Dear authors,

Please see the attached pdf for reviews. Looking forward to the revised version.

Sincerely,

Kaihe Yamazaki

5 repeated glider surveys across the Southern Boundary (SBDY) of Antarctic Circumpolar Current are used to investigate SBDY’s cross-frontal behaviours under eddy and non-eddy regimes. Eddy presence enhances cross-frontal density gradient supressing the cross-frontal mixing whereas eddy absence, comparing to eddy presence, is accompanied by a weaker cross-frontal density gradient. These results are interpreted under the context of a multidecadal evolution of SBDY speed/location derived from satellite data. Authors concluded that the enhanced eddy activities and accelerated SBDY are occurring at the same time in opposition in affecting the meridional exchanges of tracers cross SBDY at Greenwich Meridian. I found this work is interesting and potentially important for the community in understanding the Weddell Gyre heat content evolution under the context of climate changes.

I have one concern about this manuscript. This work highlights that the different cross-fontal properties are associated with eddy presences exemplified by comparing transect A and transect C. These contrasting results between eddy and non-eddy regimes need to be strengthened by a quantified uncertainty that could be raised from different glider sampling intensity along the transect because it seems to me that the transect C does not take profiles as frequently as transect A by looking at the station distribution from two transects. See also the relevant comments below. I am happy to see this manuscript published once my concerns herein are addressed properly.

General comments:

1. Most results present in this manuscript based on the comparison between transect A and transect C, where the authors argue that eddy presence/absence is the reason for the observe difference. The glider station (marked as triangle on top of cross-section plot, most evident in Figure 8) distribution between A and C is different. Can author quantify the potential uncertainty caused by different glider station distributions on the present cross-frontal difference?
2. This may or may not be resolved by typesetting, but I found that quite a few figures are far from where they were discussed. For example, section 3 in page 8 discussed Figure 4 to Figure 8, while Figure 8 is displayed at Page 14. I suggest authors to condense down figure volume, such as, putting multiple subpanels into one integrated figure, leaving the results for transect B, D, E in Supp Mats as they were barely mentioned, T-S plots with highlighted regimes taking up one subpanel spaces can be replaced by combining mainly discussed regimes in one T-S plot and mask other data points with grey colour, etc.

Specific comments:

L6: ‘quite rapid’→‘high-frequency’ or ‘transient’?

L35: delete ‘globally’, the word ‘globally’ is misplaced as the SBDY is not a global feature, is it? ‘Climatologically’ is sufficient here.

L42: ‘…further represent the southernmost boundary to mixing’. I found this sentence a bit ambiguous… I believe that the mixing process in general is happening everywhere, and I don’t think authors have set the context of using the term mixing to refer the cross-frontal mixing happened at the SBDY.

L60:’The majority of studies almost entirely…’, need refs here or is author referring to aforementioned studies? If so, please indicate.

L160: ‘converge’. The T-S plots do not show this ‘convergence’ particularly clear. Adding arrows to indicate this in the T-S plots.

L164-167: I do not fully understand this. The similarity of the properties between eddy and south of SBDY suggest eddy originated from south of SBDY, okay, then what is the meaning of mentioning the slight temperature/salinity difference below/above the thermocline? Plus, why do authors mention the salinity difference in reference to thermocline?

L179-182: Mention the criteria and the table somewhere earlier in the section. This section has covered many figures that use such color-coding criteria. Best to mention it in the first place to avoid confusion for readers.

L185: the eddy passage could be one of the reasons for the difference in horizontal density gradient between transect A and C. Figure 8 shows a smooth ADT for C and rough ADT for A which makes me realize that the profiling intensity of A and C is also different. Transect A has more profiles in general than Transect C across the front. Does this fact play any role? Authors should quantify the uncertainty on horizontal density gradient caused by different sampling intensity by subsampling a high-res model results/reanalysis or any other sensible measures.

L204: It is not clearly stated how the temperature fluctuation, θ', is computed.

L285: The discussion on the long-term behavior of the SBDY and its core speed is sufficiently supported by literatures. However, the sea ice extent seems to be a bit out of place here. I suggest authors to either specify the reason for examining sea ice extent and discuss it extensively in the context of past literatures or simply not to show the sea ice extent at all since it does not correlate well with the available data here and authors just briefly mentioned it… Sea ice advancing and retreats on yearly basis is also controlled by large-scale wind variability, thermal forcing and also internal sea ice dynamic, so it perhaps requires some extra effort to decipher sea ice extent in the context of enhanced frontal jet.

L294: If authors are referring to the positive SLA blobs into the 2010s, then perhaps the phrase ‘anti-cyclonic eddies’ is more appropriate than warm core eddies? Studies have shown that not all anti-cyclonic eddies have a coherent warm core structure throughout the vertical extent.

L309: ‘…. consistent with Williams et al. (2007) who demonstrated …’

L301: ‘… in all transects …’, authors mainly discussed transects A, relevant results for B, D, E should be included at least in Supp Mats to make this claim.