I am grateful to the authors for their patience during editorial consideration, and for their responses. The paper is recommended for major revisions. Please address all reviewer concerns directly, including input from the 2K data stewardship team and input from the public review ("short comments"). This may include, but is not limited to, the changes you proposed in your author responses. The revised manuscript is subject to approval by the reviewers and the editor before a final decision on publication is made.

In particular, please give careful attention to the description, incorporation and consideration of uncertainties when reporting the results and in the discussion of correlation, spectral and other analyses of data.

In addition, your response to input from the PAGES2K Data Stewardship Team is judged to be insufficient. Please be sure that the following changes are made in the revised manuscript:

- Data-review comment #1 asked the authors to add a "Data availability" section (as does CP instructions to authors). Please create a "Data availability" section, rather than adding a statement to "Acknowledgments".

- Data-review comment #2 asked authors to add data citations to Table 1. Please do so, rather than adding a missing publication citation.

- Data-review comment #4 asked authors to submit the data that are used in the paper to a public archive. Your response was that the data will be submitted "after the manuscript was published", but the data need to be submitted before the manuscript is published. Note that data may be archived with NOAA prior to acceptance of the paper, but only made fully visible to the public once the paper is published.

Thank you.

Dear Dr. Sano (msano@aoni.waseda.jp),

Thank you very much for your revised manuscript. I am in receipt of 4
anonymous reviews and a review from the Data Stewardship Team.

Although three reviewers have recommended acceptance and minor
revisions (please do address their suggested revisions), a fourth
reviewer has suggested major revisions (document:
cp-2016-132-referee-report-3.pdf). I have reviewed these suggestions
carefully, and I believe that making these revisions will result in a
stronger and more important contribution. In particular, they ask you
to consider: (2) justification for the stacked record, given Fig 6;
(1abc) strength of conclusions relative to propagated error,
especially for the low frequency interpretations (cited figures); and
(3) giving greater weight and refocusing the paper's discussion and
conclusions around the most-substantiated "ENSO" timescales and
patterns evident in the results, making the discussion of the
low-frequency variability more speculative, given the uncertainties in
the results (cited figures). Therefore I am asking you to revise the
paper once more in light of the specific requests from this reviewer.

In response to comments from the data stewardship team on the Data
Availiability Section, we have confirmed that the Hulma, Wache, and
Manali d18O datasets are available from the NOAA/NCEI repository at
the URLs noted in the paper. Please confirm that the JG and Ganesh
d18O chronologies and the composite regional d18O chronology developed
in the present manuscript are also available via the NOAA/NCEI
repository: per Climate of the Past and PAGES2k Special Issue
policies, these revisions are required prior to acceptance of the
manuscript. Currently we see only "shell" urls for the JG and Ganesh
datasets, and in your prior response, you have not indicated a url for the
composite o18 chronology reported in the present manuscript. For
these latter three datasets, please forward a copy of the data that
were delivered to NOAA/NCEI. If these data have not yet been
delivered to NOAA/NCEI, please cc: me on your email which delivers
those data. The goal is to get eyes on the data and to make sure that
they are on their way or already at the repository before the paper is
officially published.

Thanks and Sincerely,
Mike Evans
(guest editor)

ear Dr. Sano (msano@aoni.waseda.jp),

I am pleased to recommend acceptance of your manuscript to the 2K Special Issue in Climate of the Past, under the condition that the last 2 datasets and the 5-site composite are made publicly available via the URLs indicated in the "Data Availability: section of your paper.

I am grateful for your work to address review concerns so thoroughly and comprehensively through the revisions of this paper.

Sincerely,
Mike Evans
(guest editor)

In The Data Availability section, change wording "should be" to "is", e.g.

"The Hulma δ 18 O chronology is available at this link:
https://www.ncdc.noaa.gov/paleo/study/22547 (Sano et al., 2017b), which was described by Sano et al. (2012).; The Wache δ 18 O chronology is available at this link: https://www.ncdc.noaa.gov/paleo/study/22548 (Sano et al., 2017c), which
was described by Sano et al. (2013); The Manali δ 18 O chronology is available at this link:
https://www.ncdc.noaa.gov/paleo/study/22549 (Sano et al., 2017d), which was described by Sano et al. (2017a); The JG δ 18 O
chronology is available at this link: https://www.ncdc.noaa.gov/paleo/study/22550 (Xu et al., 2017a); The Ganesh
δ 18 O chronology is available at this link: https://www.ncdc.noaa.gov/paleo/study/22551 (Xu et al., 2017b). The Regional tree-ring celluloseδ 18 O chronology is available at this link: https://www.ncdc.noaa.gov/paleo/study/23830
(Xu et al., 2017c).

I don't know if this is handled by authors or by copyediting at Climate of the Past.