General Assessment

Overall, the manuscript is well-written and the sections are clearly defined. The topic is relevant and timely, given the increasing computational capacity that now allows high-resolution simulations of mountain glaciers. The approach is interesting, and the study addresses an important question: how spatial resolution affects the simulation of ice thickness, flow, and thermal regimes in mountainous regions.

However, while the manuscript is organized into clear sections, some aspects of the results and discussion are difficult to follow, partly due to reliance on references to supplementary material. Certain patterns and the interpretation of figures could be presented more cohesively in the main text.

Besides, there are aspects that require clarification or more explicit justification. While the authors explore a wide range of resolutions (50 m to 2 km), the study relies on a single model (the Instructed Glacier Model, IGM). It is not clear whether generalizations about the impact of resolution on mountain glaciers are fully warranted based on this single framework. Explicit discussion of model limitations and how the resolution interacts with the underlying Blatter-Pattyn approximation, compared to other approaches (full Stokes or others), would strengthen the study.

Introduction

• The introduction’s opening paragraph reads as a series of descriptive statements rather than a connected narrative. Consider revising for smoother flow.
• The knowledge gap and justification for the study could be clarified more explicitly. For example, are there other studies on the effect of the spatial resolution using similar or different approaches or applied to a different setting?

Methodology

• The choice of model parameters requires more justification. For example:
• The retraining frequency of the neural network is mentioned, but no explanation about the selection of the values is provided. Explaining why this frequency was chosen would help readers understand its impact.
• Elevation bands for low, mid, and high altitudes are defined as 0–1200 m, 1200–2400 m, and >2400 m. A brief justification or reference would improve clarity.
• While the study explores a wide range of resolutions, all resampling is performed using cubic interpolation. Would the authors expect differences by using other techniques (bilinear, kriging…)? Could this lead to some misleading results?

Results

• Some aspects of the results are difficult to interpret solely from the figures, particularly given the reliance on supplementary material. Consider highlighting the key patterns in the main text to improve readability.
• While the manuscript identifies three resolution “modes,” the evidence is not entirely clear except for Figure 8 and not for all the outputs and elevation bands.

Discussion

• The discussion is generally relevant but could be more explicitly connected to practical implications. For example, explaining how these findings could guide modelers in selecting appropriate spatial resolutions would increase the impact (other than just going for finer resolution as a general rule).
• I feel that maybe a more detailed exploration of this 400-800 “critical” mode should be addressed before to clearly talk about strong sensibility.

Conclusions

• I like the bullet structure, but I think that some points could be rearranged for more clarity.

Figures and Supplementary Material

• While the figures themselves are informative, the main text often relies on references to supplementary material to interpret the results. This can make the narrative difficult to follow. Consider ensuring that the key patterns and findings are clearly described within the main text, using the supplementary figures as support rather than as primary evidence.

Minor Comments

Here I will try to skip the comments raised from the other reviewer so we don’t need to repeat the same things.

L108: Please, justify this choice of (0.01-0.04).

L178: Please, justify the selection of the elevation bands.

L184: effects -> affects?

L209: This 10 times more is not clear to me, at least from the cited figures.

L254: It is not clear to me looking at Fig. S12.

L426-7: I really don’t see those greatest relative changes at high altitudes. Probably you have the numbers, but looking at Fig. 8f such a thing is not clear to me.

L429-31: Again, this is not obvious to me just looking at Fig. 8.

L474-8: Nothing to object here, the opposite, really clear statement and good point resulting from your work.

L500: Here you suggest to go to finer than km- and ideally 500-m scale- but your defined critical mode ranges from 400 to 800m. I found this a little bit contradictory.

Overall Recommendation

Even if this is a relevant and well-executed study that addresses an important question in glacier modelling, especially in regional or global studies, I consider the manuscript as major revisions after considering some points:

1. Clarifying model limitations and generalization potential.
2. Justifying some methodological choices, such as elevation bands, retraining frequency, and resampling method.
3. Streamlining results and discussion to highlight the most important findings and improve readability. Exploring a bit more this critical mode sensibility. 
4. Enhancing figure clarity and reducing overreliance on supplementary material.

I believe that addressing these points will help the resulting paper become a solid and valuable contribution to the field.