Dear Authors,

I have now read your revised manuscript (MS) and yours responses to the referees. All three referees were very positive about your study and you generally addressed satisfyingly their comments.

I still have myself some comments before the MS can be accepted. I will review your response and how the MS and its supplement were modified accordingly. Thanks for considering them carefully.

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

Best regards,

Etienne Berthier

1.12 « nearly ». Provide % of coverage in brackets to illustrate what is meant by « nearly ».

4.10 South of 88° south

4.17 « too » instead of « to »

4.18 maybe I missed the info, but if not tell when/on what criteria the 31 % remaining strips were excluded.

4.27. accuracy. Ref for the 10 cm ? Shuman et al., GRL, 2006 maybe ?

4.33. point cloud locations. Of the reference (CS2/GLAS) altimetry data ? Clarify.

4.34. Linear trend to what ? The time series of the CS2 at this location? I do not feel the sentence is clear enough and this is an important aspect of your method.

5.4 A reference is needed to back up the moderate elevation change in the interior. An altimetry study. Maybe the recent Schroder et al. study in TC (https://www.the-cryosphere.net/13/427/2019/) ?

5.4 the term resolution is not clear. In the context of this paper, it makes us think about horizontal resolution. Do you mean rather vertical precision ?

5.6 Why is the altimetry point cloud « corrected » ? What correction did you apply to the altimetry data ?

5.7-8. Why using different metrics for CS2 and ICESat? Not clear to the reader. Either use the same terminology to describe both corrections, or explain why different thresholds are applied. And in one case you stated that the DEMs are used for mosaicking and in the other case that the correction is applied. I assumed the correction is also applied to the DEMs compared to CS2…

5.12. « precision ». Is not it rather « accuracy » here ? As the mean bias is the main criteria. A thorough check of the terminology is required. Precision and accuracy are two different concepts (as you know I am sure).

5.15. « were negligible ». I would say « are assumed negligible ». You did not verify it but rather made this assumption. A reference is required here to back up this statement.

5.20 how can you be sure this is a sensor related bias. Why not a real elevation change between the ICESat and the CS2 periods ? If you can discard this possibility, tell why and provide a reference.

6.11 a reference is needed to back up this statement.

6.18. double check the grammar of this sentence

6.21 why using the 70th percentiles here while using 68th percentile elsewhere. Seems inconsistent.

6.27. « tend to cover » I think.

6.32. Here again a reference is needed.

8.18 Fig. X (?)

Figure 6. Legend. Inconsistency between 68th percentile and LE70...

Figure 2. The 1-to-1 line (dashed) would be a useful addition to the figure.

Supplement. It would be best to generate a single high resolution PDF including all the supplementary figures and their legends. Also the figures need to be annotated, especially the figures showing issues of ghosting and shadowing. Many readers will not know where the artefacts are in the images.

Dear Authors,

Thanks for taking into account my additional minor comments on the text, improving the supplement and its referencing in the main text.

I am delighted to accept your paper for publication in our journal.

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

Best regards,

Etienne Berthier