The authors have addressed the majority of the referees comments in a clear and thoughtful way. Some final small points to be addressed:

- there are some small typos in the text, usually where new text has been added during the review process (e.g. lines 377 on page 4). The pre-prints should be thoroughly proof-read.

- in response to reviewer 1 comment 3: I can only see a note to 'assuming modern day estimates of d30Si(OH)4 are valid for the past' as the addition the authors note (line 350 p.4). Somewhere in this section the authors should include the confirmation given in their response document, that this assumption is valid given the processes that we expect ought to change values of d30Si(OH)4.

- in response to comment 5 of reviewer 1 (regarding the thermocline), the authors posted a 2nd interactive comment which notes the presence of a low resolution SST data set for ODP site 882. This data is not cited in the revised manuscript, which instead suggests no data is available. It would be better to note that although there is data, its resolution is not appropriate for comparisons to be made here, and nor is there a surface-subsurface depth transect of temperatures which would allow confirmation of changes to the thermocline.

- reviewer 2 sought for clarification that the SST cooling with ONHG was in annual SSTs given that alkenones show a warming. This hasn't been corrected on the revised ms (line 44 p.1)