Dear Dr. Fogwill,

I now received two reviews for your revised manuscript, which are available for you at the journal web site. The reviewer #1 is overall positive; this reviewer sees that your revision is adequate and the manuscript is now publishable as is. However, the reviewer #2 is overall negative, and especially concern the modelling component of your work. The reviewer #2 argues that modeled features presented in the paper (Fig. 2C and 2D) do not match at all with the observed results (Fig. 1B and 1E), and I agree with this reviewer. The authors clearly mentioned that the model outputs for pre and post calving epochs are for two stable geometries, and do not account for the evolution (P7L13-15). It is a reasonable approach, and some features cannot be replicated due to this limitation. However, disagreements between Fig. 1B/E and Fig. 2C/D are fundamental. I am not yet convinced how much modelling is helpful to demonstrate impacts of developing polynya off Commonwealth Bay.

In general, presented model results (Figure 2) are inadequate to judge whether or not the model successfully replicated the observation, and whether authors’ implications are valid. For example, at P6L24f, the authors say “The cold and salty water mass simulated pre-calving at the NW Commonwealth Bay XCTD positions is advected from the Mertz polynya and Commonwealth Bay post-calving.” Similarly, at P7L3-4, the authors claim “…. that originates from the advection of HSSW from the Mertz polynya and Commonwealth Bay”. I speculate that the authors use the vertically-integrated horizontal velocity presented in Fig. 2A/B as supporting evidence of these statements. However, the text is unclear about this point. More importantly, if the modeled velocity is used to support the HSSW advection, the modeled velocity should be presented at depths where HSSW is expected.

The author also claimed that there is an active polynya in Commonwealth Bay NW and generate HSSW that is convected to the sea floor (P7L11). However, it is unclear for me whether this water mass with the uniform salinity is locally made or advected from somewhere.

The results section includes discussion-type statements, beyond the results, which should appear in the discussion section. This confuses readers to judge what the data and model really show respectively, and what are author’s arguments based on the observation and modelling. In Section 3.2 (P6L18-20), sea ice production is presented as “modeled” but it is basically taken from Tamura et al. (2016) (P4L23-24). So, cumulative sea ice production presented in Figure 2A and 2B are somewhat observations, not really modelling outputs. I see a similar problem at P8L5.

I would like to invite you to submit a revised version. You can find reviewer’s comments on the journal web site, as well as more editorial suggestions that I made below. Reviewer #2’s comments are not really specific, but please adequately respond overall concern that this reviewer made. If the page limit for brief communication is too tight to make adequate revision, we can consider the revised manuscript as a research article, not as a brief communication. If you think that the modelling component remains weak, please consider removing it and present only the observational side of your work. Both reviewers say that the observed feature is of high interest and worth to publish, which I completely agree.

I hope that these suggestions are helpful. We look forward seeing a revised manuscript.

Thanks again for submitting your work to TC/TCD.

Kenny Matsuoka
Editor of TC/TCD

Specific comments from the editor:

P2L24: remove one “)” and double “the”

P4L12: Why is it necessary to show the Microcat CTD data? The main body of the manuscript does not use the Microcat data at all, and it appears only in Supplement Figure S1.

P4L23: model/ocean coupling?

P5L24: Which depth range do you expect to see HSSW layer? It is said that the observed depth range is too shallow to observe it. The observed fact is the absence of HSSW, and inferring HSSW at greater depths is author’s speculation in this case. Please more clearly distinguish observed facts and interpretations/discussion.

P6L10: add a reference to support a statement on the absence of HSSW prior to the B09B grounding.

P6L18: as the reviewer #2 pointed out, Tamura et al. (2016) is a boundary condition for the modelling, not a modelling output.

P6L24-P7L1: It is confusing. Fig. 2C/D show that shallow water is warmer and fresher (red solid curves) than deep water, which is said immediate above. So, these statements are not consistent.

P8L8-10: Revise “The effect this change of…”

Figure 1:
- The caption can be clearer, starting for example “XCTD data. (a) locations of ….”. It’s better to say clearly what Figure 1 is about at the beginning, rather than what the panel a is about.
- Blue labels on black parts of satellite imagery are very hard to read (especially on hard copy). Change the colour, or use a lighter colour to outline these letters.
- For panels (B)-(G), legends are common so it is adequate to show it in a single panel. In the text, the authors emphasized that salinity/temperature can be changed seasonally so data can be compared when they are collected in the same season. So, please indicate the month of these data together with the year; i.e. “December 201x” or such.
- It’s hard to compare panels (B)-(G), because ordinate/abscissa of individual panels are not scaled. Are there any particular reasons not to scale these panels? Indeed, we need only one left axis for two panels next to each other. So, make only one left axis (not two left axes) and put one right axis to show approximate depths so that more readers can easily compare the observation and model results.
- Panels (B) and (E) do not show the post calving 2012 data (green curves), and post calving 2011 data (black curves).
- Show the seafloor in panels (B) to (E).
- Many CTD positions are shown but only for the 2013. Can you show 2013 XCTD positions only if the data are presented in panels B-G and add previous data points that are currently shown in Figure S2? In other words, please consider presenting information in Figure 1a and Figure S1 in a single panel more efficiently.

Figure 2:
- Depth should be all positive numbers.
- I see two curves for each legend in the each panel; e.g. there are two blue solid curves in panel C. Why?
- Please add a short sentence at the beginning of the caption to emphasize that you show model results in Figure 2.
- Show the seafloor in panels (B) and (E).

Supplement P2L13: “fluxes”

Supplement P2L15: Fig. S3 is cited before Figs. S1 and S2 are cited. Please make sections in
the supplement that are associated with Figures S1 and S2 so that the points of these figures are clearer. Showing just figures in supplement does not really help.

Dear Dr. Fogwill

Your manuscript was reviewed by one reviewer and myself, and now I am happy to accept this manuscript subject to minor revision.

Your revised manuscript will go through my own review for the final decision. Please provide the revised manuscript with editorial tracking, as well as point-to-point responses to all four points brought by the reviewer.

Thanks for submitting your manuscript to The Cryosphere.

Kenny Matsuoka
TC/TCD editor

Dear Dr. Fogwill

I am pleased to inform you that the manuscript is now accepted by The Cryosphere. Through the review process, the manuscript has been largely improved. Thanks for all of your work to address suggestions brought by reviewers.

Kenny Matsuoka
TC/TCD editor