Dear Catrin Stadelmann and co-authors,
Thank you for submitting your response to the referees' reports. Based on the changes you outline, I encourage submission of a revised version of your manuscript.
Best,
Nanna

Dear Catrin Stadelmann and co-authors,
Thank you for submitting the revised manuscript. I find that you have satisfactorily incorporated the suggestions from the reviewers and this new version is much improved. I still have some minor concerns that I would like to see addressed before approving the manuscript for publication. My comments are mainly related to clarity of writing. See below for details.
Best,
Nanna, handling scientific editor


Line numbers refer to revised manuscript

Line 11: "...and that allow glacier-specific calibration". Unclear what "that" refers to.

Line 32: Change "illustrate" to for example "found"

Lines 42-44: I suggest changing to: "The change in surface height between 2000 and 2011 was found by differencing a merge of two TanDEM-X radar images (28 January 2011, 4 April 2011) and the SRTM DEM."

Line 57: Strictly speaking, isn't this an extrapolation? The surface elevation change is extrapolated to the point in time when the GPR thicknesses were acquired.

Lines 58-65: I am not sure that this (very useful) paragraph belongs in the Data section. Consider if this would be a better fit in Section 3 (for example under the description of the experimental set up) or in the Discussion section.

Line 71: change: "...being driven by..." -> "...driven by..."

Line 73: delete "observed"

Section 3.2: This section is now very short and one of the lines is basically repeating what has already been mentioned in Section 2. Consider merging this section with the data description, since the manuscript does not deal with the details of TanDEM-X processing anyway.

Line 135: It is good to remind the reader that Experiment 3 is in a different year than Experiment 1+2. For example, the sentence could say "In Experiment 3 (Fig. 1) we now move forward in time to 2011. Here, KG is split into two parts... "

Line 141: Change "orientation" to "guideline" otherwise a reader might think that it is the orientation of the glacier.

Lines 154-155: Does the consensus map use a constant viscosity? If yes, is that the explanation for the similarity?

Line 161: Which Experiment is being discussed? Exp 1?

Lines 164-165: "Increased mismatch values, especially for borehole C3, might as well be explained by the very flat plateau." What does this sentence refer to? It implies a change between two values but the mismatches decrease from Experiment 1 to 2. Please clarify. Should "Increased" be "Large"?

Lines 171-173: This sentence: "Despite that no GPR measurements were considered on Mt. Kilimanjaro, the complex topography posed a similar challenge for the models participating in the consensus." combines two seemingly unrelated sentences. It is not clear how the absence/presence of GPR measurements is coupled to the topography. Please clarify.

Line 180: Please mention the mean ice thicknesses found in Experiment 2: "..increased in comparison to Experiment 2 from XXm to 9.3 m and from 23.4m to 26.6 m, respectively."

Lines 181-182: "Resolution can be excluded from a 25 m reconstruction in 2011 (not shown)." - > "Resolution can be excluded based on results from a 25 m reconstruction in 2011 (not shown)."

Lines 182-185: "Concerning the 2011 outlines, some internal ice-free areas (on both NIF and KG), present in the RGI, could not be confirmed from the coarse Landsat imagery, resulting in thicker ice. The quality difference between SRTM and KILISoSDEM is certainly also a contributing factor explaining part of the larger thickness values. "
This part is really crucial in order to understand why there is a difference between the 2000 and 2011 reconstructions. The present version leaves several questions unanswered. Firstly, are the ice-free areas real? If yes, is the difference then due to the (incorrect?) outline from the Landsat imagery? Wouldn't you then have gotten better results if you had incorporated the RGI ice-free areas in your 2011 outline?
Secondly, what is meant by quality difference? Quality in what? How do the DEMs differ and how does their difference impact your results?

Lines 196-197: The term "data assimilation" is not used in the methods description at all. Considering that the reviewers also asked for more details on this, I ask that you explicitly mention this method, for example, in Section 3.5, e.g., "The method of Fürst et al. uses data assimilation to...". Alternatively, rephrase this part in the conclusion.

Lines 210-211: "We therefore speculate that thickness information from retreat is most useful in areas that have been dynamically more active in the past." This is a very important point and I think it should also be directly mentioned in the Discussion. As it is now it is buried.

Section 6: This section could do with some streamlining and tidying up. At the moment, report on results is mixed with discussions of the utility of retreat observations and ice thickness measurements. I am also confused by the word "assess" - the new ice thickness map is compared to the consensus map but since the latter is likely wrong it seems counterintuitive to assess with consensus map?
I suggest restructuring the section:
Firstly, report on the findings. Mean ice thicknesses for 2000 and 2011 and how the results compare to the consensus map.
Secondly, the use of observations of glacier retreat to constrain ice thicknesses. From your results it would seem that direct ice thickness measurements are most valuable, then glacier retreat observations and then (in the absence of any data) the consensus model approach.
Thirdly, outlook: In the future multi-temporal satellite information might aid in reconstructing glacier thicknesses with the caveat that complex topography (or other factors) might complicate matters.

Fig. 2: I don't see any coloured dots with a magenta outline? I can see magenta contours. Please revise figure caption.