You have addressed most of the comments by the reviewers but I think that there are a few points that require some attention and can improve the paper before publication:

- in the tables (or at least in Table 1) it would be better to have the information on the vertical top of the model and the dz value between vertical levels, instead of simply the number of levels;

- the color bar in Fig. 2 continues to be white in the reviewed version of the manuscript. I have tried 3 different .pdf viewers. I think that something happens with this image, the .pdf page where it is takes a lot of time to load. Please, fix it;

- in your reply to Dr. Ray you say that now you cite one of his works, Ray et al. 2017, however I have not found it in the list of references or cited along the text;

- as you devote some discussion to the changes in the vertical and the point of vertical wave sources and the Brewer-Dobson acceleration and relation to mixing has prompted in the discussion, I think that it would be worthy to have a couple of lines discussing how the deltas/variations that you show can compare to the values of Diallo et al. 2017 (https://doi.org/10.5194/acp-17-3861-2017). I can see readers interested in both papers and otherwise comparison/assessment would be complicated;

- I find noteworthy that the model is under a Free Software License compatible with the GPL, you could consider adding it to your Code Availability statement: "The LMDz model is available from http://web.lmd.jussieu.fr/trac under the CeCILL v2 Free Software License."