Dear Sebastian,

Thanks for your careful reply top both reviewers. Both reviewers have acknowledged the quality of the data set presented in your paper but they both also made a number of remarks regarding the analysis of he data, especially the link between fabric and state of stress.

Based on both review and your responses, I encourage you to submit a revised version of your paper that I will send back to both reviewers.

Best regards,
Olivier Gagliardini

Dear Sebastian,

The new version of your paper has been send back to both reviewers. If both of them clearly acknowledge the improvements in this new version in comparison to the initial one, they have still made a number of comments, especially reviewer #2. Could you please provide a new version of your paper as well as a point by point answer to these new reviews? Based on this new version, I will take a final decision regarding the acceptance of your paper to The Cryosphere.

Regards,
Olivier Gagliardini

Dear Sebastian,

Thanks for the revised version of your paper as well as the point by point reply to both referees. I found that the overall paper has been improved after this second round of review but there are still a number of weak points, that I have listed below. I invite you to provide a revised version to my comments. based on this new version I will take a final decision regarding the acceptance of your paper for publication in The Cryosphere.

Best regards,
Olivier Gagliardini

- line 17: valuable insights on the **current** stress and strain distribution... I think it is more strain-rate than strain, as strain cumulates all strain-rate since snow deposit (which cannot be reconstructed from the current ice fabric as many different strain-rate history cumulating the same strain might give the same fabric in temperate ice). There are a number of places elsewhere in the manuscript where strain is used in place of strain-rate. Check this.

- line 27: strains -> strain-rates

- line 63: with studies from the last century -> with previous studies

- Figure 3: Not so easy to visualise the compressive state with this plot. Why not plotting a map with the longitudinal strain-rate computed from the velocity field? (and also other components of the surface strain-rate using 3 subplots: Exx, Eyy and Exy). Explanation from lines 91-93 would be simplified.

- line 105: was chosen with n=3 -> was chosen to n=3 (?)

- line 108: what is the value of m?

- line 113: bedrock velocity -> sliding velocity

- lines 129-130: you should mention first that fabric will induce an anisotropic viscous response of ice, especially because the viscous response is more anisotrope than the elastic one mentioned here (and is of better interest for glacier flow).

- line 160: I don't understand the z = y′ and y = −z′ notations. Is that true for all three types of thin sections? To help the reader, the two reference frames (x,y,z) and (x’,y’,z’) should be plotted in Fig. 4.

- Figure 5: You should indicate the limit of the sample for c-axis orientation measurements. There are a number of very small grains at the limit of the sample that I guess are not measured (seems to be a result of the drilling technics which melt the ice)? Why not plotting the midpoint on this figure?

- lines 218-219: No the Cauchy stress is not the deviatoric stress! And p is the isotropic pressure. So it should write: "model to derive the deviatoric stress tensor, which is defined as the Cauchy stress tensor minus the isotropic pressure, i.e.:"

- Table 2 would be more useful on the form of a graphic to visualize the evolution with depth of the stress components? Indicate also somewhere the sign convention that negative is compressive.

- Caption of Table 3: I don't understand why you are using the borehole measurements to derive the strain rates? For a given depth, the ratio between a stress component and the corresponding strain-rate component should be constant (which seems the case). More explanations are needed on how the strain-rates are calculated.

- line 226: Not sure to understand what you mean by the direction of stress and strain-rate are parallel. More than parallel, using an isotropic law, you assume the collinearity of the two tensors (each components is proportional to the other S_ij = A D_ij with A a scalar).

- lines 228-229: The orientation dependent response of an anisotropic ice is certainly the main deviation, the alignement of the two tensors comes after. You can get up to an enhancement factor of 10 for anisotropic ice.

- line 231: I don't understand the "is too complex for the multi-maxima patterns". There are a number of micro-macro models that could (and have already) computed the anisotropic response of an ice sample for a given fabric. One could use for example the simple Static model assuming homogeneous stress in the crystals to evaluate the polycrystal response for such multi-maxima fabrics.

- line 234: eigenvector of which tensor?

- line 244: Figs. 3 and 7

- line 245: My understanding of recrystallisation fabrics is that they adapt almost "instantaneously" to the current state of stress. Not sure you need to mention the last four decades?

- line 271: strain or strain-rate?

- line 272: I don't see easily why the presence of simple shear would induce that the principal directions are not anymore aligned, even for an anisotropic material (e.g. orthotropic)? May be add a reference here.

- lines 277, 280, 285 strain or strain-rate?

- line 280: Why for the depth 79m, the strain (rate?) is used instead of the stress? And how do you get the principal direction?

- line 287: open bracket not closed.

- line 321: what do you mean by high temperatures? A temperate glacier is at the highest possible temperature ice can be?

- lines 326-332: why not discussing this above when these processes are introduced.

Dear Sebastian,

Thanks for this new version of your manuscript and for your reply to my last comments. I have still few comments that need to be clarified before the paper can be accepted for publication in The Cryosphere.

I think there is still an issue with the definition of the reference coordinate system you are using for expressing both the components of the stress and strain-rate tensors and plotting the fabric. It should be mentioned in a consistent way, where x, y and z are first introduced (i.e. for the strain-rate, line 96 of the tracked changes version) how x and y are defined (line 166 it is mentioned that z is pointing upward, should be mentioned vertical, also). But is x or y pointing to the geographical north should be specified. Also I am wondering if the results would not be more readable if x was defined as the main flow direction? At least, in term of stress and strain-rate, it would be easier to interpret. Moreover, line 232, in the discussion about the stress, \sigma_{xx} is referred as the longitudinal stress, which would suppose that x is aligned with the main flow direction (if not, then \sigma_{xx} is not the longitudinal stress). At the end, I am not sure that you have a consistent definition of the coordinate system all along the manuscript. If you are using different ones (which I think is not a good idea), then different notations should be used. In short, you should clearly define the reference frame (x,y,z) at the beginning of the paper and express all components (stress and strain-rates) in this reference frame, as well as plot the stereo plots in the same frame (instead of N-S plot if the reference frame is not linked to the North).

The color scale of Fig. 3 should better reflect sign changes (important for \epsilon_{yy}). In particular, is the sign change of \epsilon_{yy} aligned with the main flow direction should be highlighted, by either assuming a reference frame aligned with the main flow direction or by indicating on this plot the main flow direction.

Regards,
Olivier Gagliardini

Dear Sebastian,

Thanks for this new version. I think that it is now much clearer regarding the different coordinate systems used in the manuscript. I am happy to accept your paper for publication in The Cryosphere.

Best regards,
Olivier Gagliardini