Dear Authors,

thank you for your revisions to the manuscript. As you will see, the expert reviewer still raises major concerns about your study. I share the reviewer's opinion, and I think that the paper still needs more work before it can be considered for publication. Thankfully, the reviewer agreed to have another look after their points have been addressed.

I am looking forward to receive your revised manuscript.

Best regards,

Fabien Maussion

Dear Authors,

I have finally received the review on your manuscript. As you will see, the reviewer is positive about your work, but still has many concerns about the text and the interpretation of the results.

After reading the review and parts of the manuscript, I believe that, at core, what the reviewer is asking is for the argumentation to be more precise. It is obvious that the reviewer read the paper with interest, and their comments needs to be addressed and answered. I am confident that you can implement these changes relatively quickly.

For my part:
- I'd recommend against the use of too many subjective adjectives and adverbs (e.g. "extremely high", etc.)
- please check the figures one more time for readability (e.g. font size, which should not be too small in comparison to the legend text, for example).

I am looking forward to receiving an updated version of your manuscript soon, together with a point by point answer to the reviewer's comments.

All the best,

Fabien Maussion

Dear Authors,

Thanks for your work on the manuscript, which has improved substantially in the review process, in particular in the writing and data interpretation. The result is a solid, comprehensive study of the performance of these two RCMs in the CORDEX-CAS region. The paper is very detailed and descriptive, so that one is sometimes lost in all these analyses, but I'm sure that it will be of interest for impact studies in the region and for future modelling efforts.

I'm happy to accept it for publication in GMD, provided that the following technical modifications are made:

L286 "extremely high" -> delete extremely. In fact, delete all occurrences of subjective adverbs (if others are left, I didn't check thoroughly)
L290 regional trends -> regional patterns?
L391: typo, double dots
L450: typo, dot missing
Figure 10 left row: consider using a non-divergent colortable for precipitation (also applies to temperature in a way, divergent colortables should be used for differences only I believe). The decision is left to you though.

I would like to thank both reviewers for their thorough assessment of the paper, and the authors of course for their efforts.

Best regards,

Fabien Maussion