Thanks for your attention to the reviewer comments, the revised manuscript incorporates their feedback well and I just ask one very minor correction before publication: some clarifications to the paragraph on code availability. Also, I think the reviewers both gave very good and constructive criticism overall, but I agree that the focus of one query on whether a fundamental science advance had been made is not necessary to the remit of this journal, there are both clear incremental advances, and some computational efficiency advances on top of a well constructed and documented code becoming publicly available. The modular structure will allow other researchers to test and compare various structures and other model decisions to make further scientific contributions. The paper fits very well within the remit of GMD to my mind.

My very minor technical correction required is: please change the last line of the code availability section to "A zenodo archive with the exact version of the Shyft version described and used in this paper is available with DOI: 10.5281/zenodo.3634622."