Dear Hélène and Mathieu,

Thanks for this new version of your paper and your careful reply to both reviewers. Both of them have positively evaluated the work presented in our paper and made pertinent remarks to improve the manuscript. I think that your paper can be accepted for publication in The Cryosphere, but after a careful reading I still have few remarks and found some typos that should be corrected for the final version.

My main remarks is on the strong conclusion of the paper that recommand to use the NMP parametrization. I think this should be discussed a bit further, based on the following points:
- Among the 4 parametrizations, the most physical one is the SEM2: it is the one that reproduce "exactly" the melt function you want to apply everywhere on the domain. None of the other do that (you might think adding a figure showing how the melt evolve in the vicinity of the GL for Exps 1 and 2 and the 4 melt parametrizations, and compare it to the mathematical function of melt).
- Only SEM2 and SEM1 are mass conservative for the total melt (total melt is independent of the mesh discretization). The NMP will give less total melt as mesh size increase, and it is the reverse for FMP. In other words, the forcing (or total melt below the ice shelf) is dependant on the mesh size for NMP and FMP, not for SEM1 and SEM2.
- The conclusion that NMP works better than the 3 others is obvious for the Weertman friction, but this is less clear when looking at Tsai results (see below). Isn't it a bit hazardous to build the conclusion on the friction law that has been shown not to be adapted for the vicinity of the GL? At least, this should be discussed.
- For Tsai friction law and Exp 1, I would recommand to use SEM1 or SEM2, not NMP, if only looking at the mesh dependency. For Exp 2 and Tsai, the dependency to mesh is very comparable for NMP, SEM1 and SEM2, but the ice mass change are much larger for SEM1 and SEM2 than NMP (and SEM1 and SEM2 gives similar results). Which mass changes are the most correct? Based on the previous remark, I will be more confident in the results given by SEM2 (confirmed by those obtained with SEM1) than those of NMP.

In short, I think you can draw different conclusions from your work and you should be very careful in bringing the whole community to use an inappropriate melt parametrization for bad reasons.

In the supplementary document you will find an annotated manuscript with some typos that you should correct in the revised version.

Best Regards,
Olivier

Dear Hélène and Mathieu,

Thanks for this new version of your manuscript and your reply to my remarks. I now happy to accept your paper for publication in The Cryosphere.
Thanks for having choosing The Chryosphere to publish this nice piece of work.

Regards,
Olivier Gagliardini