Review of

“The effect of storms on the Antarctic Slope Current and the warm inflow onto the southeastern Weddell Sea continental shelf”

by Dundas et al.

The authors wind-derived anomalies of surface stress caused by storm events over the southern Weddell Sea, upstream of the Filchner continental shelf. They then investigate the impact of those periods of high surface-stress on the Antarctic Slope Current (ASC) near the Filchner sill, and on the flow both of warmer waters onto the continental shelf, and the southward flow of warm waters already on the shelf toward Filchner Ice Front.

This work is a continuation of observational and idealized numerical studies by many of the same authors. Here the mooring time series has been significantly extended. Seven moorings, with time series up to four years in length have been used. Obtaining those moorings has been a colossal effort, and they represent a very impressive resource.

As a continuation, the study is in some ways incremental, providing confirmation of key findings from the previous work, but also raising some interesting questions. I would like to see it published in this journal, after some relatively minor revisions.

Overall, the English is good, in that it is entirely understandable. However, the text could be substantially tightened up, perhaps by a co-author? I’ve submitted a marked-up PDF with many comments and an incomplete list of minor textual suggestions, but very often sentences could be redrafted more concisely. That is perhaps an editorial decision. Some of the comments are more substantive but most are requests for clarifications that can be very easily dealt with.

A couple of more significant questions.

1. Section 3.4

This reviewer was a bit confused about what the authors were trying to say in this section, where they describe a shift in July 2019 in the response to storms events: the response on the shelf to storms went from being inconsistent to consistent, while the reverse was the case for the response at the sill. At the same time the flow direction on the shelf migrated from being primarily north-eastward to primarily eastward.

In line 275 they mention the importance of changes in the upstream wind forcing as a possible reason behind the shift as discussed in an earlier study, but later in the paragraph note that the mean surface stress over the Upstream box doesn’t seem to change during the shift. In the next paragraph (line 282) there is a comment about the correlation between wind direction and the current direction at M_CS2; the correlation shifts from negative to positive. Where is this wind? Is it over the Upstream box? If so, I don’t see how the mean direction of the stress isn’t changing, but the correlation between wind direction and current at M_SC2 is switching sign: the current direction is only changing by 45 degrees.

The paragraph starting at line 297 then seems initially to repeat the statements about the Daae et al paper’s findings mentioned in the para starting in Line 275.

I think this section needs to be tightened up considerably. Clearly, the authors have an interesting finding, and haven’t yet got an explanation that satisfies them. I feel that it could be explained very much more concisely and clearly.

1. Certainty in the ocean response

I think the authors have generated a time series of the strength of the westward component of surface stress and used an algorithm to identify storm events. They then calculate the strength in the response of the mooring time series around the time of each storm. To be reassured of the robustness of the identification of the response, would it be helpful to carry out a randomized test: create a set of random times of pseudo-storm events, and carry out the same calculation of the strength of the “response” as measured by the mooring time series. Carry out the same test for a many different sets of pseudo-storm events. Highly variable currents as measured by the moorings will often have peaks that will occasionally correlate with peaks in storm forcing, regardless of whether they are being caused by the storm events. A Monte Carlo-like approach such as this will make clear whether the relationship between ocean response and storm forcing is robust. If this analysis is not possible for some reason, perhaps sample time series from the current data would help give confidence in the relationship.

The authors present an observational analysis investigating the effect of storm events on warm inflow at the Filchner Trough. I am not that familiar with the existing literature on this topic in this particular region, so can not comment on the novelty of the work. However, I found the results interesting, the manuscript well-written and figures nicely presented. I recommend publication after my comments are addressed.

Major comments:

1. The authors appear to treat ERA5 as ‘truth’ and do not discuss how inaccuracies in the wind stress they use from ERA5 may impact their findings. How good is ERA5 in this region? Have any past studies validated it against nearby in situ weather station data (if any exists)? Is it possible that inaccuracy in the representation of wind stress in ERA5 could be the explanation for the complicated ocean response. e.g. Imagine ERA5 overestimates the wind stress for some storms but not others, could that explain why there is no response to some storms?
2. I am confused by the specific choice of the ‘Upstream box’. It would be helpful to discuss further the sensitivity of the storm response to the location of the ‘Upstream box’. The authors argue that it is not sensitive to location based on Fig A1. However that figure shows perhaps a correlation of only ~0.5 between the upstream box and the region closer to the trough inflow. Intuitively I would have thought that moving the box say 20 degrees west (to 20-25W) would result in a stronger connection to the inflow. If there is a physical justification why the chosen upstream box has the strongest connection to trough inflow, it would help future studies to explain further why this particular region is dynamically important.
3. I found it unclear whether total stress or westward stress is used in the identification and characterisation of storms. Mostly throughout the text just “stress” on its own without a direction is used. I would interpret this as total stress. But sometimes (e.g. line 130) “westward stress” in particular is used. Please be clear throughout the manuscript whether stress is total or westward only.
4. In relation to the last point, does the direction of the wind stress have an impact on the storm response? Based on the mechanism described in the introduction, I would have thought that along-slope wind stress would be more important than westward wind stress (which is what I think has been used).
5. Out of curiosity, how barotropic is the response in the moorings? From what I can tell, only depth-averaged flow is used in the analysis. Also, is there any response in temperature or salinity after storms? This point does not need to be included in the manuscript if the authors do not wish to. I am just asking in case there is something interesting to say there that could add to what is currently in the manuscript.

Minor comments:

1. Line 10: “increased southward current speed”. It would help readers who only read the abstract to clarify that this is speed of mWDW, and not e.g. surface speed.
2. Figure 1 caption: The units of ‘5 cm s−1’ are not formatted correctly.
3. On page 10: The authors use “fall” as the season. Please change to “autumn” or MAM.
4. Figure A2c needs an x-axis label.

The authors present a study into the effects of storms on the Antarctic Slope Current and southward heat transport towards the Filchner Ice Shelf based on data from a network of moorings. The study is a continuation of a previous model-based exploration of storm-driven flow in the region. While the mooring data had been used previously in various studies, the authors did a commendable job in pulling them together and providing a combined analysis. The manuscript is well organised and written. I only have minor comments for improvement before publication.

• Formatting of units: I’m sure the journal will apply their typesetting before publication, but there’s quite a mix of i) space or nos pace between number and unit, and ii) inconsistency in use of superscript, e.g., cm/s versus cm s^-1, even within the same figure (e.g., Figs. 5, 6, 9, A3).
• Line 23: typos/grammar: «intrusionS» and «extend»
• Line 49: what sort of distances does this traverse of the continental shelf imply?
• Lines 60-63: What is the Berkner mode then?
• Section 2.1: Do you have any information on the types of instruments, calibration and processing procedures?
• Figure 2: Introduce ADCP and explain the variable T, S, v...
• Line 92-93: Why do you treat M_CS2 differently?
• Line 96-102: How do you justify the choices to use depth-average currents (do you include the bottom sensors even though they stopped early?), longest time series, or depth with highest velocity? And which depths are those then?
• Table 1: Maybe add the relevant references for each mooring so that it’s easier to find information on the sensors and the processing?
• Line107: Should this be Figure 6 instead of 5?
• Line 123: Add that rho_water and rho_air are densities
• Line 165: Explain what cp and L_f are
• Sections 3.1 and 3.2: Check the figure references, I think they point to the wrong figure(s)
• Line 192: change “which is” to “which are”
• Line 231: correct bracket around reference
• Section 3.3: The text in this section should be streamlined a bit, could be more concise and precise.
• Lines 247-257: What about potential influences of seasonality in hydrography on the storm reponse?
• Discussion & abstract: I miss a broader impact discussion or statement – how much does this storm-driven heat transport contribute to the total heat transport towards the Filchner ice shelf, i.e., how important is it actually? And what are implications?