Dear Fraser King et al.
thanks for providing online replies to the reviews. I now invite you to resubmit a revised version. Please provide a consolidated revision with a consolidated response letter to all reviewers detailing where changes were made in the revised manuscript.
The three reviewers all find your manuscript publishable in general after addressing a number of concerns, incl. important methodological clarifications. They all chose minor revisions, but together the points of criticisms are a bit more substantial. Hence I may opt to ask one or two reviewers again to check the revised version.
One aspect that still needs to be looked into a bit better is the bias change pointed out by reviewer 1. Please make sure that there is no data inhomogeneity behind it - and if so, I agree with the reviewer the effect needs to be analysed and discussed more thoroughly.
I am looking forward to seeing the revised manuscript.
Best wishes,
Kerstin

Dear authors,

thanks for providing the updated manuscript and apologies for the (coronavirus induced) delay in the decision process.

Most of the reviewer's points have been addressed but I still see need for some minor improvements. More generally, the manuscript needs a conclusion. This can be brief, but it needs to be clear where this conclusionn can be found (heading) and the conclusion needs to clarify what the contribution of this study is (beyond summarizing results and discussion).
In addition, I found a few remaining points that are important and cannot only be rebutted in the response letter, but require to be addressed in the manuscript. Perhaps it just wasn't indicated in the response where they are addressed and this may be real quick to let me know so that I can check it more easily.

In summary, besides the missing conclusion, please revise the ms to explain how and where you added clarifications on the following reviewer comments:

Regarding comments of Rev 1

Rev 1 point 3) & Rev 3 point 6)
These generally major criticisms should be discussed more detailed in the discussion section - was that added? If not, it needs to be added.


Regarding Comments of Rev 2

2) The representativeness of the observation location is mentioned. A discussion of this needs to find its way into the manuscript. Are there documents and studies on that representativeness? I know form other snow regions, that for example, such observations are mostly taken in flat terrain and locations that may be not so wind exposed hence biased to that....etc. Such issues may help explain part of the bias.

6) How does the manuscript now ensure that others don't misunderstand this as well? The methods section needs to be clear on the procedure followed.

Rev2 point 7) and Rev 3 point 3) Training/Validation partitioning needs to be clearly explained in the methods section and not 'hidden' in the results. Please change this to make the manuscript fully transparent.


Regarding Comments of Rev 3

4) At least this needs to be discussed as caveat/weakness. The representativeness of the station has nothing to do with the issue of autocorrelation influencing the statistical methods. These are two different issues.

5) Where is this clarified in the manuscript?

Thanks,
Kerstin

Dear authors, thanks for making these final changes and for choosing HESS.
Regards,
Kerstin