Major requests from the two referees were to better describe the impact of forn, sublimation, RACMO models. The revised manuscript was reviewed again by the same reviewers, and satisfied them.

However, this revised manuscript has several points that should be improved before the publication. Therefore, I request a minor revision. The next manuscript will not be sent to the reviewers but will be promptly evaluated by myself.

The list below includes many items but most of them are editorial and I think that the authors can make this revision smoothly. When the revised manuscript is submitted, please include a response letter, which show point-to-point response (including revised text), OR marked manuscript.

Kenny Matsuoka
TC/TCD editor

P2L16: Kuipers Munneke et al. (2014) seems to appear here first time but it is specified as 2014b. (I cannot be sure as your pdf crushes my pc every time I search for the word)

P3L6: Give coordinates immediate after Jason Peninsula (rather than the end of the sentence, which is more confusing).

P4L16-23: Were sonic height rangers able to measure the true snow surface during heavy drift of snow? It is reasonable to use snow heights measured by SHR to determine SMB. However, although other factors are described in depth here, snow drift is not mentioned at all.

P6L1: provide a reference for neutron-scattering probe.

P6L5: use SMB, instead of surface mass balance, throughout the paper (except for abstract and conclusions). Please check this point for the entire manuscript.

P6L11: Please include approximate depth (and/or two-way travel time) of reflectors that are detected by GPR and discussed in this manuscript. Without this knowledge, it is hard to judge the appropriateness of the procedures described in this section. Also, the authors implicitly assume that the layer depth represents the SMB, which is the case only if accumulated dynamic strain is negligible. Here, dynamic strain is not densification but changes in ice-equivalent reflector depth or ice-equivalent thickness of layers bounded by radar reflectors due to stretches of the ice body. This is less important if the ice above the radar reflector was deposited only over the ice shelf (rather than the case that this part of the ice was deposited on the ice sheet and transported over rough terrain that could cause highly variable vertical strain). See:
Waddington, E. D., Neumann, T. A., Koutnik, M. R., Marshall, H.-P., and Morse, D. L.: Inference of accumulation-rate patterns from deep layers in glaciers and ice sheets, J. Glaciol., 53, 694–712, doi:10.3189/002214307784409351, 2007.

P7L1: FAC = Firn Air Content? Please define it clearly. Here and some other places in this section, please cite Holland et al. (2011) in a more readable way. E.g. “Firn Air Content (FAC) derived by Holland et al. (2011)” here, and “FAC thickness (Holland et al., 2011)” at L10 of this page.

P7L10: Show the pore close-off depth and its reference.

P7L23: Is it better to say “AWS data show that the melt starts only after these flight dates so that the Ku-band data were collected prior to the onset of the melt season.” (melting can happen early November but I assume that the authors can say this point more definitely using the data they have).

P8 Fig. 2: The y axis in the figure shows the relative depth in the snow? The number is shown for the air as well and thus it’s confusing. Add “last” so that the third sentence can be read as “… is interpreted as the end of the last melt season”

P8L25: Tex error. \sim

P8L26-27: Delete the sentence starting with “By assuming….” This is not a point necessary for the rest of the arguments. If the authors really want to keep this sentence, please add a reference (in my opinion, this sentence is incorrect).

P8L34-35: The authors assume that the strong reflector is likely generated by the presence of ice lenses and metamorphosed firn created during the prior melt season. However, this argument is not immediately supported by the data (Fig. 4). In my opinion, this is a reasonable hypothesis that can be mentioned here but should be tested more clearly in the result section. Alternatively, the authors can re-organize Section 2.5 so that interpretations and discussion on the origin of the strong reflector is mentioned in Section 3.4 (if yo do so, onsider changing the section title for 3.4 accordingly). Please consider similar re-organization of Section 2.4 as well. My main message is to separate methods and data-supported arguments more clearly.

P8 3 lines form the bottom: “Section 2.4 and see Section 3.1 for results).”

P9 7 lines from the bottom: Is it better to say “but all stations show similar year-to-year variations.”

P9 5 lines from the bottom: Tex error. Figure 5.

P9 3 lines form the bottom: hard to understand so please revise. The authors probably want to say “Forn is enhanced during negative SAM, which also increased temperature and summer melt (Cape et al., 2015). However, our data show that SAM is not a good indicator for the occurrence of precipitation.”

P10 middle of the page: show approximate thickness of annual snow layer.

P11 Fig. 4 caption: Include descriptions of thin blue curves. Make the description of shading panel a specific (now it can be read that shading is shown in both panels).

P11 Fig. 5: Unit of SMB should be updated (mm w.e. /year) in the axis label and figure caption.

P12 Table 2 caption: Add the number of years over which the mean was derived.

P12L5: the reflector discussed here locates around 30 m w.e. According to the annual SMB shown in Table 2, this reflector likely represents isochrones about 70-80 years old. So, please update the title of Section 3.2. If I understand correctly, the SMB pattern (or relative SMB) was derived for a single period using this reflector. Then the section title (“decadal” time scale”s”) seems inaccurate.

P12L8: “permittivity” -> “dielectric properties of the firn”. Dielectric properties = permittivity + conductivity.

P12L9: remove “ash”. Ash (or tephra) doesn’t make a reflection but contrasts in chemical constituents do make a reflection.

P12L10: remove “temperature” as temperature varies only gradually and thus cannot make a reflection surface (if the ice body is not polythermal).

P13 Table 3 caption: Probably better to say “Comparison of reflectior depths from OIB radar data with snow height changes for the corresponding period.” Partly because this table appears before the main text mentions it, it is quite hard to understand what this table really means. Also, add “Trusel et al. (2013)” as the reference for scatterometry.

P13 1 and 2 lines from the bottom: “the OIB radar data over LCIS” and “characteristics of the firn including melt features at the much higher OIB radar frequency”

P16L20-21: This sentence is confusing. Please revise.

P19 Figure 11: Include the scale of arrows.

P20L13: previous -> last

P20L16: change to “track the shallowest melt layer and derive the winter SMB immediately prior to the radar survey”

P21: Please move the data availability statement that currently appears in the acknowledgements to the end of conclusions.

Dear authors,

The revisions that author made are satisfactory, so I am pleased to accept this manuscript as is. The manuscript was improved significantly through the review process so I would like to suggest the authors to acknowledge the anonymous reviewers.

Kenny Matsuoka
TC/TCD Editor