GENERAL COMMENTS:

The article by Thomas et al. presents a collection of ice core chemistry data from 105 uniformly distributed sites in Antarctica dating back to the last 2,000 years. As part of the CLIVASH2k project, which aims to reconstruct climate variability in Antarctica and the Southern Hemisphere over the past 2,000 years, all previously published and unpublished sodium [Na⁺] and sulphate [SO₄^2-] records were requested and collected by the authors through a specific call addressed to the paleoclimate community. When the snow accumulation rate for a given site was available, the sodium and sulphate concentration was converted to deposition flux. In addition, based on the assumption that all Na came from the ocean, excess sulphate or xs [SO₄^2-] and its flux was also calculated. Finally, based on statistically significant correlation with three climate variables (sea ice concentration, atmospheric circulation, and wind), the data were filtered with the aim of identifying potential sites suitable for future proxy reconstructions.

This paper represents the synthesis of a major community effort to create a useful and comprehensive dataset derived from hundreds of snowpits, and ice and firn cores retrieved over the past decades across Antarctica. Although some minor aspects listed below need to be clarified before publication, the article itself is adequate to support the publication of a dataset, and the overall structure of the article is well structured and clear to the reader. The annual resolution data are likely to be useful for future studies, thus I support their publication. I particularly appreciated the so-called "first-pass filtering" of the data provided, which makes it easier to identify the potential of a given ice core site for past environmental reconstructions and/or comparisons between data and models.

SPECIFIC COMMENTS:

-Line 80: Please rephrase the sentence to avoid the repetition of the word “contain”.  

-Line 81: Specify which are the climate observation/variable used to filter the dataset.

-Line 96: Consider adding the words in capital for clarity:" value of large datasets in reconstructing climate and sea ice VARIABILITY over decadal to centennial TIME scales”

-Line 100: Please fix the reference format for the “Wais divide project members”.

-Lines 105-106: Add additional Ref. to Minikin et al., 1994. A more recent study could be “Rhodes, R.H., Yang, X., Wolff, E.W., 2018. Sea ice versus storms: what controls sea salt in arctic ice cores? Geophys. Res. Lett. 45, 5572e5580.”

-Line 114: In my opinion the annotation “xs” used to define the excess of [SO₄^2-] is a bit confusing through the text. Please consider to change it with the annotation [SO₄^2-]_Exc  according also to other ice-core based studies.

-Lines 132-134: The sentence “… 2) the data comprise two distinct chemical species which do not have a well-established relationship with climate (beyond the episodic sources of [SO4^2-] noted above)" is not clear, please rephrase it for clarity.

-Line 145: What do you mean for “… providing a threshold of half a year of data was achieved”, please rephrase it for clarity.

-Lines 156-158: Change “two” with “few” and “the third” with “another”.

-Lines 173-174. This statement is not entirely correct because is not enough to have an ICP-MS to see the total concentration. Once the sample is acidified, you can see part of the total, or the dissolved particles. In this regard, it is important to mention that there are different protocols for acidifying samples by making use of different acids (HCl, HNO3, HF, etc.) and different acidification times (days, weeks, months) based on which different absolute concentrations might be obtained as illustrated by Gaspari et al. (2006), Koffman et al. (2014), and Burgay et al. (2021).

-Lines 180-182: Insert Ref. to both sentences.

-Line 196: Insert the equation used to calculate flux from concentration with the units presented in the dataset (i.e. ppb kg m-2). It would also be appropriate to explain why these units were chosen instead of the more common (mg m-2) or (ug m-2). In this way, indeed, a user can easily infer the annual snow accumulation (in kg/m2 or mm w.e.) simply by dividing the flux with the concentration.

-Lines 210: Add a Ref. to the sentence “Other ratios may be more suitable for coastal sites (Ref.), …”.

-Lines 241-244: This part is a bit misleading. You start the paragraph saying, “A total of 117 records were submitted, representing 105 individual ice core sites (Fig. 1)”, however, looking at Table 1, the Total records are only 105 and the Total ice core sites are 90 with 15 duplicates. Where are listed the remaining 12 records? Furthermore, please change “Table A1” with “Table S1” as indicated in Appendix A.

-Line 260 (Figure 1 / Panels A-B): Since all the records, except for one, are homogeneously distributed across all the continent it would be worth zooming in a little bit more and create a separate small box only for Bouvet Island.

(Figure 1 / Panels C-D): Because of the low resolution of the graphs, also after zooming in, it’s still very hard to distinguish each single record based on its duration (the file looks like formatted as a jpg or jpeg file). Besides that, the right part of the graphs (i.e., around Year CE 2000) appears oblique going from the top to the bottom of the graph. Is it due to some indistinguishable records present from 2000 CE to the end of the graph? Please submit a new version of Panels C-D increasing the resolution (i.e., uploading the original vector file) for clarity.    

-Lines 334-335: In the sentence “Proportionally, more records are correlated with SIC when using flux THAT (than?) concentration (78 % compared to 72 %)” percentages are not corrected. Please adjust them according to the values listed in Table 3 (if I understand correctly, it should be 86% compared to 78%).

-Line 338: The additional records marked as uncertain in Table 3 are 3 and not 4 as stated in the sentence. Fix the table or the sentence with the correct value.

-Line 349 (Figure 3): Similarly to what was suggested for Figure 1, in my opinion the panels shown in Fig. 3 would also deserve a zoom to better visualize the individual ice core sites and their spatial distribution across Antarctica (with a separate small box for Bouvet Island). Also consider changing the color of both the fill and the edges of the diamonds representing sites with no correlation; gray might be difficult to distinguish from sites with uncertain and/or significant correlation.   

-Lines 361-362: The records of SO4 indicated in Table 3 are 40 and not 39 as stated in the sentence. Also, the percentage need to be fixed accordingly (they seem to be reversed in the sentence). Fix the table or the sentence with the correct value. Same in Line 369, “26 records (out of 59)” should be out of 61, as indicated in Table 3. Fix the table or the sentence with the correct value.

-Line 371: Flux should be concentration or, alternatively, the percentages must be inverted, and the sentence adjusted accordingly.

-Lines 373-396 (Figure 4 and 5): Same comments as for Figure 3.

-Line 407: For consistency, add the ionic form for [Na⁺] (also in lines 411,431, and 473).

-Line 408: I would replace “thus” with “this suggest that”.

-Line 410: For an easier identification of the “most useful” sites for climatic studies, it would be nice if the full list of sites presented in Table S2 would be ordered from top to bottom according to the number of correlations with the others climate variables. In this way, sites with a larger number of correlations (i.e. “ITASE-02-6”) would be listed in top rows of Table S2, while site with a lower number of correlations (i.e. WHG (ICPMS)) would fill the last rows.

-Lines 419-420: Replace “In East Antarctica” with “In contrast, in East Antarctica, ”

-Line 425: Replace “to covert to” with “needed to calculate the flux”.

-Lines 431-432: Please rephrase this period for clarity. I suggest something like: “Overall, the records of [Na⁺] exhibit the highest number of correlations with the climatic variables considered (179 out of 264) , followed by xs [SO₄^2-] (or [SO₄^2-]_Exc  )(164 out of 243), and ([SO₄^2-] (152 out of 252).”   

-Line 474: Rephrase the sentence about Na flux in a more conservative way, such as “The Na flux exhibits the greatest proportion of records that correlate with sea ice, atmospheric circulation, and winds. Therefore, among the ice-core chemical species considered in our analysis, we propose Na+ flux as the best candidate for reconstructing all three climatic components”.

General comment

Thomas et al present a unique data base of previously published and unpublished sodium [Na⁺] and sulphate [SO₄^2-] records from all over Antarctica, collected within the CLIVASH2k project. The records span several 100 years back to 2000 years. Age scales were made consistent where possible and age transfer functions are provided. A data validation is provided and recommendations with respect to uncertainty. In a second step, a comparison to the climatological parameter sea ice condition, wind and geopotential height is conducted to serve as a first level filter for the data base. To this end the correlation of each record to the climatological parameter is evaluated.

The authors present the data set along with a careful description of uncertainty and potential usage. The provided data base presents a well needed contribution to data coverage related to Antarctic climate. The only weakness of the paper is the attempted interpretation, i.e. the presentation of the correlation analysis of single records.

Specific comments:

Line 147ff: It would be beneficial to have the records published with both –age and depth. Having access to the original record over depths allows re-using the record and comparing it to new/upcoming records in the future.

Line 156: how are these cores dated/synchronized with the others?

In line 265ff (chapter 3.1.1) the correlation of single records with the three climatic parameter sea ice concentration, wind and geopotential height is used for validation. It has been shown for stable water isotope records from Antarctic ice cores, that single records contain little climatic information on annual time scales, i.e. that it takes several cores to increase the meaningful time period to interpret (Münch and Laepple, 2018). The authors recommend to average over several cores earlier in the text (line229) but interpret their single records with respect to climatic parameter. Assuming that sodium is deposited with the snow (like stable water isotopes) these findings (and the approaches to solve the issue) could be considered similar – i.e. that it makes sense to look and larger time scales than annual for records from the East Antarctic Plateau.

Line 277: It is not explained well, why the correlation analysis includes wind/ geopotential height. What do you expect and why? Is your expectation the same everywhere in Antarctica?  

Line 288ff: The authors write, that an “interpretation team” is evaluating the results. The objective of this approach is to provide a first level filter for the database. It implies, that the expert’s opinion is counted very high, whereas the statistical evidence is not relevant. On what basis did the experts decide? Has there been an objective measure? Is this finding reproducible by others? Maybe it is possible to mention some of the measures taken by the interpretation team here.

Line 308ff: In previous studies it is shown, that climatic fields inherit patterns and correlations which lead to p>0.05 probabilities by chance (see Livezey, R. E., & Chen, W. Y., 1983). The presented results have to be taken with care – i.e. to be redone accordingly and/or require a more in-depth discussion.

Overall, the interpretation is the weakest part. Maybe it makes more sense to work on the interpretation in more detail in an extra paper and leave it out here.

 References:

Münch, T. and Laepple, T.: What climate signal is contained in decadal- to centennial-scale isotope variations from Antarctic ice cores?, Clim. Past, 14, 2053–2070, https://doi.org/10.5194/cp-14-2053-2018, 2018.

Livezey, R. E., & Chen, W. Y. (1983). Statistical Field Significance and its Determination by Monte Carlo Techniques. Monthly Weather Review, 111(1), 46–59. https://doi.org/10.1175/1520-0493(1983)111<0046:SFSAID>2.0.CO;2