This manuscript by Si et al., titled "Late Holocene cooling and increased zonal asymmetry in the mid-latitude North Atlantic," presents new records of alkenone-based sea surface temperature (SST) estimates and planktonic foraminiferal analyses from two sediment cores in the North Atlantic. The authors integrated these new data with a suite of paleoreconstructions to examine the spatiotemporal trends of SSTs in the North Atlantic inter-gyre zone from the early to late Holocene. The study addresses an important topic related to the discrepancy between model simulations and proxy data, an important and still-debated “Holocene conundrum”, and it has the potential to make a useful contribution to the field. However, in its current form, the manuscript requires revision before it can be considered for publication, specifically regarding the overall structure and flow of the manuscript.

Below I outline several concerns and suggestions that require the authors' attention.

Specific comments

Title: I think “Middle to late Holocene cooling” will be more appropriate.

Abstract: The authors should first present the results from their own cores (as it currently reads more like a synthesis of previously published data) and then demonstrate how these new data are integrated at the regional scale (mid-latitude North Atlantic) to support their findings regarding an increased zonal SST gradient in relation to inter-gyre circulation.

Introduction: The introduction is well written; however, in my opinion, the authors should better summarize previous studies in the North Atlantic and further discuss the current understanding of the mid- to late-Holocene cooling trend recorded by proxies in this region in order to more clearly identify the existing gaps.

Sections 2 and 3: While Section 2 can be accepted as a regional setting, except for the last paragraph, which should be moved to the Materials and Methods section, Section 3 is problematic and, in its current form, makes the manuscript more difficult to read. This section describes some of the methods adopted by the authors, namely the calculation of productivity-weighted SSTs and their comparison with (i) annual-mean SSTs to compute the differences and (ii) core-top SSTalk, and it also presents the results of these calculations. To improve the overall structure and flow of the manuscript, I suggest that the authors split this section between the Materials and Methods and Results sections, after first presenting their own reconstructions from the two IODP cores.

Materials and methods section:

As highlighted above, a large part of the content currently included in the previous two sections should be moved here. I also recommend organizing this section into subsections to improve clarity and readability.

Age models: It is unclear why the authors used the Marine13 calibration curve rather than Marine20. The authors should also explain why the ¹⁴C dates of the other cores included in their compilation were not recalibrated. Given that some of these records have low temporal resolution, it is not clear whether this choice could influence the calculation of mean SSTalk over the two selected time intervals.

Heaton, T. J., Köhler, P., Butzin, M., Bard, E., Reimer, R. W., Austin, W. E., ... & Skinner, L. C. (2020). Marine20—the marine radiocarbon age calibration curve (0–55,000 cal BP). Radiocarbon, 62(4), 779-820.

Results: Please present your new reconstructions first and then the result of your compilation as I suggested before. Why only SST-alk from core U1304 are presented here and in figure 2? I couldn’t find those from U1308! Also, for paragraph 165, it’s not clear from which core the alkenone concentrations are. This section should be substantially revised to make it clearer for the readers.

Discussion and conclusions: Overall, I found these two sections well written, except for some typos, which I have listed below.

Figures: The figures are of high quality; however, in general, their captions lack sufficient detail.

Figure 1: Please mark your two cores with different color. I also suggest to move Figure 1b to a new Figure 3.

Figure 2: Please add the name of the core for 2d and move the legend for foraminiferal assemblage to be in parallel.

Technical corrections

Lines 36-37: Please rephrase and further explain this sentence. How can variations in Earth's precession simultaneously cause both cooling and warming at lower latitudes?

First paragraph of the Materials and Methods section: Please provide more information about the two cores (e.g. water depth, core length, and geographical location). In addition, please indicate here that core ODP U1304 was also used for planktonic foraminiferal analyses.

Line 217: Please add references.

Line 218: What do you mean by “Starting ~11 ka”?

Line 225: Please rephrase this sentences for more clarity.

Lines 233-234: Please add the core names used by Colin et al., 2010 or add "Rockall Trough" on Figure 1.

Lines 236 and 238: Please be consistent when you write "subpolar mode water".

Line 245:  Please explain how “the termination of deglaciation after ~8 ka” has strengthened the subpolar gyre.

Line 247: “Nevertheless, the effects of stronger subpolar convection appear to have been limited to subsurface layers or winter conditions.” Please further clarify the underlying assumption and support it with appropriate references from the existing literature.

Line 262: “a weaker subpolar gyre over the last 6 kyrs” are you sure? According to Tornalley et al., 2009, the subpolar gyre oscillated between weak and strong phases during the last 6 ka.

Line 267: please change “millennials” to “millennia”.

Line 279: Maybe I am missing something here, but I can't see any Red arrow in Figure 1a.

Lines 280-283: We can’t see this information from Figure 2e!

Line 284: “Enhanced summer heating” when and where?

Line 287: “during this period” please indicate the time-interval here.

Line 290: Please change “In the late Holocene” to “During the late Holocene”.

Line 294: "Consequently, the east-west SST gradient reversed" from what to what? Please explain further.

Line 395: Please change “The late-Holocene summertime cooling” to “The mid- to late Holocene …”.

First all apologies to submit my recommandation so late.

The article ‘Late Holocene cooling and increased zonal asymmetry in the mid-latitude North Atlantic’ by Si, Herbert and Toggweiler present new alkenone-derived SST and concentrations from the North Atlantic and collate them with other records for the region to decipher how the present-day sub polar frontal zone has evolved during the Holocene. The aim is to help understanding better the global Holocene conundrum with a focus on the mid-latitudes North Atlantic, with a special emphasis on a recent hypothesis that states that precession could directly trigger warming and coolings in the tropics (Toggweiler and Si, 2025).

The article reads well and, in my opinion, could be published after minor to moderate revisions. I am listing these suggestions in the order of appearance along the article.

Line 61: should read ‘SW-NE’ tilt I guess?

Chapter 3: The productivity-weighted SST calibration is a very interesting approach. There are already efforts to accommodate regional calibrations in the North Atlantic (Filippova et al., 2016) and productivity-weighted (Schneider et al., 2010) that perhaps should be discussed, along with the limitation that any past changes in regional paleoceanography would inevitably alter the SST estimation based on the Uk’37 but, at least for the North Atlantic, I find it quite convincing. Since you suggest a similar approach that in Schneider et al., 2010 for their productivity-weighted seasonal signal, but have different results in the southern ocean (opposite signs), please explain a little more with appropriate equations (line 95) how was computed your map in the supplementary information.

Results: You should show the U1308 data at least once, even if not all the Holocene has been measured, as it is, as I understand, a new set of unpublished data. Only a gradient testifies that those data actually exist in Figure 4, but the full U1308 Holocene SST record should at least be shown also in Figure 2, aside that of the one derived from site U1304. Also, I would recommend plotting the age control points directly on the two new sediment records you publish at the bottom of the timescale (e.g. triangles), so the reader could easily evaluate the quality of the age model. Finally, sometime around 6 ka you have a step-like decrease in SST that is poorly commented. Is this change seen in the other record? It probably means something interesting, such as a SST front movement around core location(s). Please be less eager to dismiss the structure of your SST timeseries and invest more in commenting your results before rushing on a broader interpretation.

In Figure 5 you show the early/mid-Holocene result extracted from the TRACE transient run. It would be useful to compare it with the PMIP ensemble-mean multi-model output for the mid-Holocene since all models could capture those frontal zones quite differently. A nice example that you did not discuss, too, is the model output from the MITgcm model that was used by Sachs, 2007, Geophysical Research Letters. In fact, the striking difference in the magnitude of the Holocene cooling btw the East and Western part of the Atlantic are all driven by an extreme coolin recorded in the 3 cores published in Sachs, 2007. These cores are apparently strongly influenced by the slope waters of the Northwest Atlantic region, that are rightfully captured by the MITgcm model seen in the publication by Sachs. 20 years later, models have evolved so much that they could help documenting what is happening in this NW corner of Nova Scotia probably much better than the MITgcm seen in the Sachs publication. This is another point that should be scrutinized in more details, too, in my opinion

References:

Filippova, A., M. Kienast, M. Frank, and R. R. Schneider (2016), Alkenone paleothermometry in the North Atlantic: A review and synthesis of surface sediment data and calibrations, Geochem. Geophys. Geosyst., 17, 1370–1382, doi:10.1002/2015GC006106.

Sachs, J. P. (2007), Cooling of Northwest Atlantic slope waters during the Holocene, Geophys. Res. Lett., 34, L03609, doi:10.1029/2006GL028495.

Schneider, B., G. Leduc, and W. Park (2010), Disentangling seasonal signals in Holocene climate trends by satellite‐model‐proxy integration, Paleoceanography, 25, PA4217, doi:10.1029/2009PA001893.