Stable isotopes of nitrate preserved in ice cores hold the potential to reveal past variability of stratospheric ozone over Antarctica. However, there are many factors affecting ice core nitrate concentration as well as its stable isotopic composition. Efforts to understand those processes and their influence are therefore much needed. 

In this manuscript, Cao et al. presents such an effort using two shallow ice cores from the South Pole dating from 1944 to 2005. Because the time span of the cores nicely encompasses the period of the Antarctic ozone hole since 1976, the nitrate isotope records within serve as a nice archive to investigate the relative contribution of different factors on the nitrate isotopes. Observationally, the authors find that the d15N of nitrate has large variability and the D17O of nitrate displays a long-term decline (on top of the variability). Aided by a snow photochemical model, they conclude that:

(1) Ozone hole—which enhances UV flux arrived at the ice sheet surface—alone cannot account for the large variability of d15N, so accumulation rates must be the dominant factor here.

(2) Nonetheless, if snow accumulation rates are somewhat stable, the variability could potentially reflect post-depositional processes driven by UV—and by extension by ozone variability.

(3) Finally, the trend in D17O seems to be compatible with a change of atmospheric oxidant ratios in the extratropical southern hemisphere.

Overall, this paper is timely and interesting, and falls within the scope of Atmospheric Chemistry and Physics. It is well-written and easy to follow. I enjoy reading it and believe it could be published on ACP after making some minor revisions and adding some clarifying statements. I should say, however, that the photochemical modeling is out of my area of expertise, so I am may not be qualified to assess the robustness of the model. I hope other reviewers could comment on the modeling aspect more authoritatively.

General comments:

First, in the Introduction (from Line 51 and onward) there seems to be no mention of other attempts to reconstruct ozone and the authors proceed to discuss the principles of stable isotopes of nitrate preserved in ice cores as a potential ozone proxy. Non-ozone specialists may wonder if there are other ways to know ozone in the past. A quick review of the existing methods with their strengths and limitations discussed could be helpful here. The readers will also be able to understand the value of the isotope records in ice-core nitrate.

Second, in the Discussion 4.3, the lines of reasoning could benefit from a simple re-structure: why not putting the 4.3.2 and 4.3.3 first? This way you could discuss the reject the alternative hypotheses, leaving the most plausible explanation (changes in the O3/HOx ratio) on the table.

Third, one key point of the paper is that Dome A might be a good place to study nitrate isotopes because of its low snow accumulation rates. This might foreshadow a follow-up study from that very site, which is great. For the present study, however, can you also calculate the expected d15N variability induced by stratospheric ozone in other East Antarctic sites such as Vostok, Dome C, and Dome F where deep ice cores have been drilled? This could be summarized with a new figure. Though it does not necessarily mean that you have those samples, I think this exercise could benefit the ice core communities in general.

Specific comments:

Line 21: “but” and “nevertheless” are repetitive. “Nevertheless, this enrichment is small and masked by …” sounds better.

Line 21: the second half of this line could be simplified by saying “… masked by the effects of snow accumulation rates at the South Pole ...” In essence the snow accumulation rates have two parts: internal variability superimposed on a long-term trend. They could be discussed in greater detail in the main text without complicating the message here in the abstract.

Line 32: consider changing “protecting life on land” into “and protects life on land”. No need for using the nonfinite verb here.

Line 44: missing an “of” after “shifting”.

Line 57: “ozone which determines surface UV radiation.” This seems to suggest that there are lots of “ozone” and the one being talked about is the one that determines surface UV radiation. Yet, in fact you are just describing stratospheric ozone, so no need for the defining relative clause here, and there should be a comma “,” before “which”.

Line 61: missing an “as” after “deposited”.

Line 78: this sentence is not very clear. By saying “it is a mass-independent fractionation signal” it is implied that photolysis is a mass-dependent process. If this is the case, please explicit state so.

Line 129: can you specify which years were binned to the adjacent samples? This could be provided as a supplementary table.

Line 190: “were from data extrapolation” could be better phrased as “were extrapolated”.

Line 216: missing a blank between “years” and “1944”.

Line 218, 231, 237, 245, and 251: please specify the meaning of the number after the ± sign. Is it one standard deviation?

Line 263: there are two “similar to the observation”. Please consider rephrasing.

Line 278: change “pronounced” to “reproduced”?

Line 283: I would appreciate you putting the numbers into a greater perspective here. At face values, about 75% of the primary nitrate was lost, leaving 25% nitrate behind. On the other hand, you mentioned that re-deposited nitrate contributed to the preserved nitrate. Does this mean that the loss of *primary* nitrate exceeds 75%? Similarly, please specify what the ~40% nitrate loss calculated by the photochemical model refers to, perhaps with the help of Figure 2: is the nitrate in the combined surface and photic layer?

Line 305: the shading area in Figure 4 does not correspond to the periods with an ozone hole.

Line 307: is this from the sensitivity test? If so Figure 5 should be mentioned. Alternatively, you could just discuss d15N of nitrate exclusively in section 4.2 (which now needs a new title of course), and leave the discussion of D17O entirely to the next section.

Line 364: “discern” might not be the proper word choice here. “Investigate” or “Examine” sounds more logical.

Line 455: “Had snow accumulation rate at Dome A stayed …” Technically this sentence shouldn’t be in subjunctive mood, because by doing so you are implying that accumulation rates at Dome A were, in fact, not stable. Yet, the accumulation rate history is not known, so you could just use “If” instead of “Had” to indicate a possibility.

Figure 2: should be “Archived” instead of “Achieved” layer?

Figure 3: per the text, the “ozone hole period” begins right after 1976, but in the figure here, the ozone hole starts around 1979 C.E.? Please make them consistent with each other.

Figure 4: please add some visual guidance to mark the ozone hole period.

Figure 5: same as Figure 4, a little visual cue of the ozone hold period (or simply the beginning of it) would be nice.

Cao et al. measure nitrate isotopes and concentrations in a 60-year firn core from South Pole and perform air-snow nitrate transfer simulations using the TRANSITS model to investigate whether nitrate isotopes at the site reflect changes in stratospheric ozone. The results are similar to previous Antarctic studies of ice core with similar snow accumulation rates that indicate d15N(NO3-) is insensitive to total column ozone. Decreases in the D17O(NO3-) record during the ozone hole are qualitatively attributed to atmospheric oxidization changes in the extratropical Southern Hemisphere nitrate source regions. The new dataset is a valuable contribution however, the manuscript could be improved by furthering our understanding of ice core nitrate isotopes in Antarctica which have a unique and not fully understood fingerprint. As such, I believe the authors have an opportunity to use the ice core dataset and the TRANSITS model to advance our understanding of ice core D17O(NO3-) to make a new and valuable contribution to the literature. I look forward to seeing the published. 

Suggestions for improvement

A paper on nitrate isotopes in a snow pit (1960-2000) from the low-accumulation Dome A site was just published in June (Shi et al., 2022) and the authors conclude that nitrate isotopes (d18O, D17O, and d15N) record stratospheric ozone depletion and ultra-violet radiation at the Dome A site. The authors have discussed the modelled response of d15N(NO3-) to total column ozone at South Pole versus Dome A sites. Please update the manuscript in light of the newly published paper.

Shi, G., Hu, Y., Ma, H., Jiang, S., Chen, Z., Hu, Z., et al. (2022). Snow nitrate isotopes in central Antarctica record the prolonged period of stratospheric ozone depletion from ∼1960 to 2000. Geophysical Research Letters, 49, e2022GL098986. https://doi.org/10.1029/2022GL098986.

Now that there are a number of d15N(NO3-) measurements across Antarctica, a discussion on the sensitivity of d15N(NO3-) and D17O(NO3-) to total column ozone at various ice cores sites, including the new Dome A record, would be valuable addition for the community to make progress on the use of d15N(NO3-) and D17O(NO3-) as a UV or total column ozone proxy.

Another recently published study (July 2022) on nitrate isotopes in relatively high accumulation rate sites (Summit Greenland) also highlights the importance of understanding post-depositional effects of ice core nitrate and it would be worth citing this paper.

Jiang, Z., Savarino, J., Alexander, B., Erbland, J., Jaffrezo, J.-L., and Geng, L.: Impacts of post-depositional processing on nitrate isotopes in the snow and the overlying atmosphere at Summit, Greenland, The Cryosphere, 16, 2709–2724, https://doi.org/10.5194/tc-16-2709-2022, 2022.

There are extremely scare measurements of e-folding depth in Antarctica. A much shallower e-folding depth of 2-5 cm was observed at DML. This was also shallower than estimated by Zatko et al. (2013). What is the uncertainty on your estimated e-folding depth of 20 cm? How appropriate is that estimate in the context of measurements and modelled estimates? Given that recent studies have shown the importance of e-folding depth on nitrate recycling, a discussion and sensitivity analysis of a range of possible e-folding depths for South Pole site is highly encouraged.

Please add a section of assessing the validity of the TRANSITS model especially in regards to D17O(NO3-). The model doesn’t simulate the observed decreasing D17O(NO3-) trend from ~1976 to 2000. Why is this? How much can you take away from the simulated D17O(NO3-) results? How can you improve the model? How does the model help you understand D17O(NO3-) at South Pole. TRANSITS simulations of D17O(NO3-) would be an area where the authors can contribute new understanding to the literature.

Introducing the South Pole site in terms of the snow accumulation and also atmospheric nitrate isotopes (Walters et al., 2019) in the introduction would be helpful to put the site into context of other records given that the nitrate isotopes are sensitive to accumulation rate.

It is not always clear in the discussion if the authors are talking about the results from TRANSITS or observations.

Specific comments

L1 The title is misleading as nitrate isotopes at South Pole do not reflect changes stratospheric ozone changes.  

L26 HCl and ClONO2

L65-67 The photic zone at DML is 15 cm which is less than Dome C (Winton et al., 2020).

L71-73 This sentence focusses on fractionation constants on the EAP. Relevant to this study are fractionation constants in the “transition zone” characterized by snow accumulation rates typical of sites located between the EAP and coast (5–20 cm yr−1 w.e.; Erbland et al. 2015).

L86-102 Recent studies have shown the importance of e-folding depth on nitrate recycling. This is important to mention here.

L108-111 See the recently published paper by Shi et al. (2022)

L131 Did you decontaminate the samples?

L134 Suggest moving reference to Geng et al. further up in the methods section.

L119-136 Please add protocols for minimising contamination. Please state the sample resolution in terms of depth and age here.

L133 UW

L138-140 This sentence seems out of place.

L164 How did you calculate the e-folding depth?

L223-226 Seems out of place.

L234 Add the dates of the pit

L282 Can you use the approach of Weller et al. (2004) to calculate nitrate loss? And then compare to the TRANSITS estimate of nitrate loss?

L295 Heading should reflect that this section is about TRANSITS modelling

L334 This assumption ignores other factors that influence e-folding depth. While we don’t know how e-folding depth changes over time, based on changes in grain size, snow density and impurity content it is fair to assume e-folding depth at any site is not constant through time. Sensitivity studies show that nitrate isotopes are sensitive to changes in e-folding depth.

L346-357 Update in light of the published work by Shi et al. (2022).

L359 The ice core data in the figures suggest interannual variability.

L391 A concluding sentence about oxidation for this paragraph would be helpful here.

L424 The EAST ANTARCTIC PLATEAU snow sourced

Figures: It would be very helpful for the reader to visualise the TCO and nitrate isotope trends on the same figure.  

Fig. 5: please add in the nitrate isotope observations.