Dear Daniel and co-authors,

Thank you for submitting a revised manuscript and for responding to the Executive editor and reviewer comments.

In relation to the Executive editor comments on model code availability, I think you have clearly addressed this comment by making both the Extrapolar and Polar SWIFT model codes available through a Zenodo repository. Congratulations on taking this step in adhering to GMD requirements.

Reviewer 1 was highly complementary about your paper and although they had a comment in relation to a volcanic eruption and the training data used, addressing this comment was not a requirement for acceptance for publication in GMD. However, it's good to hear that you're planning to extend the range of conditions of the training data, including the possibility of a volcanic eruption.

Reviewer 2, too, was positive about the manuscript although had some comments that required addressing. Firstly, they asked whether the SWIFT model would be robust to a 30% change in Cly loading. Although you answered this question in your response, I felt that you could have included some of your answer in the revised manuscript. In particular, may I ask you to include that the model can cope with upto a 10% change in chlorine loading in your conclusions? It would fit well in the paragraph where you refer to the model being able to cope with some variability and before you discuss extreme values.

Secondly, Reviewer 2 asked about the computational cost of the SWIFT model relative to a full chemistry model. Again, although the estimates you have are rough, I would ask that you include something about the relative costs (i.e. 6-8 times faster than ECHAM + full chemistry) in the revised manuscript. A single sentence in Section 5.4 will suffice.

Congratulations on your manuscript and on the positive reviews that you received. I recommend that your manuscript is accepted for publication in GMD once the above very minor comments are addressed.

Regards,
Fiona O'Connor