Please refer to the attached review document.

Fokkema et al. reported orbitally-resolved SSTs and associated data from a tropical Atlantic site. They first confirmed that the Eocene “hyperthermal” events are present in the tropical ocean. Then, these TEX86-based SSTs were used together with benthic foram d18O to constrain Polar Amplification in the ice-free greenhouse world. They identified a persistent PA factor that appears to be robust across different archives, proxy calibrations, and timescales. They concluded that feedback other than ice-albedo, probably involving the carbon cycle, was in play. This is a solid study with a large number of measurements used to provide robust paleoenvironmental reconstructions, while considering potential caveats (such as upwelling at the studied site and different calibrations of TEX86). These results were used to strengthen the series of studies conducted recently by the authors and other folks to constrain PA over different periods of the Cenozoic. I actually reviewed an earlier version of this manuscript for a different journal and don’t have many remaining comments. I suggest publication after minor revisions.

The only thing I can think of is to use this opportunity to expand a little bit more on the discussion of feedback mechanisms operating on orbital timescales in the early Cenozoic. This should be feasible given the expertise of the climate dynamics in the author’s team. What could be important but currently missing in the model? Also, any thought experiment on the potential drivers of these carbon cycle changes on orbital cycles (I know we still haven’t really figured out the late Pleistocene G-IG CO2 changes, but what are the potential ways to balance the carbon budget of the Eocene? What factors could be more important in the greenhouse world than the Pleistocene)?