Dear authors,
I have received 2 detailed reviews and I also looked at your responses in the open discussion. AS you can see from the reviews, there are a many points that require major revisions, including new computations, to bring this manuscript to its full potential. I think this is possible and I would propose that you provide a detailed response to each reviewer comment and also provide a revised version of the manuscript with the changes marked that you propose in your response. I will send out the revised paper again to the 2 reviewers.
Best regards
Markus Weiler

Dear authors,
as you can see in the 2 review reports, both reviewers appreciate the improvments of the manuscript, however, they are still questioning some parts, see needs for additional changes and in particular one reviewer is disappointed that the authors focused rather on the ‘simple’ revisions than the more substantial ones, which would need new computations, etc. I agree with the proposed need for additional revisions and probably additional model runs and I can only consider the manuscript for publication in HESS, if the authors seriously consider the proposed changes, in particular of reviewer #2.
Best regards
Markus Weiler

Dear authors,

as you can see, the referee report still asks several valid questions. I took the time to have a more detailed look than normal to the revised paper and the your responses. I also believe, that there are several issues with the paper, that need to be fixed before the paper could be accepted for publication in HESS. I know, some of them are not easy, but being partly a field hydrologist myself, I cannot accept that we claim to develop better models but at the end, we just use some simple fixes like adding a soil storage underneath a glacier to represent the much more complex process of englacial liquid water storage, where conceptual models have already been developed. When looking in more detail into the sensitivity analysis in chapter 5.3, I was also wondering about the presented NSE and KGE values for changing the "soil depth" under the glacier (at least you should give this a different name!!) and the values in Table 3. Why are some of the values higher than the best model in Table 3. The sensitivity of changing this parameter is actually larger than the improvement of the model from v0 to v3. If this is the case, why improving the other parts if this process would be so much more relevant to improve streamflow prediction.
From the three points proposed by the referee, I would at least see that the following points should be addressed:
• Explicitly adding storage in the glacier (see above), which might reduce the need to add a thick soil under the glacier
• Performing a limited number of Monte Carlo runs with reasonable parameter value ranges - as you clearly demonstrate, there are many uncertainties and then you parameterise some parameters by "These were fixed by optimizing the recession shape of the hydrographs." This is not repeatable by other scientists.

I hope you see the need to reconsider some parts of your model and to change the analysis and provide a clear way to explain the model and its implications.

Best regards
Markus Weiler

Dear authors,
the revised version of the paper is now close to be accepted. However, I would like to see the following additions to the new sections:
p.10 l. 6-10 : ....These parameters are kept constant through the simulations and were optimized according to the constraint of minimizing the differences (in terms of least squares) between the recession shape of the simulated hydrographs and the observed one.” - It would be helpful to the reader to cite some references that discussed the assumption of keeping these parameters constant of changing it in time as has been done in other models simulating the water storage within the glacier.

Best regards
Markus Weiler

Thanks you for the final changes to the manuscript. I think the paper will make an interesting contributions to this field.
Best regards
Markus Weiler