Summary: The authors use a high-resolution sea ice-ice shelf-ocean model to study the freshwater fluxes in coastal polynyas over the Weddell Sea continental shelf. With the help of sensitivity experiments, the authors show that the presence of an icescape, formed by an ice mélange-bridge between a grounded iceberg and the ice shelf, leads to a blocking of the sea ice movement and, thus, strongly influences the freshwater fluxes through convergence and divergence of the sea ice. Further, a correct representation of ice growth and freshwater fluxes can only be simulated with the use of a high-resolution regional atmospheric forcing that correctly represents local winds. The freshwater fluxes within the coastal polynyas determine the density structure on the continental shelf and consequently the circulation and basal melt rates of the ice shelf. With the help of the different model simulations, the previously observed switch between a Ronne-sourced HSSW production and a Berkner-sources HSSW production can be shown, revealing the importance of a correct representation of the atmospheric forcing and sea ice conditions on the underlying ocean.

General comments: The model setup and simulations are well described and the resulting ice growth and freshwater fluxes agree with observations. The figures and tables are comprehensive and the text is well written. Given the importance of coastal polynyas in this region for ice shelf basal melt and deep water formation, the study is relevant and interesting. However, I suggest some changes mainly in the structure of the paper for easier readability and for an improved highlighting of the relevance of the study.

The paper feels at times very much focused on an evaluation of the model and very packed with comparisons with previous studies and different values, which could be shortened or summarized (e.g. when the values themselves are also in a table, or considering to first provide the results from this study, then comparing with previous results f.ex. lines 220-226). Some parts could be considered to be moved to the discussion section (e.g. lines 212-216 or some of the comparisons such as lines 245-272). Instead, there could be a stronger focus on the physical results, such as the implication of the salt production on HSSW production, circulation inside the cavity/ on the continental shelf and basal melt. The impact of the polynyas on HSSW seems a very important part of this paper, especially in regard of the new studies about the Berkner and Ronne mode (Janout et al and Hattermann et al). The impact on HSSW is described in the discussion and nicely summarised in the conclusions, but in my opinion most of the discussion section is actually a result and would gain more weight by moving it to the results.

Specific comments: 

Title: It might be useful to have the term polynya in the title, as people looking for sea ice production might search for the term polynya in google scholar.

Abstract: It could be condensed or restructured to better highlight the main results (e.g. some details such as the MODIS satellite in line 10 is enough to have in the methods, and line 5 can be merged with line 8-10.. ) and there could be a larger focus on the impact of the polynyas on the HSSW production (which is mentioned in the last sentence), as it currently sounds as if the last sentence is more of a general statement rather than an actually conclusion based on the model simulations and calculations done in this study.

Line 5 and 8-10: It is unclear from the abstract what an icescape is. In line 5 is seems like an alternative term for the grounded iceberg rather than the ice bridge between the ice shelf front and the grounded iceberg. Also, it is unclear to me what the difference between an icescape and an ice bridge is...

Line 15: i) treatment of the icescape – not clear what you mean with treatment; better “existance of the icescape”?

Introduction (Line 66-67/ line 71): Since this study has a strong focus on the icescape and the sensitivity of the freshwater fluxes on its existence, it would be easier for the reader to understand the motivation for these sensitivity experiments, if there was a description of the icescape – is it a permanent feature, does its shape change, what do we know from satellites? etc... I suggest to move (parts of) the first paragraph of 2.3 into the introduction.

 In general, the introduction could benefit from a more physical motivation of this study. There is currently a strong focus on what other studies have not managed to resolve/look into, but there is much more value in your simulations: How have icescapes/ fast ice/ icebergs influenced coastal polynyas and salt fluxes in other regions around Antarctica? Why is it important to study the impact of the icescape in the Weddell Sea in detail? The switch between the Berkner and Ronne mode for HSSW production is well described later, but it provides a great motivation for this study and could already be mentioned in the introduction.

Line 107: Is there a reason to use ERA-Interim instead of ERA5?

Lines 109: Reading the description of the sensitivity experiments the first time, it was not very clear to me that BRIDGE means an experiment with the existence of the time-varying icescape as observed from satellites (the name seems arbitrary if you otherwise call it icescape).

Line 113: “without prescribing the blocking effect of the varying icescape”  – noBRIDGE is an experiment where the ice melange between the iceberg and the ice shelf is removed, right? It would be useful to describe a bit more what is done in the noBRIDGE experiments (either here or in section 2.3).

Line 116: The experiment statBRIDGE is listed in Table 1, but not mentioned here.

Line 120: “until beginning of 2022” – until present (?); it reads as if it eventually got ungrounded in 2022, which I believe was not the case?

Line 183: Is the 70% threshold or 20cm ice thickness a common criterion? Do you have any reference?

Line 192: A more recent basal melt rate study is Adusumilli et al 2020, how does the pattern compare with it?

Fig 4c: Are the values and average over the whole are shown in Fig. 4a?

Line 242-244: It is really nice to have these values of ice production in the different regions... How about adding a figure with the ice production for each of the regions? Either the annual production for each experiment or maybe a figure like 6a split into the different regions to also see the interannual variability.

Line 270: “tends to reduce polynya area and ice production”... Possibly explaining the smaller values in Paul et al.

Line 318: Fig. 3b shows that ERA is also in general warmer. How can you distinguish between the changes in ice production due to temperature vs. wind?

Lines 378-380: If there is an output of velocity from the model it could be useful to add velocities/ streamlines in Fig. 7

Line 425: This is not clear... You mean “Using a high resolution/ more realistic region atmospheric forcing ...”

Fig. 5a & b: Are only the velocities from August 2009? Why are you showing that months rather than an average?

Technical comments:

Line 120: hyperlink does not work with the line break.

Line 162-163: remove “by”: ERA has 2C ... while it is 0.5C ... and 2C cooler...

Line 367: “too strong salinity” -> too high salinity

Line 371: “is in BRIDGE is 3.2Sv” -> remove one “is”

Line 376: reference in brackets

Summary: This study examines the impact of the distribution of fast ice and grounded icebergs and atmospheric forcing on sea ice production on the southern Weddell Sea continental shelf. These processes are investigated using an ice shelf-sea ice-ocean model with prescribed icescapes and atmospheric forcing. The sensitivity of sea ice production to the presence of fast ice is diagnosed by imposing zero sea ice velocities in regions of fast ice (as determined by satellite observations). This analysis finds that the representation of fast ice influences the spatial pattern of sea ice production within polynyas, with enhanced ice production in the lee of fast ice. The patterns in ice production also correspond to the distribution of High Salinity Shelf Water (HSSW), which has implications for bottom water export and ice shelf cavity circulation. The study finds that using atmospheric forcing from a regional high-resolution atmospheric model yields a more realistic representation of sea ice production.

General comments:

This is a clearly presented assessment of the impact of the stationary ice features on sea ice production in the southern Weddell Sea region. The results provide convincing evidence that the relatively small-scale details of fast ice and grounded icebergs can impact the spatial patterns of sea ice production, bottom water export, and ice shelf melting. I believe this study has the potential to become a valuable addition to the Weddell Sea literature. However, I would like to highlight a few issues for the authors to consider and address.

- The manuscript motivates the importance of the southern Weddell Sea shelf by highlighting its outsized role in sea ice and HSSW production, which have far-reaching impacts on the global ocean overturning circulation. Though the manuscript makes a convincing argument that the details of the icescape are important for local ice production and water mass formation, it does not make a compelling case that these processes matter much away from the southern Weddell Sea region. From Figure 6a, one could argue that none of the sensitivity experiments had a meaningful impact on the total amount of sea ice leaving the region. While this is certainly a valid result worthy of publication, it is tempting to conclude that the nuances of the icescape in these high sea-ice production regions do not have an appreciable impact on the circulation and water mass properties of the Weddell Sea and the broader Southern Ocean. If the authors dispute this conclusion, I encourage them to clarify this point in their abstract and conclusion.

- Compared to the sensitivity analysis of the icescape, the assessment of the atmospheric forcing feels like an afterthought. While Section 3.3 highlights many differences in the experiments where the ice-ocean model is forced with ERA-interim versus a high-resolution local atmospheric model, the significance of these discrepancies is not clear. It is well-established that sea ice production within coastal polynyas is sensitive to local winds. Thus, it is not surprising the two experiments yield different sea ice production rates. The value of this comparison is further diminished by the fact ERAinterim is an outdated global reanalysis. Further, since neither experiment is rigorously compared to observations, this assessment does not provide any validation of CCLM. I do not see much value in discussing the cATMO experiment.

- This is a somewhat minor point, but it would appear that the use of the term "icescape" in this study differs from its common usage in the literature. Here, icescape refers to stationary ice features, such as grounded icebergs, ice shelves, and land-fast sea ice. However, in other studies, the term is used as a catch-all descriptor of all forms of ice, including snow and transient sea ice, regardless of its motion or lack thereof.  For consistency and precision, I recommend that the authors clarify that their study explores the influence of stationary ice features on sea ice production.

In addition to the above points, I have detailed some minor comments and suggestions below.

Detailed Comments:

- Lines 31-32: I generally associate basal ice shelf melting with the intrusion of modified CDW or Warm Deep Water. In some studies (e.g., Hazel and Stewart 2020), the presence of HSSW within the ice shelf cavity is described as a "cold state" that is characterized by low levels of ice shelf melting. Please clarify.

- Lines 80-90: While referring readers to previous studies that have used FESOM is appropriate, more details need to be presented here. In particular, a description of the ice dynamics scheme and specifications along the open ocean boundaries should be provided.

- Line 87: How does using a time-varying S_ref impact freshwater conservation?

- Line 93: Please specify that "A" represents the sea ice area.

- Line 100: Please comment on how well this horizontal grid spacing represents mesoscale eddies.

- Line 107: Why not use ERA5?

- Line 110: At this point in the paper, it is not apparent why these experiments are referred to as BRIDGE and noBRIDGE. Perhaps add a sentence clarifying the names are in reference to an ice bridge that will be described later.

- Line 113: Please clarify what is meant by "blocking effect." Blocking has various meanings in oceanic and atmospheric literature.

- Line 132: I suggest the authors delete "best possible" or provide evidence that you have rigorously optimized your methodology.

- Line 138: How can one distinguish between warm surface temperatures caused by surface melting versus ice thinning? It seems like this approach may misclassify a region with substantial snowmelt over thick sea ice.

- Line 188: Or, more precisely, "where the vertical entrainment of Warm Deep Water into the surface mixed layer drives basal melting."

- Line 208: Here and elsewhere, I suggest explicitly defining non-SI units when they are first introduced (e.g., mSv and Gt/a).

- Line 218: Delete a^{-1} or "annually."

- Line 219: Same issue as above. "With the annual net sea-ice export of 1041 km3 a−1..." Using a^{-1} or "annually" is redundant.

- Line 236-238: "substantially higher ice-growth rates..." This sentence is hard to follow. Please re-write for clarity.

-Lines 245-260: It would be helpful to reference Table 2 somewhere in this paragraph.

-Line 276: "where" should be "were."

-Line 290-291: In addition to the triangle icon showing the grounded iceberg, it would be helpful to add contours or shading outlining the time-averaged area where the ice velocities are modified.

- Line 293: I don't fully understand the ice thickness and production anomaly maps in Figure 5b,d. From ice velocities, one would expect less ice accumulation upstream (to the west) of the ice bridge when the blocking effect is removed. However, Figure 5b shows the opposite pattern. Perhaps related is the fact that the ice velocities in Figure 5a are almost 90 degrees to the right of the surface winds in Figure 3c.

- Line 293: Also, why do Figures 5a and 5b only show values from August 2009? What is unique about this period?

Line 319: Please clarify what the 19% refers to.

Line 320: It is not apparent that these ice production values are statistically different.

Line 325: A few words appear to be missing. "For 2002-2017, we find polynya ice production [to be]..."

Line 330: Are these trends for the 2002-2017 period?

Line 361: Please specify the density range used to define HSSW.

Line 384: It would also be helpful to show the circulation anomalies of HSSW.

Line 387: Please see my earlier comment about the heat supply associated with HSSW.

Line 400: From the results presented thus far, it is not evident that the density variations in the southern Filchner cavity reflect an increase in the volume flux of Ronne-sourced waters or a change in the property of waters produced at Berkner Bank or something else entirely. It might be helpful to show the shelf cavity circulation in an earlier figure.

Figures and tables:

- Figure 4a: please note in the caption that the colormap does not linearly scale with the numerical data.

- Figure 5: (a,b) Why only show values from August 2009 here? (b, d, and f): Note in the caption that the colormap does not linearly scale with the numerical data.