EDITOR COMMENTS on cp-2020-155, Green et al. “Evaluating seasonal sea-ice cover over the Southern Ocean from the Last Glacial Maximum” submitted 03 Dec 2020

The authors have put careful thought into addressing the comments of 3 reviewers and one posted comment. Having read their responses and noted the positive inclusion of the new, PMIP4 analysis, I think that the authors are in a good position to move forward to “reconsider after major revisions.” In addition to the well-documented changes planned by the authors, I have noted the following for the authors to consider in their final manuscript revision:

1) The authors have indicated that one of the goals of this manuscript is to provide updated LGM proxy datasets. I suggest that the authors could better document in their methods section how they harmonized chronologies (what was the choice of LGM time period), methodologies, and determined consistencies (if needed).

2) Similarly, as the LGM proxy data set update was an identified goal, the authors should devote at least one paragraph at the outset of their Results section outlining what the new compilation shows (prior to their comparison to simulations). Such a paragraph could also help to address concerns of R2 regarding zonal asymmetry. How prominent is zonal asymmetry in WSI cover for both time periods in the proxy data?

3) R1’s comments about the feedbacks between winds and sea-ice are important ones. In their response, the authors do a thorough job of clarifying that the goal of the paper is not to address (a) how accurately the models reproduce winds, or (b) how sea-ice feedbacks ultimately influence wind positions. Rather, the authors are focussing on each model’s relationships between winds and sea ice.

The first point (a) appears to be well-addressed by the proposed text for Results section 3.4 (below), but the additional sentences added to Results section 3.3 to address point (b) (also) below jumps topics and does not appear positioned well. I suggest adding this response to paragraph just above section 3.2 (ca. line 155 in the original manuscript), and to state it more clearly as “we recognize that the presence or absence of sea ice also has a direct influence on surface winds (Kidston et al. 2011; Sime et al. 2016). However we are focused on the influence of winds on sea ice….”

TEXT FROM AUTHORS: Results section 3.4:
“While latitudinal position and magnitude of southern hemispheric westerlies at the LGM is poorly constrained (Kohfeld et al., 2013; Sime et al., 2016), here we want to assess the impact of the simulated windstress curl on ocean dynamics in each model. We thus use the simulated windstress outputs to estimate the location and strength of the SO upwelling, and its potential impact on sea-ice cover. “

TEXT FROM AUTHORS: “Results section 3.3:
“The strength and location of the southern hemispheric westerly and polar easterly winds impact Southern Ocean circulation, sea ice transport and therefore sea-ice distribution (Purich et al., 2016; Holland and Kwok, 2012). On the other hand, the presence or absence of sea ice also has a direct influence on surface winds (Kidston et al. 2011; Sime et al. 2016). The divergence created by the wind stress curl over the Southern Ocean leads to an upwelling of warmer deep waters and thus heat loss to the atmosphere. This upwelling can therefore also impact Southern Ocean sea-ice distribution.”


4) R2 is concerned that zonal averages will smooth out zonal asymmetries in the sea-ice edge, but authors are unclear on a better way to conduct this analysis. Note that when using paleo-data we get around issues of zonal asymmetry by using a feature like the modern-day polar front and expressing changes in variables (e.g. LGM minus PI) against location, where location is expressed as latitudinal difference from the modern-day front. This would be one approach for addressing this concern. Understandably, although this approach may be very useful, this new analysis may be beyond the scope of this manuscript at this stage. However, it would be useful for authors to address this limitation in the manuscript by (a) describing any zonal asymmetries observed in the paleo-data and (b) explaining that their analysis does not address how these asymmetries can affect the overall (zonally averaged) estimates of summer and winter sea-ice extent.

5) Please include latitude/longitude grid lines in your new figures so that readers can see latitudinal positions of data and modeled sea ice extents (e.g. Figure 1).

6) The authors might better address R2’s comment about regridding of model simulations by simply indicated in figure captions where this has occurred, or putting one sentence in the methods.

Thank you for this revised manuscript which provides a thorough response to all reviewer and open comments and therefore is ready for publication following four minor corrections:

LN 175-176: for clarity, please change "extending past" to "extending equatorward of" to indicate direction of change in sea ice extent
LN 247-249: the definitions of proxy/model lines appear to be reversed in the text (proxy should be solid black line; model dotted) relative to figure 3.
LN 348-349: For clarity, can you please specify that the relationships are with SSI, rather than the more general use of "sea-ice extent"?
LN 358: To make your sentence parallel, can you please also provide the approximate, seasonal range of the sea-ice extent in degrees latitude for the modern day as well?

Thank you for an interesting and thorough paper.