Dear authors

Thank you for your replies and detailed comments in response to the Reviewers concerns. The detail of the replies gives much better detail on the approaches that you took for handling the different data sets (e.g. the ice volume corrections) as well as the decisions you made about methods (e.g. why select the Mueller et al. 1998 temperature calibration). In a revised version of the manuscript, can you please ensure that these details are incorporated into your text, so that readers can follow the decisions made about particular proxies and the analysis of their results.

A comment from Reviewer 2 was that for the d18Oivf record there is a strong influence of the SST record in generating differences (whereas the original d18O are more similar). You provide a good description now of the steps you took with the data and why the approach is sensible - could you also re-iterate this point of the reviewer about the similarities in the original d18O data (i.e. before corrections)? From the reviewer's comments it seems that some of the finer (but important) details of the records may have been unclear e.g. the importance of the 'bump' and so you may want to carefully check over the descriptions of the patterns in the data.

Dear authors

Thank you for responding to the reviewers comments in detail, and for incorporating their comments into your revised manuscript. The manuscript is improved and you have addressed many of their concerns. I have a comment here about the structure/duplication comment from Reviewer 2, then some minor comments to consider.
Reviewer 2 commented on some duplication of information about the circulation. In the revised manuscript I can see where this arises – it is not word-for-word duplication, but you have a detailed discussion at the start of page 3 about the different water masses in this region and how they interact, which then feeds into how different previous work might have found differing results. I can understand the rationale for providing this information, since it gives a good justification for the study. But when I was reading this I thought “this would make a good section on it’s own”, and yet in Section 2 there is a section called “regional settings” which also talks about circulation patterns and their controls in more depth. I recommend that the authors consider shortening the introduction e.g. by finishing page 2 with a statement of the study aims and then moving onto a “regional setting” section (having used Figure 1 to flag that there is complex circulation). Or, if the Introduction stays as it is, section 2.2 could be more focussed on the direct controls over circulation at these two sites (e.g. section 2.2 describes processes happening at 10-15*S, at 33-38*S etc but some of this is visible in Figure 1).

The final paragraph of the Introduction includes text (from “our multi-proxy reconstruction indicates...” which explains the findings of the study. I recommend to remove this text which is “Discussion” to keep the focus of the Introduction on the rationale and aims for the study and then what was done.

I recommend swapping #2.1 and #2.2 so that the information on the sediment cores is directly adjacent to the age model information (#2.3).

page 4 lines 3-4: place the information about selecting the top 250 cm into it’s own sentence rather than in brackets.

You discuss the uncertainty on the d18OIVF-SW results. Could you please flag these in the caption to Figure 3 (and Figure 4), so that it is clear that the variations you are calculating exceed those propagated uncertainties? (even better – show a vertical bar for the size of the uncertainties if you can’t show them on each data point, but I can see that you don’t have much space on this graphic). It could be particularly valuable in Figure 4 to show that both data sets ‘overlap’ or not.

page 8 line 16: some duplication in the text (“the 101” is stated twice)

In Discussion (p.10 and Figure 5), where the regional signals are compared, it would also be good to consider the record of Roberts et al. (https://www.sciencedirect.com/science/article/pii/S0012821X17303825) from the south west Atlantic. Roberts showed alkenone SST changes upstream of the Drake Passage, and made comparisons between the Santos data cited here as well as data from the Chilean margin, including observations of anti-phase SST patterns (rather than synchroneity noted here). The Roberts paper discusses circulation implications for the south west Atlantic so could provide some interesting comparison.

Thank you for your revised manuscript and the replies. I am pleased to recommend publication once some very minor technical corrections are made (they can be found in the attached pdf - see highlights in green and comments).