Dear Patrick and co-authors,

First of all, I would like to thank the two referees for their constructive reviews of your submitted manuscript.

Reviewer #1 raised 3 major points:
1. Concerns about inclusion of Section 2 due to its technical detail
2. Experiment names, a concern that I too shared on reading the first draft
3. Which simulation to use in multi-model analysis

I will take each of these points in turn. For 1., I think you gave a sufficient number of reasons to retain this section and I find these acceptable. In relation to the names of the experiment names, given that other papers have been accepted with these names and that the names themselves are included in the metadata of the output files uploaded to the CCMI database does preclude these being changed. However, the changes to Table 1 have improved the categorisation of the simulations which should help readers. I also liked the addition of the colours for the simulations. For 3., this was an excellent point raised by Reviewer #1, and I'm pleased that you addressed this in the revised manuscript.

Referee #2 raised the following points:
1. Sheer volume of simulations
2. Inclusion of sensitivity runs, with little difference in results
3. Lack of clarity between simulations and the CCMI Reference simulations

This first point is one that the two reviewers and myself all shared and I can accept why you wish to include all simulations - this paper can then act as a single reference for the performed simulations. I also believe that the inclusion of sensitivity experiments, in which you've corrected some errors and even in the absence of significant impacts, to be a valid one. And in relation to the lack of clarity and the CCMI reference simulations, this has been addressed through the changes to Table 1.

On balance, I believe your responses to the reviewers' comments are adequate and I recommend acceptance for publication in GMD.

Thank you again for submitting your manuscript to GMD.

Regards,
Fiona O'Connor