BG-2018-518

Dear Maria-Elena Vorrath,

First of all I would like to thank both reviewers for reviewing this manuscript and you for your reply. I think the reviewers made some good suggestions concerning your manuscript and I think you should definitely address these. There are two main issues I think you really need to address. The first issue deals with the analytical details. In response to my earlier comments you have already addressed some. I assume that when silylating sterols you added both BSTFA and pyridine, for instance? Another question I have related to analytical details is if you could give a more thorough description on how you calibrated the HBI response on the MS in order to be able to compare them with other compounds? How does this affect the use of the c-factor, sometimes you use it, sometimes you do. There is a relatively large number of “zero” points, how does this affect your correlations? How do you deal with the large variability or standard deviation in your sea ice concentration? Make sure that readers can follow your approach and even reproduce it. My second issue has to do with the set-up of the manuscript. You now discuss the data in light of your new sea ice proxy PIPSO25, the alternative would be to discuss your data in light of all the environmental parameters including sea ice and end with by suggesting a potential new sea ice proxy for the Antarctic region, PIPSO25. Basically reorganizing the discussion to a more general discussion rather than the focus on the proxy. The data you have generated is clearly interesting, the issues are with the PIPSO proxy, by shifting the focus you would generate a more acceptable manuscript. It might be that the oceanographic setting around Antarctica is so complex that a sea ice proxy similar to the Arctic is just not feasible, for instance. When discussing your data in light of the oceanography, sea ice extend and other environmental parameters this outcome would be fine. In the current manuscript with the focus on the PIPSO25 proxy, the proxy has to work, any other conclusion cannot not be accepted. I think this is why the manuscript as is feels a bit forced, PIPSO25 has to work and this is not necessary. Please also check if you site all relevant literature, also the more recent publications and in the right place. Again, with a slightly different focus I think you have a great dataset and a very interesting manuscript. I am looking forward to your rebuttal and revised manuscript.

Best regards,

Marcel

Dear Maria-Elena,

Thank you for taking all the comments seriously and modifying your manuscript accordingly. I find it has improved significantly and is now much better balanced. I would like to accept your manuscript for publication.

Best regards,

Marcel