Review of Nissen et al., 2023 - Altered Weddell Sea warm- and dense-water pathways in response to 21st-century climate change

The formation of deep and bottom waters on the continental shelf in the Weddell Sea is a key mechanism for sequestering carbon and exporting heat to the abyssal layer. This study by Nissen et al. explores the pathways of warm waters onto the shelf and dense waters flowing off the shelf under both historical and future conditions. They used output from the global ocean-sea ice model FESOM-1.4 to conduct Lagrangian experiments in the Weddell Sea in 1990-2009 (historical period) and in 2080-2099 under the high-emission scenario SSP5-8.5 (future period). The depth where particles reach the Filchner Ice Shelf front and where they originated in the open ocean and crossed the shelf break is evaluated as well as the pathways. In the historical period, particles mostly reach the ice shelf front within the upper 200 m and originate from the upper open ocean, suggesting weak inflow of modified Circumpolar Deep Water into deeper parts of the ice shelf cavity. Under a high-emission scenario, more particles reach the ice shelf front by the end of the 21st century, originating from greater depths offshore and reaching the ice shelf front at greater depth compared to the historical period. This increased onshore transport has important implications for future ice shelf basal melt rates and deoxygenation. In the future period, the transit time from the shelf break to the ice shelf front decreases by 6 months and cross-shelf flow becomes more likely in winter. Particles leave the Filchner Ice Shelf front mostly below 600 m (that would be within the ice shelf cavity), flow down the western shelf break of the Weddell Sea and enter the open ocean as dense waters mostly below 800 m. In the future period, the particles trajectories are overall shallower and especially in the open ocean the depth reached by the particles is greatly reduced with only 60% below 500 m and none below 1500 m, but their horizontal distribution is similar. Most particles reside on the shelf for a year and leave the shelf in autumn and winter in the historical period, while in the future period the median residence time on the shelf is 21 months and spring is as important as winter for the export across the shelf.

This is an interesting study and important contribution regarding the cross-shelf transport in the Weddell Sea using Lagrangian methods for the first time. The manuscript is very well structured and written with the figures illustrating the results and supporting the conclusions. I have a few minor suggestions which might help to further improve the manuscript.

As a paper should be self-contained, more details should be given when referring to results from your previous papers (Nissen et al., 2022, 2023) which are fundamental for understanding the conclusions. Especially the mechanisms for the changing pathways in the future period are not discussed in much detail and this leaves readers with open questions when they are not fully aware of these previous studies. Overall, I think that this manuscript is a valuable contribution to the community and should be published in Ocean Science after some minor revisions detailed below.

Specific comments:

• ll. 6-11: Rearranging these sentences (and potentially including additional results) would clarify the results and the conclusions made. The results should be described more clearly by e.g., referring to depth changes in meters as well. In addition, the implications could directly follow individual results and not be stated a few sentences afterwards.
• ll. 84-85: A few sentences should be added on what has been evaluated in Nissen et al. (2022, 2023) and what biases were found to make the paper more self-contained.
• ll 174-176: Be more specific about the mechanisms detected in Nissen et al. (2022, 2023), e.g., increasing on-shelf heat transport with negative cross-shelf break density gradient in future period.
• Figures 2 and 6: These figures illustrate the results really well, but they could be improved. The thickness of the arrows doesn’t seem to be “scaled with the relative importance of different depth intervals for the origin of these waters before the crossing and their fate after the crossing, respectively”, as stated in the figure caption. For example, in Fig. 2a the thick blue and yellow arrows representing flow from the open ocean on to the shelf in the upper layer should represent about 80% of the particles but this is not represented by their combined thickness. Can you depict the changes in the vertical distribution offshore better by changing the position of the arrow at the right axis of panel a and e? For example, in Fig. 6e the red arrow should point at a shallower depth offshore (approximately 600-700m) compared to panel a. In my opinion the size of the figure is too small, especially the text in panel a and e is hard to read, and the lower panels should be enlarged. This could be achieved by having the panels corresponding to the historical and future simulations in separate rows instead of columns.
• ll 271-273: Be more specific about what changes were detected in Nissen et al. (2022), e.g., reduced ventilation rate along the slope, reduced surface water mass transformation and decline in density in future period.

Technical comments:

• caption of Fig. 1: Include description of black and grey contours in the figure caption.
• ll 309: Be more specific, what percentage of particles is referred to as “sizeable”?

This is a fine study with a focus on the Weddell Sea. It is well written and the only complains I have is that is that the reader is expected to fully trust the model. To do so, she must be given much more details:

line 83:

What is the quality of the present day forcing? It seems to come from a coupled model, so there should be figure with biases in sea-ice and winds at least.

What is fidelity of the ocean circulation? Can you show a comparison with some data, in particular with velocity- if available?

The model has some 5km resolution, so if the particles are advected with daily (mean, snapshots?), the structure of the mesoscale is lost. Does that matter? What is the advection time step?

Section 4. I would probably move it to section 2.1, as part of the model setup. Your statements here suggest that open ocean convection is not important in the Weddell Sea, and the z - levels make a downslope current unlikely. So how do the particles get deeper? Is it simply flow along isopycnals? Or are they advected across density fronts (see comment above)? Can you provide a plot that shows some of particle trajectories in a vertical plane?