We are grateful to the two reviewers for their detailed assessments and also their encouraging remarks on our manuscript. We also thank Dr. Alastair Curry for providing additional feedback on the open forum. Below, we provide responses to their comments and describe the changes we propose to make to the manuscript so as to address all reviewer concerns fully.

Reviewer 1

The authors present 11 new Be-10 exposure ages with very tight internal consistency on glacial landforms from northwest Scotland that constrain a late glacial re-advance/standstill of a remnant of the British Ice Sheet. The authors also utilize independently published radiocarbon ages from a nearby isolation basin to provide stratigraphic support for the new ages. Conceptually, these radiocarbon ages at the transition from marine to lacustrine sedimentation should provide a minimum age constraint on the timing of deglaciation in the region as ice evacuated and the local land rebounded. The authors also spend considerable time discussing their previous Be-10 production rate calibration site to the southwest of this study area as there has been recent debate with another research group over the implications of deglacial ages calculated using this local production rate. I found the manuscript to be well-written with figures that are easily understandable and organized. I note below some issues I find mainly with the independent nearby isolation basin radiocarbon constraints as well as their approach for comparing ages using different PR sites. I would be supportive of publication of this manuscript in Geochronology after updating the manuscript to address my main comments and addressing other minor comments.

 Main Comments

1. I find the use of isolation basin radiocarbon constraints from a previous independent study lacking what I feel is critical context. From Simms et al. (2022), I note in section 3.1 of their paper that the radiocarbon constraints are from “sieving … and picking under a binocular microscope either large plant fragments, individual Charophyte oogonia seeds, or masses of small plant fibers and fragments” - are readers to assume that the “plant fibers” used to measure the two radiocarbon constraints recalibrated in this study (LBA18-11R, 530 and LBA18-11R, 536) are from the same freshwater species as the seeds collected, or because the samples are collected from the marine to freshwater transition, is it possible some of the sampled plant fragments are marine species and/or were capable of uptaking marine water? I note in the supplement to Simms et al. (2022) they present a diatom assemblage plot in Figure S5 that may help inform if there was a potential for sampling marine plant fibers. Any marine influence would drive radiocarbon ages down by as much as 1500 years (if using Marine20 and a default global reservoir correction of 550 years), which would have considerable implications for the interpretation of the overlap between these radiocarbon constraints and the new Be-10 ages.

Were those previous authors consulted on their results to confirm marine versus freshwater sourced macrofossils? As the authors state in this manuscript, and I fully agree, PR sites need to be carefully chosen and methods for constraining independent ages need to be as robust and meticulous as possible. The authors here need to specify with full confidence whether the radiocarbon gathered from previous work is from a fully freshwater source, or if marine radiocarbon could have potentially influenced the samples. Otherwise, it is difficult to accept the overlap between the radiocarbon ages and Be-10 ages as well as their interpretation that lower production rates produce better agreement at their site, which has significant implications for their interpretations in discussion in section.

As a minimum-limiting benchmark for local deglaciation, the radiocarbon ages published previously by Simms et al. (2022) are a foundation of this study and we appreciate that the reviewer would want to know more about their quality. We have discussed the nature of their ¹⁴C-dated organic samples with the lead author of that study (Dr. Alexander Simms), who confirmed that, although identifiable remains were used for some samples, formal identification of plant remains was not possible for many samples, including the basal samples (530cm and 536cm) used in our assessment. However, as highlighted by the reviewer and discussed with Dr. Simms, Simms et al. did report diatom assemblage counts for the Loch Bad na h-Achlaise record (summarised in their Figure S5), and these data provide valuable environmental context for the two basal ¹⁴C ages. Stratigraphically, both ¹⁴C samples are derived from green-brown organic muds that are compositionally distinct from the underlying grey minerogenic silts. The higher of the two samples, at 530cm, is from a layer characterised by 100% freshwater diatom species; the complete absence of salt-tolerant diatoms and the abundance of Halophobe species indicates there was minimal, if any, marine input during deposition of this sedimentary unit and thus it is very unlikely the plant material was of a marine origin. Sample 536cm, which is indistinguishable in age from that at 530cm, is from lower in the same sedimentological unit and also strongly dominated by freshwater diatom species (>90%), with only trace numbers of salt-tolerant (but not marine) species. Given these diatom assemblages and the stratigraphic and chronologic concordance of the two ¹⁴C ages, we conclude that these samples are unlikely to be of marine origin, thus justifying the use of the terrestrial IntCal20 calibration.

Regarding the approach for determining a preferred production rate calibration site to calibrate their ages, I urge the authors to take a more robust, quantitative approach. Perhaps visually, the Rannoch Moor site produces a tighter overlap with the radiocarbon constraints that adhere to the minimum age constraint rule (assuming my point above is addressed), but many of the age ranges using the different PR choices (and propagating PR uncertainty) overlap fairly well with the radiocarbon (especially in Figure S3) and visually produce uncertainty bounds and Be-10 ages that could also plausibly be older than the radiocarbon. I request that the authors demonstrate with statistical confidence if the respective populations of Be-10 ages and isolation basin radiocarbon ages adhere to the minimum age constraint rule from the radiocarbon (within total uncertainty, i.e., analytical + PR uncertainty; the horizontal whiskers in Fig. 8 & Fig. S3) for each PR calibration using perhaps a one-sided Welch t-test or equivalent approach. If a PR dataset produces Be-10 ages that cannot be proven with confidence to be statistically younger than the radiocarbon ages within uncertainty, the authors should note that for readers.

We agree with the reviewer that our evaluation of production rate output will benefit from the addition of statistical assessment. Consequently, we have performed both one- and two-tailed Welch’s unequal-variances t-tests for each derived dataset relative to the ¹⁴C benchmark and plan to include a description of those findings in the revised results section. Specifically, these tests (see the supplementary table included with this reply) confirmed what is shown in Fig. 8: production rate datasets based on calibrations made against independently dated surfaces (‘direct’ calibrations) give exposure ages that are statistically indistinguishable from the ¹⁴C control, while those based on calibrations made against surfaces of assumed age (‘indirect’ calibrations) give exposure ages that are statistically younger than the ¹⁴C control. The Global Primary ¹⁰Be dataset is a mixture of both (direct and indirect). Although the Global Primary dataset does technically pass the t-test, this is on account of the relatively large total scatter of that calibration dataset (6.9%). But, because that compilation includes individual sites that we have shown to produce exposure ages that are untenably young (i.e., Glen Roy, Isle of Skye & Highlands, Mt. Billingen, Norway), our findings suggest that using the Global Primary production-rate calibration dataset will also result in exposure ages that are too young stratigraphically if not statistically. We propose to report that observation in the revised Discussion.

In reference to the reviewer’s comment about Fig. S3, we stress that this figure is intended to illustrate how production rates based on independently dated (direct) calibration surfaces, regardless of how remote from Redpoint Peninsula, return exposure ages that are statistically and stratigraphically consistent with the ¹⁴C control. The results of the one- and two-tailed t-tests confirm this alignment (see table) and will be included in the revised supplement as a table of statistical output.

Line 67: I notice that some of the PR sites cited above (e.g., the Arctic PR from Young et al., 2013 and the site in Germany from Hoffman et al., 2024) are not included in the comparison, is there a specific reason why?

Given the mid-latitude setting of the Redpoint Peninsula, we have focused our assessment on calibration datasets developed in similar latitudinal settings (plus the Global Primary dataset), and for this reason decided not to include the Arctic rate of Young et al. (2013). That said, as a good example of a directly calibrated dataset, the Arctic rate would make a logical addition to Figure S3, in which we compare surface-exposure output from a suite of globally distributed, directly calibrated datasets; we propose to make that addition in the revised version and cite the Arctic dataset accordingly on line 414. The German dataset reported by Hoffman et al. (2024) is not included in our assessment because those authors highlighted the stark discrepancy between it and most other calibration datasets; we do note, however, that this production rate calibration dataset would produce ages at Redpoint Peninsula that are consistent with the ¹⁴C.

Line 135: as mentioned above, I highly suggest the authors ensure or report a specific justification for using intcal versus marine20 and/or the mixed marine and NH atmosphere calibration curves as the resulting PR calibration relies heavily on this.

Our full response to the reviewer’s concern is given above.

Line 264: this is appreciated context, but I am curious if the samples at this site are of a comparable lithology (and thus comparable erosion susceptibility) to the samples at the Rannoch Moor PR site?

The two sites exhibit different lithologies: out Redpoint samples comprise coarse-grained arkosic sandstones, whereas the Rannoch Moor samples are granite. Recognising the strong climatic similarities between the two sites (both are relatively low elevation maritime settings), it is of course possible that the different lithologies experience differing erosion rates. Observational evidence suggests that any such differences are minor; the 2-3 cm of apparent surface erosion over the duration of exposure on the Redpoint boulders (indicated by protruding pebbles) is broadly similar to the magnitude of erosion at Rannoch Moor suggested by protruding feldspar crystals (though these are minimum values only) and boulder form is similar. Recognising the challenge/impossibility of quantifying average erosion rates among multiple sites, we are confident that any such differences would not result in significant divergence in beryllium-10 concentrations between Redpoint and Rannoch Moor.

Line 371: here, as stated in my main point #2, I disagree that ages using at least some of the calibration datasets in the comparison (mainly NENA and the Globally averaged rate) necessarily produce exposure ages that are (within external uncertainty) statistically younger than the radiocarbon ages based on a one-way Welch’s T-test at 95% confidence. I did the test on a rough estimate of the means and external 1 SDs from figure 8 so it’s not necessarily accurate, so I would like to see what the authors find.

Our full response to the reviewer’s concern is given above. To summarise, our one- and two-sided t-tests reinforce the findings displayed in Fig. 8 and show that NENA gives a relatively good fit with the ¹⁴C, while the Global Primary dataset is statistically indistinguishable due to its relatively large total scatter. In the revised Discussion, we plan to suggest that, according to our assessment, the Global Primary dataset does not give the optimal fit to the ¹⁴C control due to the inclusion of sites where indirect calibration sites were involved (e.g., Isle of Skye & Highlands calibration dataset).

Line 525: more of a curiosity, but is there any potential hard water influence at the Redpoint Peninsula lake core site or is the site similarly dominated by the plutonic core? This could likewise have a potentially significant impact on the calibration at the Redpoint Peninsula.

The potential for a hardwater effect at Loch Bad na h-Achlaise is considered to be low, since the area is underlain by arkosic sandstones of the Torridonian Group and higher-risk lithologies (e.g., limestone) are absent. Furthermore, in discussing their radiocarbon chronology, Simms et al. (2022) specified that they explicitly avoided the most common sources of hardwater effects (aquatic algae, fine organic detritus, etc.) by sampling “only macro- and micro-plant fragments and not the bulk sediment, which tends to be the source of much larger reservoir offsets.” (section 5.2.1 of Simms et al., 2024). We propose to highlight this in the revised manuscript via a short addition to the Methodology (section 3.2).

Line 720: missing the Hoffman et al., 2024 citation

Thank you for noting this omission. We will add the citation to the revised references.  

References cited:

Hofmann, F.M., Rambeau, C., Gegg, L., Schulz, M., Steiner, M., Fülling, A., Léanni, L., Preusser, F. and Aster Team: Regional beryllium-10 production rate for the mid-elevation mountainous regions in central Europe, deduced from a multi-method study of moraines and lake sediments in the Black Forest. Geochronol. 6, 147-174, 2024.

Simms, A.R., Best, L., Shennan, I., Bradley, S.L., Small, D., Bustamante, E., Lightowler, A., Osleger, D., and Sefton, J. Investigating the roles of relative sea-level change and glacio-isostatic adjustment on the retreat of a marine based ice stream in NW Scotland. Quat. Sci. Rev., 2777, p.107366, https://doi.org/10.1016/j.quascirev.2021.107366, 2022.

Young, N.E., Schaefer, J.M., Briner, J.P., and Goehring, B.M.:. A ¹⁰Be production-rate calibration for the Arctic. J. Quat. Sci. 28, 515–526, 2013.

Thank you for highlighting these typographical errors. We will make the recommended changes in the revised version. 

Reviewer 2

The authors present a dataset of 10Be measurements from Northwest Scotland and use different 10Be production rate calibration datasets to determine which datasets yield 10Be exposure ages consistent with a previously published limiting radiocarbon age for deglaciation. The authors then take this a step further and assess the use of different production rate calibration datasets in Scotland (and the N Atlantic), concluding that a dataset from Rannoch Moor (Putnam et al., 2019) is the most appropriate for this and other studies. The authors also include an in depth discussion on deglaciation at Rannoch Moor, a subject that has been the focus of a significant amount of back and forth in the literature. 

The paper is a joy to read and includes many well-presented figures. I really like the straightforward approach of the paper, particularly the extensive use of ICE-D and the Washington online calculators. That being said, I agree with the first reviewer that a more quantitative exploration of the different production rate calibration datasets (/production rates) and the radiocarbon age constraint would strengthen the paper.   

Please see our responses to these concerns given above. 

Line 69 - should this be “archived *in* the ICE-D: Production Rate…” . 

We will make this change.

Figure 1 & caption - “AP” is used to represent both the Applecross and Aultbea peninsulas in the figure and caption. I suggest changing one of these so that we can tell which is which. 

Thank you for catching this error. We propose to change the Aultbea Peninsula tag to ‘AbP’.

Line 112 onwards - Exposure ages from Everest et al. (2006) and Ballantyne et al. (2009) are included here. Scaling schemes are cited for the Ballantyne et al ages, but not those from Everest et al. Are the Everest et al. (2006) ages also calculated using one of the Lm/De scaling schemes, and is an erosion rate used? It’s unclear if the ages have been calculated with the same inputs and are therefore directly comparable. 

The exposure ages reported by Everest et al. (2006) were not calculated or scaled in the same way as those reported by Ballantyne et al. (2009); the former employed the scaling factors of Stone (2000) following Lal (1991), which was subsequently termed ‘St’ scaling by Balco et al. (2008). Everest et al. (2006) discussed, but did not incorporate in their age calculations, the potential impact of surface erosion. The reviewer has highlighted that this information was included in our original submission and so we propose to adjust the text in this section to include these details. 

I have the same comment for the ages discussed in line 118 onwards (though you do mention “a range of CRONUS-Earth production rates”). However, I also get that part of the point being made here is the complex situation arising from different/evolving production rates being used over time, which, in part, helps necessitate the present paper! A comment on that might be helpful context here (instead of making sure each age talked about in this part of the paper is calculated in the exact same way). 

We agree that a short comment of this nature would help frame the issue more clearly, and propose to add it to the revised manuscript. It will also be straightforward, without cluttering the text, to specify the scaling models applied in each prior study.  

Line 116 - It might be helpful to include an age range (and maybe a citation) for the Older Dryas here. 

We will make this addition.

Line 141 - Same as above, it might be helpful to include an age range (and maybe citation) for Heinrich Stadial 1. I might’ve spent too much time thinking about the Southern Hemisphere, but I think this would add useful context regardless. 

Thanks to the reviewer’s comment, we found that the original manuscript included neither the definition of the acronym ‘HS1’ nor an age range (17,800–14,600 years before present [NGRIP Members, 2004; WAIS Divide Project Members, 2015]) for the event. Both will be added to the revised version.

Line 187/Table 1 - So that readers are able to fully reproduce/reduce the 10Be concentrations, I think it’s important to include Be blank information (e.g. AMS ratios and 10Be concentrations) in Table 1 or elsewhere. Also, how were the 10Be concentrations corrected? For example, was one blank correction used for samples extracted in Ireland and another for those extracted in Maine? Was there a batch-specific blank correction, some kind of running lab average, or something else? I don’t doubt that a sensible choice was made, but I think including that info here would be a helpful addition. 

We employed batch-specific blank corrections for all samples and agree that such information is important to include. The revised Table 1 will include all process blank information as requested.

Line 266-267 - Similarly to how the production rate calibration data sets haven’t been corrected for erosion (helping justify your lack of erosion correction, which I think is totally justified), it might be worth highlighting here that the production rate calibration data sets you use have also not been corrected for post-depositional uplift (helping justify your lack of uplift correction, too). I think that’s a point being made in this response by Young et al. (2020) (https://doi.org/10.1016/j.quascirev.2020.106329) which you could cite. 

We thank the reviewer for raising this important point and will add a statement to the revised manuscript concerning our choice not to include an uplift correction.

Line 329 - What’s the rationale for using St scaling here and in Figure 8 vs, for example, LSDn scaling? 

We employ the St scheme for the purpose of this paper because it exhibited the closest agreement among reference production rates determined from Rannoch Moor and other calibration sites from around the world based on independent chronologies (Putnam et al., 2019). We propose to add a clarifying statement to this effect in the revised section 4.3, while also noting that, given the proximity of Redpoint to Rannoch Moor, the three default scaling schemes used in version 3 of the University of Washington online calculator give results that agree closely with one another.

Figure 7 - The vertical bars are described as grey in the figure caption, but they seem more purple/lavender to me. 

Yes they do. In the revised version, we will adjust the colour scheme to ensure they are grey as described.

Line 427 - You mention a Table S3 here, but I can only find Tables S1 and S2 in the supplement. 

We thank the reviewer for noticing that Table S3 was missing from the original submission. This table (which includes all data needed to calculate the maximum-limiting production rate) will be included in the resubmission.

Line 473 - Should this be “which is located immediately adjacent *to* five of the seven…”?

Yes. We will make this change.

Line 529 - Sorry to sound like a broken record, but an age range for the Younger Dryas would be nice here, too. I think this is the first time you mention the Younger Dryas, and you use “YD” in line 561. It might therefore be useful to add “(YD)” after the first mention. 

Agreed. We will make these changes, specifically providing the age range 12,800–11,600 years before present for the Younger Dryas (NGRIP Members, 2004; WAIS Divide Project Members, 2015).

Line 614 - should this be “corresponds to a short-lived…”?   

 It should. We will make this change.

References: 

Balco, G., Stone, J.O., Lifton, N.A., and Dunai, T.J.: A complete and easily accessible means of calculating surface exposure ages or erosion rates from ¹⁰Be and ²⁶Al measurements. Quat. Geochronol. 3, 174–195, https://doi.org/10.1016/j.quageo.2007.12.001, 2008.

Ballantyne, C.K., Schnabel, C., and Xu, S.: Readvance of the last British Irish Ice Sheet during Greenland Interstade 1 (GI-1): the Wester Ross Readvance, NW Scotland. Quat. Sci. Rev., 28, 783–789, https://doi.org/10.1016/j.quascirev.2009.01.11, 2009.

Everest, J., Bradwell, T., Fogwill, C.J., et al.: Cosmogenic 10-Be constraints for the Wester Ross Readvance moraine: insights into British ice-sheet behaviour. Geog. Ann. 88A, 9–17, https://doi.org/10.1111/j.0435-3676.2006.00279.x, 2006.

Lal, D.: Cosmic-ray labeling of erosion surfaces: in situ nuclide production rates and erosion models. Earth Planet. Sci. Lett., 104, 424–439, https://doi.org/10.1016/0012-821X(91)90220-C, 1991.

Members, N.G.R.I.P.: High-resolution record of Northern Hemisphere climate extending into the last interglacial period. Nature, 431, 147-151, 2004.

Putnam, A.E., Bromley, G.R., Rademaker, K., and Schaefer, J.M.: In situ 10Be production-rate calibration from a 14C-dated late-glacial moraine belt in Rannoch Moor, central Scottish Highlands. Quat. Geochronol. 50, 109–125, 785 https://doi.org/10.1016/j.quageo.2018.11.006, 2019.

Stone, J.O.: Air pressure and cosmogenic isotope production. J. Geophys. Res., 105, 23753–23759. https://doi.org/10.1029/2000JB900181, 2000.

WAIS Divide Project Members, 2015. Precise interpolar phasing of abrupt climate change during the last ice age. Nature 520, 661–665.

Young, N.E., Schaefer, J.M., Briner, J.P., and Goehring, B.M.:. A ¹⁰Be production-rate calibration for the Arctic. J. Quat. Sci. 28, 515–526, 2013.