Dear authors,

I have now received the reviews the new version of the manuscript.
Can you account from Referee #1 comments ?
I report them below :
"I would like to thank the authors for their careful answer to all my comments and suggestions. The paper has been significantly improved. The results are much clearer thanks to the better evaluation metrics and the reference to the open-loop ensemble, and the description of the experiments in the methodology significantly improves the paper structure. The main learnings are also better emphasized in the conclusions. As a result, I think that this paper will be a very important contribution for data assimilation in snowpack models. However, I still identify 4 issues which deserve a last minor revision of the manuscript before publication :

1) The new description of the openloop ensemble behaviour is extremely useful (now in Section 2.5.3). However, I am not sure whether the results (from page 12 line 23 to page 13 line 10) are well positioned in the manuscript as it would be more common to present these results at the beginning of section 3.

2) In the same part, I find the rank histograms very useful. The right one relative to the rank of the deterministic member is well described in the text but unfortunately the left one relative to the rank of the observation is not described. I think a short description will be very useful. Indeed this rank histogram is typical of an underdispersive system and it suggests that the perturbations of the meteorological input will not be sufficient either because they are lower than the real uncertainty either because perturbations of model parameters are also required.

3) Several answers to my questions improve the understanding of this work but unfortunately some of them were not included in the manuscript whereas they would be very useful. This is the case about :
- the 2 layers characteristics because it helps understanding what is the simulated surface temperature ;
- the way the incoming longwave radiation is parameterized because it appears as a potential source of uncertainty which is not accounted for in the perturbations applied in this work ;
- the explanation about Equation 10 because it is critical for the reader to understand what the authors say in their answer ;
- the fact that precipitation intensity is not perturbed ;
- the time of assimilation and evaluation for surface temperature ;
- the definition of the snow free periods

4) It should be specified how unique KGE values are obtained in Table 7. Is it the KGE of the ensemble median or the median of KGE for all members ?"


Thanks a lot,
Best regards,

Marie Dumont

Dear authors,

Thanks a lot for accounting for the reviewers comments.
I think the paper is now ready for publication.
Is it possible to just rephrase line 12-14 page 15 to something like : "The distribution of the SWE ensemble simulation shows that the ensemble is under-dispersive, which warns about the limited representativeness ... "

Best regards,

Marie Dumont