Dear authors

Thanks for your careful revisions. I am overall satisfied with your changes, but have a few more comments, which require some more work before acceptance of the paper in CP.

The most important point is the criticism of reviewer #1 and my own assessment that it represents a problem that the Hulu cave record you are using, is not publicly available yet. Moreover you cite also a series of other papers “in preparation” or “in review”, which are not available. This issue has also been discussed by the chief editors of CP and we come to the following conclusion:

In case of papers “in preparation” or “in review”, which are not at the heart of your manuscript, papers “in review” can be cited like that, papers “in preparation” should be better cited as “personal communication”. As for the Hulu record, this represents an absolutely crucial part in your paper and it is not acceptable that the readers of CP cannot replicate and assess your synchronization. The references you are citing with respect to Hulu are unfortunately not sufficient as the Southon paper only covers the last 27 kyr. Accordingly, it is essential that you include a plot with this record that illustrates the synchronization, favourably in the main text of your manuscript or as a supplement. Without this, acceptance in CP would have to be delayed until the new Hulu record is published elsewhere.

My further comments are minor and are only meant to further improve an excellent paper:

Line 90: you could add here the EPICA, 2006 paper
Line 157: I assume you mean „replicate samples“, if so please specify here

Concerning your comment on measured Ca concentrations in reply to my previous editoral comments, the problem is a potential high blank value, which would not be in the ice, but would be added externally during the measurement. This is why Freitag et al use the LOD as lower limit, below which the Ca concentrations are not „in the ice“. As you use a similar CFA method (although based on ICPMS) I am quite confident that your limit of detection (LOD) or blank contribution is not substantially higher. So I do not expect there is a problem but you should specify the LOD in your manuscript.
Unfortunately this information cannot be found in McConnell 2002 and 2007 for Ca derived using ICPMS.

Line 265: para-metrizes
Line 271: The sentence in parentheses wasn’t clear to me. I would just delete it
Line 275&276: the LGM
Line 299: Cuffey et al, in prep (see above)
Line 408: sensiti-vity
Caption figure 3 last sentence: means

Concerning your reply to my comment on the paper by Bendel et al.: I have to apologize for being sloppy. What I meant is that the bubble distribution evidence for a stronger layering in glacial ice by Bendel et al. is most interesting but appears yet to be circumstantial. I could imagine that fast reformation/relaxation could be an issue in this ice, where bubble nucleation may be strongest in dusty layer, which then would not necessarily imply different density/porosity layering during bubble close-off. There is no action needed from your side. The wording should only make clear that a stronger layering in the glacial is not a proven fact yet.

Line 580: core, where … not available, we
Line 617: time series

Line 721: I would suggest a sentence like the following one: “Note that this approach represents only a first order correction of a growing offset between GICC05 and Hulu, while non linear temporal changes in the counted dating error may exist from one tie point to the next (Fleitmann, D., et al. (2009), Timing and climatic impact of Greenland interstadials recorded in stalagmites from northern Turkey, Geophys Res Let, 36(L19707), doi:10.1029/2009GL040050)

Line 783: the stretched GICC05
Line 785: wrong Figure number

Another comment from my side concerning the 2nd paragraph on page 13: The deviation is largest below 60 kyr, i.e., below the counted part of the GICC05 age scale and below the last Hulu tie point. Accordingly, the age of the WD-NGRIP tie point at 65 kyr may be not so well constrained as the other tie points, having a direct influence on the accumulation between 60-65 kyr. May be, you want to mention that.

Line 844: I would suggest to add a black dashed line in Fig. 7c that does not include the uncertainty in absolute ages.

Line 1000: Is there a reason why you use a different equation here than Freitag et al., 2013? It would be nice , if you could show how one can come from equation (4b) in Herron-Langway to (4c). In equation (4c) it actually says dp/dt not drho/dt in the Herron-Langway paper, I assume it is a typo in the Herron Langway paper.

Line 1039. Please explain what you did to get the relationship between alpha and beta.