The paper by Arnaud et al. on the ‘Precise dating of deglacial Laptev Sea sediments via ¹⁴C and authigenic ¹⁰Be/⁹Be - assessing local ¹⁴C reservoir ages’ gives a nice demonstration of the potential of ¹⁰Be/⁹Be for correlations and independent age control. The Figure 2 shows good agreement between the ¹⁰Be/⁹Be fine structures of the Antarctic WAIS Divide core and Greenland GISP2 for the period 6000 to ~14500 BP after correction for climate influences. The normalized average ¹⁰Be/⁹Be record allows a global correlation for the record of the Laptev Sea sediment core PS2458-4 and, thereby, an estimate of the marine reservoir age (MRA) for the Laptev Sea for this time period. The ¹⁰Be/⁹Be ratio may thus be used in future quantification of past MRA values in high-latitude ocean regions not amenable to modelling.

I find this paper highly interesting because it shows the potential of ¹⁰Be/⁹Be as a new global correlation tool. It is therefore very well suited for Climate of the Past.

I do have, however, a number of suggested modifications/improvements that I hope the authors will consider before submitting a final version of this paper.

ΔR instead of R: The focus of the paper shifts from Marine Reservoir Age (MRA) R to ΔR (line 60-63) to discuss local deviations from a modelled global MRA that is, however, uncertain at high latitudes. The definition of MRA is the deviation from the atmosphere (globally well mixed¸ line 47-48). Since the correlation of the sediment core with the ¹⁰Be/⁹Be ice core record provides an independent time scale and thus access to the IntCal20 atmospheric ¹⁴C record, the local MRA at any place in the sediment core record follows from a simple comparison between the measured sediment ¹⁴C concentration and IntCal20.

Local ΔR range: Comparison with Heaton et al. 2023 modelled Laptev Sea MRA will give local ΔR values for this location near the Lena mouth at the edge of the continental shelf under changing sea level and climate that can be compared with the range of values (-100 to +800 yr) mentioned.

Figure 2: Figure 2 provides ¹⁰Be/⁹Be after climate correction. These assumed corrections are based on our best understanding of the ¹⁰Be production, its distribution, and local snow accumulation but the corrections may have flaws. It is thus interesting to compare the ¹⁰Be/⁹Be record of Figure 2 with the NORTHGRIP δ¹⁸O climate record and the IntCal20 Δ¹⁴C record (supplemental figure).

From 6000 to ~14500 yr BP there is detailed agreement between WAIS and GISP2 while beyond this, to 18000 yr BP the agreement is worse. This change in character closely coincides with the resumption of a strong Atlantic Meridional Overturning Circulation (AMOC) and the start of the Bølling in Greenland. IntCal20 shows a steep decrease in Δ¹⁴C slightly earlier, more coeval with the apparently opposite excursions in ¹⁰Be/⁹Be in WAIS and GISP2. Is this a problem of the Polar Seesaw? The discussion of one high-latitude MRA and one ΔR thus does not do justice to the data presented.

Comparison of the three records further indicates coincidence of the Older Dryas climate episode around 14000 yr BP with a ¹⁰Be/⁹Be low and a little increase in Δ¹⁴C. (AMOC?) The end of the Allerød/start Younger Dryas shows again coeval cooling and Δ¹⁴C increase with a possible¹⁰Be/⁹Be low but here the rapid changes in the ¹⁰Be/⁹Be record around this time require a more detailed synchronisation to be substantiated. A challenge to the authors.

Table S1: The uncertainty intervals of several of the dated samples straddle climate changes in Greenland. The known timing of these changes may, in combination with the sample position in the sediment isotope/climate record, be used to refine the dating intervals. In this respect it would be good to know more about the mixed benthic. Is there information about endobenthic versus epibenthic contributions?

Half-life: Line 66 gives the Audi et al., 2003 half-life of 5700 yr.  The value of 5730 yr. is still commonly used in reporting ¹⁴C results. Please state clearly what has been used in tables 2 and S1

Discussion: Robustness is important for calibration and can argue for a statistical use of an averaged MRA or ΔR value (line 309-314). Yet, the time interval considered includes large climatic changes, changes in AMOC, and sea level change and, therefore, large changes in MRA and ΔR are to be expected.

 It will be good to build the discussion on the record of changing MRA values over time, obtained from the direct comparison of the synchronized PS2458-4 record with IntCal20. Using a moving time window of 1000 or 1500 years to calculate a ‘temporal’ best fit instead of the full period may give robustness and flexibility and show differences for the H1, the Bølling/Allerød, Younger Dryas, and (Pre)Boreal periods.

Figure 4b: The Y-axis should read ¹⁰Be/⁹Be

This manuscript show an attempt to utilise cosmogenic radionuclide production rate variations to improve the time scale of a Laptev Sea deglacial sediment record and to assess the local marine 14C reservoir effect during this time period. This is achieved via the comparison of the authigenic 10Be/9Be sediment record and the 10Be variations in an ice core 10Be stack record.

I think the paper is well written and it shows a promising method to improve time scales and assess the marine reservoir effect. It is well within the scope of Climate of the Past and I recommend publication after some clarifications.

I think the marine reservoir age discussion has to be clarified. Usually the marine reservoir age refers to the 14C age difference between upper ocean (mixed layer) and the atmosphere. However, this study discusses benthic 14C ages at a present depth of about 1000 m (less during the deglaciation). The setting is not an open ocean setting and, therefore, I assume that the authors have good reasons to relate their 14C offset to Marine20. However, this is not at all explained and should be discussed thoroughly.

If I understood correctly, the authors assume a constant reservoir effect in their calculations. Is there any discernible trend in the reservoir age over the deglaciation and wouldn’t one expect a trend considering the changing setting (affecting so strongly 10Be/9Be).

The authigenic 10Be/Be record is dominated by a large trend and the residual variability appears to be largely within the measurement uncertainties (see e.g. Fig 4a where few points deviate from the trend lines exceeding their uncertainties). I recommend that the authors elaborate more if these deviations from the trend can be considered statistically significant.

The authors mention replicate measurements but do not seem to discuss them. I assume that they are shown in e.g. figure 4 but it could be discussed more (e.g. where the replicates separate samples from the same depth or e.g. replicate measurements on the same sample after leaching). To which extend do the replicates agree?