Dear Dr. Berends,

Thanks for the responses. As recommended by both reviewers, I would like to accept your manuscript pending final revision. A few helpful things based on your responses and reviewers’ comments are listed here:

1) Both reviewers pointed out that the introduction is insufficient. Please do extend the introduction to cover recent publications such as Prescott et al., (2018), Tan et al., (2017, EPSL), etc.

2) Both reviewers mentioned the lack of description of model configuration, and they are both confused by equation 1 and 2 regarding how it is implemented and how the scaling factor is determined in 2. Those are valid and important points to address. From what I understand, it looks like, in both equation 1 and 2, (t+0.1ky) should be a subscript instead of an actual term. If so, please correct.

3) Both reviewers pointed out that a major caveat of this study is the lack of warm Pliocene-like climates in the climate matrix. Given that vegetation, ocean circulation, and ice sheet all respond nonlinearly to insolation and CO2 forcing, a good discussion on future work should be included beyond a few words. Note that recent studies have reported the non-linear Earth System sensitivity to forcings (e.g., Friedrich et al., 2016, Nature).

4) Finally, reviewer 2 has pretty good comments on discussing model limitation (especially the resolution and parameter sensitivity, whether bias correction is used etc), and many detailed comments on improving the clarity of the manuscript. Please do follow them when revising your manuscript.

Thank you very much! We are looking forward to the revised manuscript!
Ran

Page 2, line 6: delete “such as the late Eocene”. Given its transient nature, Late Eocene does not belong to the warmer-than-present global climates that the community is investigating.

Page 2, line 9: …a strong warming in the polar regions…Please cite the original proxy records for polar amplification. Perhaps consider two recent papers for this list of references:

Salzmann, U., Dolan, A. M., Haywood, A. M., Chan, W. L., Voss, J., Hill, D. J., ... & Contoux, C. (2013). Challenges in quantifying Pliocene terrestrial warming revealed by data–model discord. Nature Climate Change, 3(11), 969.

Fletcher, T., Feng, R., Telka, A. M., Matthews Jr, J. V., & Ballantyne, A. (2017). Floral dissimilarity and the influence of climate in the Pliocene High Arctic: Biotic and abiotic influences on five sites on the Canadian Arctic Archipelago. Frontiers in Ecology and Evolution, 5, 19.

Page 6, line 18, “…the prescribed climate is an interpolation of these snapshots…” Based on my reading, this meant linear interpolation. Please specify. If not, please explain what kind of interpolation is this.

Page 8, line 23, “…The modelled benthic d18O is calculated using ice volume, ice-sheet d18O and deep-water temperatures relative to PD for every 100 years.” This is not understandable. Please add information on how is this calculated. Have you used some kind of linear regression? If so, please specify, and write out your regression model.

Page 12, Figure 6 (Line 16): Please spell out the acronym “LGC” upon first use.