Dear Christian,

After a careful read of this new version and your response to the three reviewers, I am very happy to accept your paper for publication in The Cryosphere. I have only few remarks that you should account for in the final version (see below).

Thanks for submitting this nice, dense, piece of work in TC!
Regards,
Olivier Gagliardini

- page 3, line 20: if you can not provide a reference for that paper, I would suggest to use "will address" instead of "address"

- Fig. 2: I am wondering how you can get a ratio h_f/d_w < 1. I expect that the flotation thickness at grounding line h_f is always larger than the crevasse water depth? What did I miss here? In the legend of Fig. 2, the last sentence shoud end by h_f/d_w instead of h_f. Also, labels of horizontal axes are cut.

- page 8, lines 4-5: an other mechanisms that can enhance tensile stress is the bending of the shelf front? This should be mentioned here, similarly to the enhanced stress due to crevassing (and not few lines below as a mechanisms for models accounting for elastic fracture mechanisms).

- page 10, line 25: I am not sure it is specified at any place that d_w is assumed to be a constant? Also, why the different chosen values for \lambda are expressed as a fraction of a function of r, \epsilon and n is not obvious and should be explained?

- Fig. 3: some text is hidden in between panels d and e?

- page 12, line 17: avoid a whole sentence in bracket.

- page 12, line 7: here /lambda is called the calving parameter. I would suggest to keep all along the manuscript to the same wording, using the more precise scaled crevasse water depth or crevasse water depth?

- page 16, line 22: now redundant

Supplementary:
- above (7): whether the ice is grounded or afloat
- in (8): d_b = d_b = ... should be d_b = ....