation area of the northwestern sector.

Line 49: Direct evidence for the statement in the first half of this sentence?

Line 57: Suggest add “geomorphic-based evidence of ice stream activity”

Line 83: Clarify “the LIS and CIS ice sheets through the saddle”

Line 115, remove commas around undated

Line 126: “Range and were dammed [...]”

Line 132: Suggest “has led [...]”

Line 134: add parentheses timing of Younger Dryas Stade for context

Line 136: here is this argument still incorporating an early 30ka maximum, or are they arguing a total maximum later, at 20ka? Please clarify this point as these two comparative sentences are not exactly congruent.

Lines 150 and 152: The ice sheet wide and across the entire ice sheet in the same sentence is redundant.

Line 155: Reference figure 1 after “Smoking Hills-Horton River area”

Line 160: Cordilleran to CIS

Line 169: Clarify whether the “uniform mapping approach” was a manually conducted mapping effort or if there were automated tools involved or machine learning approaches to landscape analysis.

Line 184: clarify what you mean by “morphology of the flowset”. Do you mean here you are identifying the types of streamlined bedforms? Elongation ratios? Orientation and parallel conformity?

Lines 194-197: The comparison between these three sentences is a little difficult to follow. I would add a contrasting argument before introducing the inferred deglacial flowset (i.e., “Conversely, inferred deglacial flowsets [...]” and in the final sentence describing the fan-shaped lineations, I would suggest saying “The proposed ice-marginal formation of inferred deglacial flowsets is based on [...]”.

Line 198: There needs to be more clarification on the difference between ice stream flowsets and event flowsets and possible overlap between the two. Both have abrupt lateral margins, both could occur on the interior, both may be overprinted, and the elongated nature of the event landforms is unclear. I would choose several classifying characteristics for each of the flowset types and make sure you identify the characteristics for all the flowsets so that they may all be directly contrasted and compared.

Line 198: I would also like more clarification on the name “event” for these flowsets. Is this suggesting that these flowsets were developed very quickly in a singular streaming event and were then discontinued? This should be clarified in the definitions.

Line 277: Consider looking at ICE-D and AskICE-D for standardized and recalculated cosmogenic nuclide exposure dates – this is a global dataset that has many CIS and LIS ages from published work that can be compared.

Line 285: I commend the authors on making all shapefiles available from the paper.

Line 297: You name deglacial flowsets as the second-most prevalent flowset type yet have not named the first most prevalent type yet. I would make this clarification or reorder the presentation of flowsets otherwise this statement seems out of place.

Lines 302-307: This interpretation of topographically influenced streaming seems like it may fit better in discussion where you will have more room to justify this argument. Based on what is currently in these sentences, I am not quite clear on how you are making this interpretation and if the bedforms in the foothills of the Mackenzie Hills are more affected by topography than those on the Northern Interior Plains. Additionally, you distinguish between drumlins and mega-scale glacial lineations but did not explicitly state your classification guidelines in the methods. I would assume you used Clark et al’s 2010 identification of a 10:1 elongation ratio difference between the two bedform types, but if that is the case, then I would mention this somewhere in the methods. The sentence between lines 305 and 307 is difficult to follow. I would recommend splitting this up into more than one statement.

Lines 318-319: Again, an interpretation of the influence of local topography on diverging flow patterns.

Line 329: You use the term “topographic influence” broadly several times in this results section without necessarily defining it or naming the different influences you identify (e.g., funneling from valleys, divergence of flow around bumps in the bed, or separation between flowsets by ridges) please provide examples and be specific when discussing your identified flowsets. Please see McKenzie et al., 2022 for resources regarding topographic and lithologic influence and streamlined subglacial bedforms and McKenzie et al., 2023 for evidence of subglacial bump influence on streamlined subglacial bedform morphologies.

Line 340: This relates back to another comment about event flowsets, but what do you mean by “different events” in this context. Please provide examples (i.e. surging events, subglacial lake outbursts providing lubrication to the bed, etc.).

Line 341: please provide the “n=” value for all datasets, not just the unclassified flowsets.

Line 370: cite the “previous studies” you are referring in this line.

Line 375: I would clarify “Terrestrial ice-contact landforms” here. There are areas where marine-terminating ice contact landforms are visible at the surface from landscape evolution, so I would just be abundantly clear.

Lines 406-407: Claiming that the geomorphological evidence does not support ice stream activity during the same time as the Amundsen Ice Stream feels rather unsupported. This would take geochronological data to support the timing statement. Your argument of deglacial facies overprinting ice streaming is fine, but here is a good opportunity to pull in your findings of the most common flowset being the deglacial flowsets.

Line 414: I think further supporting this sentence would make it stronger (i.e., “[…] with periods of slow retreat as seen from the presence of periodic recessional moraines and some larger moraine crests […]”).

Line 414: Do you mean meltwater drainage by ice drainage, or do you mean ice drainage through ice stream networks? Make this clearer.

Line 415: “these changes” be clearer, are you talking about the periods of rapid ice loss or long-term stabilization that could have caused further erosion/transport to the margin?

Line 420: Were likely not active at the same time as each other or at the same time as the Amundsen Gulf Stream during early deglaciation? Make the connection between this sentence and the prior more explicit.

Line 425: I understand the connections here, but I think you need to make them a bit more explicit to readers. I understand how you determine which flow came first through crosscutting relationships, but to make it clear this occurred as a result of the Bølling–Allerød, I suggest framing the argument like “Without the source from the ice saddle, the ice streams of this region began to primarily receive input from the Keewatin Ice Dome to the west as seen in the northerly flowsets transitioning to northwesterly and eventually westerly-facing signatures of ice flow.” You do this in later lines but I would move this up to the start of this argument.

Line 444: Connect to the flowset data – potential to include “which is seen to be associated with more truncated streamlined subglacial bedforms and inferred decrease in ice flow speed and subglacial sedimentation organization (McKenzie et al., 2022).”

Line 445: the last paragraph in this section could be moved to an introduction between header 5.2 and 5.2.1 to make it more fluid for readers.

Line 470: multiple younger flowsets?

Line 509: I would break the sentence after “deglaciation” – it took me a few times of reading this to make sense of the additions to the first statement.

Line 513: Add “between 16.5 to 16.0 ka” to the end of this sentence for clarity of that is the time slice in which you’re referring.

Line 521: Change “this” to “the later flowset”

Line 523: Change “collapse of the ice saddle during the Bølling–Allerød” to “collapse of the ice saddle, which occurred during the Bølling–Allerød” for clarity.

Line 529-532: This statement is either not well supported or could be written better. If you are stating that the ice margin shows a slower ice flow regime that transitions to a faster ice flow regime as you spatially move upstream into the Mackenzie Valley, then please clarify you’re referring to the spatial variability across the region that was occurring simultaneously. If you instead are stating that the slow ice flow regime becomes a faster flowing ice regime at this single location over time, I think that statement needs more support, specifically in stating how your flowsets capture that variability (e.g., because more and less elongate elongate features co-exist in single ice stream systems (McKenzie et al., 2022)).

Line 565: What do you mean by “margin retreat was active”? Please clarify if you’re referring to sedimentary processes deforming the bed near the margin or timesteps of retreat or something else.

Line 622: I suggest this be changed to “overprinted on flowsets derived directly from the Great Slave Ice Stream.”

Line 629: “topographic complexity led to complex cross-cutting flow patterns” – maybe clarify what you mean by “topographic complexity” to reduce the use of the word complex in this sentence.

Line 633: What does “this” refer to? The topographic complexity, the complex cross-cutting relationships, or the esker? Please clarify.

Line 638: In the beginning of this section, you name the geochronological tool used to determine the timeslice. I recommend you do this somewhere in all other 5.2 sections. I assumed it was all using cosmogenic nuclides, but after it being explicitly stated only here, now I am not sure.

Line 705: References at the end of this sentence?

Line 720: At the end of this sentence maybe add something to the effect of “but it is unclear whether this mechanism of drawdown is strong enough to weaken the entire ice-saddle” to better tie this observation to the following paragraph and the opposing arguments.

Line 723: Be clearer with the word “topography” – a topographic high? Of what size?

Line 752: Provide examples of what regional stagnation would look like in the deglacial record like your examples for active margin retreat.

Line 771: What about the hummocky terrain? If there are any other possible explanations for lack of ice marginal landforms, these should be presented here as well.

Line 798: Evidence that these moraines are from the end of the Bølling–Allerød?

Line 829: The “Instead” at the beginning of this line makes the argument confusing because you used a “but” previously. Please make these two sentences clearer. Also, please provide evidence or citations for the flowsets you use to assume lower retreat rates and slower ice velocities (are the flowsets less elongate?).

Line 853: Just say “We reconstruct both extensive and shorter time transgressive […]”

Line 867-869: This mention of crevasse-fill ridge networks could use some clarification. How does the presence of these features indicate surging behavior?

Line 875: I would add a “However’ at the beginning of this sentence because these statements contradict each other. Also add a reference at the end of this sentence.

Line 889: Expand on the explanation for this. Maybe include something like “allowing for basal shear stress to increase and stabilize the ice during retreat and slow streaming.”

Line 908: clarify that “they” refers to geological conditions?

Line 921-922: This is not a complete statement. Please clarify this sentence.

Line 961-965: These lines contain a lot of statements that contradict each other. I would simplify this to say ice streaming occurs across the Canadian Shield only after xyz circumstances are met. I would also make sure you don’t mention earlier that there is not ice streaming on the Canadian Shield because there is some back and forth in this and previous sections.

Line 968: I take issue with naming “glacial lakes” as a more important control on ice stream formation than subglacial geology. What is the mechanism that is the stronger control? Because if this were marine-terminating, I argue it would be the same pattern, so it’s not the lakes but perhaps the onset of crevasse-driven ice loss, or increased ice breakage from loss of buttressing. Also, how does this tie into the topography? Can you say anything about the role of topography in relation to the geologic control. I think it could also be argued that the lakes are a function of topography because the topo has allowed for lakes to develop, so does your argument inherently agree with Winsborrow et al., 2010’s argument that topography has a higher control on ice behavior than geology?

The paper by Stoker et al. provides a detailed assessment of the geomophic features eroded and deposited by the NW sector of the LIS.  The authors provide detailed flowlines of the LIS from the LGM through the last deglaciation and from these reconstruction attempt to interpret the style of deglaciation of this sector of the ice sheet.  Overall, I find what the authors have done to be worthwhile and useful for interpreting how this sector of the LIS transgressed from the LGM to end of YD.  However, I find it difficult to ascertain how the authors have arrived at their conclusions about the time-transgressive nature of the LIS demise without the chronologic information provided alongside their reconstructions.  To me, this is the major weakness of the paper and without this information it does not allow the reader the ability to properly assess their interpretations, or at least easily assess it without digging through other publications and comparing and contrasting.  I think with this information provided in the figures and within some of the text, it will improve the paper substantially and give readers a very nice assessment of the history of this part of the LIS alongside their geomorphic mapping and interpretations.  My detailed comments are provided in the accompanying PDF with other suggestions for improving the manuscript.