This review addresses the revised manuscript by Lagorio and co-authors, based on their responses to my previous review and the updated manuscript. I acknowledge that the manuscript has improved compared to its initial version. However, while the authors recognize the complexity of the climate dynamics they describe, they still tend to oversimplify some interpretations without adequately acknowledging the limitations of their approach.
In conclusion, I have three main concerns (along with additional comments) that needs to be addressed before the manuscript can be accepted for publication.
1. My first concern relates to the authors’ response regarding the discrepancies observed among the different ENSO proxies used. I fully understand that proxies have inherent limitations, and I agree with the authors that they only discuss periods where all three proxies present a consistent picture. However, this methodological choice is not explicitly discussed in the main text. I strongly suggest that the authors briefly acknowledge the limitations of the proxies used and clarify that they focus their discussion on periods where all three proxies align in terms of the link between the Ross Sea Dipole and ENSO.
2. My second concern pertains to certain statements that should be revised or tempered, particularly regarding the links between ENSO and the Ross Sea Dipole. While I find the discussion for the period 1000–1450 CE convincing, the limitations of this connection in other parts of the record, such as during the El Niño analogue at 300 CE, should be more explicitly addressed.
I anticipate the authors in saying that this is the fourth time I raise this issue. However, my concerns from previous review remain insufficiently addressed. Given that the study presents a 2000-year record—not just a 1500-year one—it is essential to discuss the entire dataset and properly address any inconsistencies in the proposed interpretation. Simply stating that "the first part of the record is much more complicated to understand" is not a valid reason to overlook it or interpret it in a superficial manner without supporting references (see specific comment below). This also contradicts the statement in Line 368: "to accurately interpret our 2kyr record." The authors have not, in fact, accurately interpreted the entire 2000-year record, but rather only three-quarters of it.
Specifically regarding my concern: reading the revised paragraph it seems that the authors still underline that at 300 CE a similar El Nino event was in place, and consequently a similar Ross Sea dipole to that of 1000–1450 CE was still in place. However, δD is not the only proxy that does not respond as expected. As also highlighted in my previous review (but ignored), ssNa from TALDICE shows a minimum during this period, indicating that ENSO did not seem to induce a Ross Sea Dipole: both the eastern and western Ross Sea regions exhibit minimum sea ice extent. This suggests that the ENSO-Ross Sea Dipole relationship may not be as straightforward as proposed or that the proxies used for this comparison have limitations. For example, the ssNa proxy is typically averaged over several years to capture long-term sea ice changes (see Crosta et al., 2022) and may not be reliable for short-term events such as the one observed at 300 CE. While it performs well for sustained El Niño periods (e.g., 1000–1450 CE), its applicability for shorter events may be uncertain. Acknowledging this, or considering alternative explanations, is crucial—especially in light of the manuscript’s conclusion: "El Niño-dominating conditions promoted the establishment of the Ross Sea Dipole" (L539). Well, the available data do not fully support such a definitive claim.
3. Finally, in their reply, the authors stated that they would shorten the initial paragraph on the different climate periods, but they did not. Unfortunately, this is not the only instance where they claimed they would make changes but failed to do so. While it is entirely acceptable to disagree with a referee's comments and provide well-reasoned responses, I find it somewhat disrespectful to deliberately disregard certain suggestions (although answering “OK”), especially when they were made solely to enhance the quality of the manuscript.
_____
OTHER COMMENTS (line numbers refer to the manuscript version without track changes)
L26: Consider providing a brief definition of the Ross Sea Dipole here, as its introduction feels somewhat abrupt. You might condense the explanation from L171–173 to ensure clarity. I raised this point in my previous review as well.
L66: Since the manuscript presents three questions, a “(3)” should be added after “LIA?” for consistency.
L128: "Over the last 2 ka" should be written without "BP," as "over the last" already implies a time span relative to the present.
L218: Since the mean is a specific value, please provide the actual number (e.g., 16.x) or a range.
L218: If the dust record over the Holocene is not shown in this study, the phrase “this study” should be reconsidered. As currently written, it suggests that the temporal dust record is presented, while only an average value is reported. If the authors do not want to disclose the value, they should either refer to other similar sites, or remove this reference to the Holocene as it is a sentence that cannot be verified (the data are indeed not shown).
L220: The average snow accumulation rate should be given with the same number of significant digits as in L76 for consistency.
L235: Does "Supplementary Information" here refer specifically to Figure SF3? Clarifying this would be helpful.
L237: Perhaps my previous comment was unclear. I understand that the period 550–1450 was identified based on the decadally smoothed profiles of dust concentration and flux. However, my question is: How were these periods determined? Was an abrupt change-point detection method applied to the smoothed dust profiles, or was this classification made qualitatively? You mention that “a detailed comparison between dust and stable isotopes is limited by the different sampling resolution of the two records.” However, it would be interesting to see if applying the same approach used by Bertler et al. yields similar results in your dataset. If the authors choose not to use such a method, they should explicitly clarify that the step changes in the dust profiles were identified qualitatively. The validity of this classification is nonetheless supported by the boxplot in the Supplementary Material.
L246: Shouldn’t the periods be three?
L256: In their response to my previous review, the authors stated that they are not yet prepared to present the Holocene diatom record, as they are still working on it. Since these data remain unpublished and cannot be verified, the reference to “even over the Holocene” should be removed.
L263–L267: Since the comparison between SEM and optical microscopy is a methodological outcome, it belongs in the Methods section. Please move this paragraph after L211.
L265: Add "6" after "SF" for clarity.
L345: Why did the authors not calculate the average ssNa values for the two periods (800–1000 CE vs. 1000–1300 CE) and compare them? Simply stating that “it increased” without providing numerical values is open to interpretation, particularly when the change is not so immediate. The authors should calculate the mean for the period before the increase (e.g., 500–1000 CE) and compare it with the mean from 1000–1300 CE, then test the differences for statistical significance. A box plot similar to what was done with the dust concentration would be appreciated. The authors have the data, so this can be easily achieved, and this will help in giving a stronger statistical support to the discussion.
L410: The discussion regarding the lack of response in the RICE stable isotope record is appreciated. However, the following statements require references: “surface waters of the Eastern Ross Sea were much cooler” and “300 CE marks a period of dynamic changes in the Ross Ice Shelf, with the calving line either terminating or having just completed its last phase of retreat.”
L423–L426: The authors cite Figure 6 in Brightley (2017), but this figure does not exist in the version available at this link: https://openaccess.wgtn.ac.nz/. The last available figure is 5.7. Is there another version? If so, please provide the correct reference. Alternatively, are the authors referring to Figure 4.1?
Additionally, the phrase “significant increase of marine compounds” should be clarified. First the authors should know that Na+, Ca2+, K+, and Mg2+ are ions, not compounds!! I suggest the use of "marine tracers", or "marine species". Second, the comparison needs to be more specific: significant increase relative to which period? Based on Figure 4.1, the comparison appears to be with 1200–1400 CE, but this should be explicitly stated.
FIGURES & FORMATTING
Figure 1: A compass rose would be helpful
Figure 2: Panel (c) already includes FPP, dust concentration, and Loess, correct? If so, panel (d) seems redundant and could either be omitted or clearly separated. Please also specify the color coding in the caption. To enhance readability, the Loess 20-year running mean should be shown in a distinct color (e.g., red). Additionally, the arrow in panel (e) should be better aligned and centered with the panel it refers to.
Figure 5: Since the unit for concentration is provided for dust, it should also be reported for ssNa, RICE snow accumulation rate, and Fe concentrations for consistency.
Figure 6: The units should be consistent with those in other figures—either use ppb or µg/kg, but maintain uniformity throughout the manuscript.
GENERAL FORMATTING & DATA AVAILABILITY
Double spaces: There are multiple instances of double spaces throughout the text. Please review and correct.
Abbreviations: Both ka and kyr appear in the manuscript. Please choose one and apply it consistently.
Data availability: In accordance with Open Access principles (FAIR) and the journal’s data policy, I recommend uploading the data from Tables S1 and S2 to an open-access repository.
SUPPLEMENTARY MATERIAL
Figure SF1: The caption is unclear. Does this figure present dust fluxes for particles smaller than 5 µm? If so, please specify this in the caption. Also, clarify the meaning of the elements in the figure: “Mean dust fluxes (bars), Late Holocene dust fluxes (circles), and elevation of each drilling site (crosses).”
Figure SF3: Please present the p-values using scientific notation and calculate it also for 0-500. If they want, the authors can also add a small diagonal 3x3 matrix reporting the p-values for the three periods. |
