Dear authors,
Two referees evaluated your manuscript (MS), and I would like to thank both of them for their thoughtful and timely comments. You provided detailed responses and suggestions of how to address them.
The responses suggest that you now much better acknowledge the source of errors. However this apparently did not resulted into a revised uncertainty estimate. It is quite striking that for, a poorly sampled grid cell in Tien Shan (maybe a worse case scenario), the dh/dt evolved from a positive value initially to -0.56 m/yr. Are your confident that the +/- 0.1 m/yr uncertainty reflect the level of sensitivity to the data sample and the regression model? Should not this sensitivity somehow also be reflected in the error bars?
To continue with the evaluation process, I would like to ask you to upload the revised, both clean and track-changed, MS for a full assessment by me and at least one of the reviewers.
I am looking forward to receive the revised MS.
Best wishes,
Etienne Berthier

Dear Authors,

I concur with the reviewer who assessed the revised version of your article that it is now much improved. You will see that he still has a few minor comments. On my side, I copy-annotated the PDF of the article and the supplement (attached). Some of my comments are just stylistics. Others are more important and I expect an answer to them in your response letter.

Do not hesitate to contact me if you have some further questions.

Looking forward to receive your revised text soon. I will evaluate it promptly to avoid further delay.

Best regards,

Etienne Berthier

Dear Authors,

Thanks a lot for revising the manuscript and, in particular, for including a stand-alone section on uncertainties.

I am delighted to accept your study for publication in our journal. Thank you for choosing TC.

You will find my final technical comments below.

Best regards,

Etienne Berthier

L144. check that all citations are made using "(" and not "["

L154. Sorry for being picky on dates but I think this is an important point. Mid-2010 is not super clear to me. Can you provide month instead?

L311. For readability maybe break this sentence into two (after "path")?

L316. I am not native so could be wrong here but I feel that "as a function of" reads better. Ignore if I am wrong of course.

L410. Here again provide month for clarity (January 2011 to December 2015 maybe? or to January 2015?)

L755. Sorry for repeating but one of the issues raised by the RAGMAC working group is the need to clarify the start/end dates of any trend or volume change assessment. Could you also apply this to your items (figures and tables) for those readers who may focus on items first before reading the full text? This apply to all figures/table.