Dear Drs. Heister and Scheiber,

Thanks for submitting your manuscript to The Cryosphere Discussions and your response to the two reviewers’ comments. I am generally satisfied with both your responses and revisions to the manuscript. The presentation of your technique and its overall scientific significance have improved significantly and become more compelling to the audience of The Cryosphere. Below I have several minor comments for consideration prior to a final decision on your manuscript.

Joe MacGregor
NASA GSFC
9 August 2018

Title: Is “Benefits of” really necessary here, given that the processing described is uncommon at best”. I think those two words could be dropped.

1/19: “…ice sheet and bed were performed by Jezek et al., …”

2/3: “…which leverages the fact…”

2/9: “…highest SNR were then selected from further processing, aimed at improving the SNR of weak bed echoes in outlet glaciers.” Don’t include the unnecessarily negative remainder of this paragraph.

2/25: Approximately how long in azimuth are each of these “blocks”? That information is needed for the reader to assess whether the surface slope can reasonably be assumed to be uniform over that distance, which has an effect on refraction through the ice.

2/27: Give equivalent real part of relative permittivity value also.

2/31: “We now describe the algorithm inputs using…” and end this sentence with “(Algorithm 1)”.

3/3: windows

5/11: “An inverse azimuth Fourier transform is then applied…”

5/19: “can be tracked semi-automatically or automatically, …”

6/9: How much of the variability in apparently layer slope is due to the layer tracing itself? About 1 degree?

6/11: What does it mean to “rebin the echogram in azimuth by a factor of 8”? Are the data decimated of interpolated? Clarify.

6/11: Here a ramp of 15 dB/km is used, whereas later on 20 dB/km is used. A bit confusing and better to stick with one value. 20 dB/km is the reference value for ice sheets although it is probably actually lower in the regions you surveyed.

10/6: “…varies more as compared to…”

10/20: “at the –6 dB level”

14/8: It appears more logical to cite Schroeder et al. (2013) here also, since that study did exactly what is described in this sentence.

14/11: “of the bed return, Schroeder…”

14/24-30: Rephrase these sentences. RES observations of internal layers must be due to changes in electrical properties, otherwise they wouldn’t reflect (unless one can somehow mount an argument that they are due to changes in magnetic properties…). Further, these layers are clearly generally isochrones, as evidenced by MacGregor et al. (2015) and many others, and the cited studies describe a debate that is now mostly closed. Finally, it is best to avoid the term “manually” as most layer-picking algorithms are semi-automatic, rather than purely manual.

14/31-16/10: Interesting additional argument and useful illustrations. It may be worth noting that most layer slope methods are only used to predict the internal layering, which then simplifies tracing, rather relying on those layer slopes entirely.

Throughout the manuscript:

- Use “surveys” instead of “datatakes”.

- Use units of dB km–1 for ad hoc attenuation rates, rather than dB (100 m)–1. The former is simpler and is the convention for ice sheet RES.

Figures:

- In general, the figures are too often described in the text as opposed to either in the figure caption or in the figure itself (using, e.g., a legend). Shorten in-text figure descriptions as necessary.

- 3 and 4: Trace internal layer that is considered in Figures 5/6 in red in panel a. Also identify with vertical lines the positions highlighted in Figures 5/6. For all axis labels, give units in parentheses, e.g., “Azimuth (km)” not “in km”.

- 8 and 10: Clarify the difference between the top and bottom panels in both the caption and an annotation on the panels themselves.

Figure 11: Add color scale for first and third panels from top. What are the units? Add panel labels.