This manuscript provides a new vantage on Late Pleistocene variability of the southern margin of the Fennoscandian ice sheet, drawing primarily from a suite of OSL ages on periglacial sediments exposed in the Rożental gravel pit. A number of less-compelling beryllium-10 surface-exposure ages augment the chronology, though owing to the considerable scatter in the latter, they provide only background support for the overall glaciologic story. The authors link apparent stepwise retreat of the ice sheet margin with variability in meltwater discharge into the Atlantic via the Fleuve Manche, a robust marine-geologic indicator of terrestrial meltdown. Altogether, this is an intriguing dataset that is presented in a clear and transparent manner, including excellent figures and thorough descriptions of all methodologies. As such, it will make a valuable contribution to the literature and is suitable for publication in The Cryosphere. Before I can recommend publication, however, there are a number of issues that need to be tackled; these are not altogether onerous but will require the authors to reconsider some of their interpretations of the data. I list them below (along with general comments) in the order at which they present themselves in the text.

Abstract: Look out for typographical errors. This is a fine introduction/synthesis of the manuscript but the incomplete English distracts the reader’s attention somewhat. The same goes for the rest of the manuscript, which is well written but would benefit from fine tuning by a native English speaker.

Figure 1: Legend is a little confusing. The black circles with values are recalculated ages, yes? What are the black circles without values? It would also strengthen the figure to justify the positions of the LGM and Pomeranian ice margins – whose work confirms that these positions are correct? What about chronology? Without those details it looks a little like guesswork.

Line 110: Such things are subjective. How big is massive? Give approx. size.

Figure 2 caption: Do you mean 'diluvium' (or outwash)?

Figure 3: The figure is most illustrative. I suggest, for added clarity, that you specify that the yellow ‘V’ shapes in panel B and the frost wedge casts.

Section 3 is very detailed, which is excellent. Such attention to detail can help make your paper a methodologic resource for future work. On line 194, suggest you replace ‘decontaminated’ with ‘isolated’, as contamination in the cosmogenic workflow has very different connotations.

Section 3.2.1.: Why are you using the Borchers global production rate? The choice of PR is, of course, entirely up to the researchers but given the number of rates now available to use, every choice requires a strong justification in my opinion. This is particularly the case when the Borchers rate contains obviously flawed production 'calibrations' from Scotland. As detailed by Putnam et al. (2019), the inclusion in the global set of production rates calibrated against surfaces of assumed (i.e., not actually known) age artificially skews that global average rate, making it a little too high and thus the resulting ages unrealistically young. The CRONUS team themselves noticed the weird impact of the Scottish rates on the overall average (Phillips et al., 2016), pointing out some anomaly in that area, but couldn’t (or wouldn’t) pinpoint the lack of robust independent dating as the cause. This paper could be improved significantly, therefore, either by using one of the robustly calibrated production rates from the northern mid latitudes (including Europe) or by using the CREp calculator (https://crep.otelo.univ-lorraine.fr/#/), which allows users to remove dubious calibrations from the global primary dataset. I am immediately sceptical of Late Pleistocene studies that blindly use the published Borchers rate, particularly without a very strong justification.

Line 238: I think the paper needs a clear statement (at some earlier point, not here) about the relevance of an OSL age and the sediments it is dating. My understanding of your stratigraphy is that these are minimum-limiting ages for the RzB2 unit, since the wedges were emplaced into that deglaciated surface. Therefore, the OSL ages do not date the emplacement of the till itself (advance) but the abandonment of that till surface (retreat), and the lag between deglaciation and ice wedge growth/infill cannot be known. Unless I am missing something fundamental here (quite possible), all your OSL ages are minimum ages and thus should be explained as such right from the outset.

Figure 5b: As the authors point out, the Be-10 ages are not very consistent. This is a huge range, particularly in this day and age. Can authors tell us more about why these specific boulders were sampled? Were they all on specific moraine ridges, or are they randomly distributed? In other words, what is their significance? I’m pleased the authors include the beryllium dataset, for it shows complete transparency and no desire to hide ‘ugly’ data, but a little background would help readers understand better the rationale for sampling and also the potential problems.

Lines 276-277: This pattern, should it be correct, is intriguing as it pops up in glacial records worldwide. Net retreat (warming) following the LGM was punctuated by brief pauses or readvances (cooling) that could have existed for just a few years/decades. If the pattern is more widespread than just central Europe, what does that tell us about its climatic drivers and importance? A broader exploration of this possible pattern is warranted, beyond the borders of Poland; if you don’t explore it, somebody else will.

Line 286: ‘Exposition’ is not the correct word here.

Line 294: I fear the relevance of the OSL ages is being overstepped here. These K1 OSL ages are minimum ages for the till itself, yes? Because the K1 wedges have been emplaced into the till surface following deglaciation. You need to make that clear, since the Rz2a till presumably predates the wedges. Yes, you've done bayesian statistics on this age set to get ages for the till, but the truth of the matter is that, from the real data themselves, there is only minimum-limiting age control for the basal till and it could easily postdate 19 ka. Indeed, knowing what we now do about the shape and duration of the LGM, this till could have been emplaced anywhere in MIS-2. Deglaciation subsequently presented a subaerial surface on which periglacial landforms could develop. I’m harping on about this because I think it is important; you could be misrepresenting the age of the basal till by multiple millennia, and that in turn could skew the common perception of when the LGM occurred in these parts (e.g., there is an important difference between, say 24 ka and 19 ka). Already, on Line 338, you are ascribing a concise-sounding age for the basal till (19.1 ± 1.1 ka) that cannot be ascribed based on minimum-limiting OSL ages.  

Figure 6: In panels B and D, how do you know the ice margin retreated outside the study area like this? What is your geologic evidence for that? I don’t see any described in this manuscript. Likewise, in panels C and E, what is your geologic evidence for the ice margin having stabilised at these tidy blue lines? Are there conspicuous moraine complexes defining a robust, stable ice margin, or is this conjecture? If the latter, please specify and use dashed lines rather than filled; otherwise, folk might take this as true when it could be little more than speculation. I assume the glacial geology of these parts has been thoroughly mapped?

Line 316: Yet, this is essentially speculation. Again, please be careful, as Figure 6 will give the impression that these speculative margins are based on unequivocal geologic mapping - and they aren't. Again, I suggest you use dashed lines instead and state specifically in the caption that these are entirely speculative.

Lines 327-329: Again, there’s a lot of speculation here. Why not include some detailed mapping? That would make this a much stronger contribution.

Line 348: Again, this should be defined as a minimum-limiting age.

Line 385: I understand the logic here, but this relationship is again highly speculative. What is the global nature of this inferred readvance? How is it possible to correlate with a Heinrich event if we don't, as a community, yet know quite what causes H events? (Or why some stadials include them and others don’t, or some H events occur without stadials?) And when accurate and precise C-14 dating of H events in marine records continues to elude us due to reservoir uncertainties? I think this is saying way too much given both the uncertainties in your data (excellent OSL ages, but they do have sizeable error bars, as is expected) and the persistent lack of understanding regarding H events themselves. I’m not saying don’t hypothesise, rather I just urge more caution.

References cited:

Phillips, F.M., Argento, D.C., Balco, G., Caffee, M.W., Clem, J., Dunai, T.J., Finkel, R., Goehring, B., Gosse, J.C., Hudson, A.M. and Jull, A.T., 2016. The CRONUS-Earth project: a synthesis. Quaternary Geochronology, 31, pp.119-154.

Putnam, A.E., Bromley, G.R., Rademaker, K. and Schaefer, J.M., 2019. In situ 10Be production-rate calibration from a 14C-dated late-glacial moraine belt in Rannoch Moor, central Scottish Highlands. Quaternary Geochronology, 50, pp.109-125.

This manuscript provides new information on Late Pleistocene ice margin fluctuations at the southern fringe of the Fennoscandian Ice Sheet. The new evidence comprises 11 OSL measurements from fluvioglacial and periglacial deposits exposed in the Rożental gravel pit, and 5 ¹⁰Be boulder surface exposure ages, all derived from a 30km² area of the moraine plateau in northern Poland. Unfortunately, the ¹⁰Be ages do not contribute to the discussion, however it is important that they were presented here for sake of transparency.

The millennial scale fluctuations are based on the Bayesian modelling of the OSL data from Rożental gravel pit, where 3 glacial tills separate periglacial sedimentary features. The authors link their millennial scale fluctuations to evidence of meltwater sediment deposition in the Bay of Biscay. This apparent relationship is speculative but interesting. Overall, the manuscripts presents new terrestrial data on ice margin fluctuations in northern Poland and is well within the scope of TC. The paper reaches substantial conclusions which are backed up by the evidence presented. The methodology is well described and all relevant data necessary for readers to recalculate the ages is provided. References are appropriate, but some additional ones are suggested for Figure 1. The supplementary materials are generally good, boulder sizes are missing in Table A4.

Below I have attempted to correct the grammar and also make more detailed comments. LXX refers to Line XX in the submitted manuscript.

Suggested edits:

L15: Optically stimulated luminescence (OSL) was used to date…

L17: 10Be surface exposure dating was used…

L18: …resting on the surface…

L23:  “confront” is not the right word, compare is better

L35: The sentence “The last two ice sheets….Rignot et al., 2019).” can be removed. It does not add anything meaningful.

L40: geological records

L41: enables correlating

L43: glacial records…nature of the glacial

L48: see comment for L15, or use OSL

L49: see comment for Line17

L52: see comment for L23

L59: …within an elevated…

L62: denivelations should be relief

L64: sheet’s should be sheet

L65: The sediment outcrop where…

L78: …of a well-preserved…delete “a”

L86: …and it consists of…should be …and consist of…

L109: …most of surface… should be …most surface…

L110: massive is subjective, give dimensions

L110-111: Suggest changing “, which commonly occur at the surface of moraine plateau,…” to “located on the moraine plateau,…”

L114: deluvium, or diluvium, is an obsolete term. Is this alluvium?

L120: The sequence…

L122: “reveals” should be “show”

L125: “with a fossil” should be “with fossil”

L142: PCV should be PVC

L147: “remaining of other” should be “remaining other”

L157: “decay chain and potassium” should be “decay chains of potassium”

L158: delete “the reference materials, namely”

L164: “consists of sequence of sediments units” should be “consists of the sequence of sediment units”

L166: “sediments deposition” should be “sediment deposition”

L166: likehood  should be likelihood make sure you check every occurrence.

L172: “sediments deposition” should be “sediment deposition”

L175: “pre-date particular event” should be “pre-date a particular event”

L184: “with IntCal20” should be “with the IntCal20”

L187: see comment for L110

L190: “lithologies as” should be “lithologies such as”

L194: “decontaminated” should be “separated”

L232: “samples MAM” should be “samples the MAM”

L133: “as Rz1” should be “as the Rz1”

L234: “It is visible especially within aliquots distribution of…” should be “This is especially visible in the aliquot distributions of…”

L236: “deposition of Rz1 unit” should be “deposition of the Rz1 unit”

L241: “dominates” should be “dominate”

L242: “revealing” should be “shows a”, “with CAM” should be “with the CAM”

L248: “with CAM” should be “with the CAM”

L286: What do you mean by boulder exposition after deglaciation and/or significant postglacial erosion… How much erosion would you need to explain the young age? Why would this boulder erode so much faster than the others you sampled? Please provide a reasoned argument for your explanations of the young age. The most likely reason for the young age is that the boulder has moved or that it was buried and subsequently exposed. Could this be the case?

L320: “of boulder” should be “of a boulder”

L341: “for Rz1” should be “for the Rz1”

L342: “of Rz2a” should be “of the Rz2a”

L349: “formation” should be “formations”

L350: “in north-“ should be “in the north-“

L351: “be also related to regional” should be “also be related to a regional”

L354: “a large” delete “a”

L361: ~17 ka and” delete “and”

L366: “record” should be “records”

L368: “been already” should be “already been”

L376: “sediments” should be “sediment”

L377: “sediments” should be “sediment”

L403: “record” should be “records”

Figure 1 caption: provide references for the LGM and PM ice margins. The outline script for the LUB samples and Rożental reduces clarity. Either use solid white or increase the size of the labels. The Gd coring site label in the legend does not match the style on the map. Label the grey dashed box as Figure 2.

Figure 2 caption: What is the grey dashed box? What is deluvium? If you mean diluvium, it is an obsolete term. Is this material fluvial or glacial? Please check appropriate terminology.

Figure 5 caption: “(a) Sediments profile” should be “(a) Sediment profile”

Figure 6: Drawing ice limits on maps is one of the hardest processes to justify to the community. The process probably generated quite a lot of discussion and various versions of the maps. What are the uncertainties of the isochrons on the map, and how were the locations of these limits established. The explanation in the text suggests ambiguity “not equivocal” about the exact position. Could this ambiguity be shown on the map with a maximum/minimum limit?