Review of “A seasonal analysis of aerosol NO3- sources and NOx oxidation pathways in the Southern Ocean marine boundary layer” by Burger et al.

In this manuscript, the authors provide a unique dataset of isotopic compositions (d15N, d18O, and D17O) of atmospheric nitrate across the Atlantic Southern Ocean, including spring and winter samples, in addition to summer. Combined with air-mass back trajectory (AMBT) analysis, they interpreted the data to characterize each season with different nitrogen sources and oxidation pathways for nitrate formation. They conclude that relatively low d15N in spring and summer suggests a significant contribution of alkyl nitrate emitted from the ocean surface, in addition to well the established source of snowpack emission. A three isotope plot of D17O and d18O indicates the contribution of peroxy radicals (RO2 and/or HO2) to NOx oxidation in springtime, which may be enhanced by alkyl nitrate or chlorine emissions from sea ice.

Their findings are undoubtedly beneficial in evaluating modeling works regarding nitrogen and oxidants chemistry in the Southern Ocean marine boundary layer, which is of importance for constraining the background conditions of aerosol radiative forcing as the authors mentioned in the introduction. Their conclusion is well supported by the data with clear presentation, except for a few concerns regarding the influence of field blanks on their isotopic measurement and lack of discussion for limitation of their isotopic approach when considering possibility of isotopic fractionation on both d15N and d18O. I find that the manuscript is worthy and appropriate for publication in ACP, after addressing minor revisions stated below.

Specific comments:

Line 18: Here in the abstract authors mention that the major NOx source in low latitude region is lightning based on d15N, whereas in line 299-305 they conclude that the possible NOx source includes other natural sources such as biomass burning and soil emission. Biomass burning and soil microbes can supply NOx with d15N of -7 to 12 ‰ (Fibiger and Hastings, 2016) and -60 to -14 ‰ (Miller et al., 2018), respectively. Thus it would be difficult to rule them out from possible sources based solely on d15N. The abstract should be corrected to be consistent with discussion.

Line 94-95: Theoretical mechanistic of non-mass dependent isotope signature in ozone is thought to originate from the stabilization step of asymmetric molecules of excited ozone (O3*), as mentioned in Ireland et al. (2020) and initially proposed by Heidenreich and Thiemens (1986). It is not believed to be associated with photochemistry. Please revise the explanation accordingly.

Line 103: “a lack of exchange of O atoms with O3” is not correct expression because formation of nitrate is not equilibrium reaction, unlike the case of isotopic exchange between H2O and OH. I suggest rephrasing to “increased contribution from other oxidants”.

Line 166: The authors report that field blanks represented 32% and 59% of the NO3- on sample filters in winter and spring respectively. While they corrected concentration measurements for these field blanks, there is no mention about corrections of isotopic measurements. To mitigate the potential impact of blanks, it is common to collect an excessive amount of nitrate relative to the blank on each filter, or to measure isotopic compositions of the blanks to correct those of samples (e.g., Savarino et al., 2007). But in this manuscript, authors conducted ~24 hours sampling to obtain higher temporal resolution, which resulted in small nitrate loadings on the filters. What is source of the field blanks, and would it be possible to discount the significant impact on isotopic signatures? Is it possible to assume isotopic signatures of the blanks? Even if not, the potential impact and the assumption made for the later interpretation should be carefully addressed.

Line 200: Is 72-hour AMBT enough to trace NOx source? I would expect any references or discussion to certify the lifetime of nitrate for 72-hours.

Line 256: I did not see any discussion about potential influence of isotopic fractionation on d15N through NOx oxidation, even though authors points to it in the introduction (Line 81-83). To appropriately convey the limitations of their approach appropriately, I believe the authors need to address how they assumed minimal or negligible influence by isotopic fractionation in their interpretation.

Line 391: What is meant by “prior work” here? I perceive that the authors are assuming that the influence of isotopic fractionation on d18O through NOx oxidation is insignificant, and therefore d18O-NO3- directly reflects d18O of oxidants. This is an important point to interpret D17O vs d18O plots later. I would suggest including references or discussion to support this assumption.

Line 416 and Figure 6: I am confused by the apparent inconsistency in the isotopic composition of ozone as reported in the text and plotted in the figure. While the text states “a δ18O of ~114 to 138‰”, the O3-endmember is plotted around d18O of ~110‰ in Figure 6. I am unsure of how the authors employed the lowest value of possible variation in d18O of ozone. In case of D17O, ozone end-member is estimated to be 37 to 39‰ based on experimental studies determining transferring factors of O3-term to products for NO + O3 (Savarino et al., 2008) and NO2 + O3 (Berhanu et al., 2012). Would it be possible to do similar calculation for d18O? Or, is there another reason for d18O of ~110‰ in Figure 6?

Technical comments:

I would suggest combining Figures 2, 3, and 5 so readers can refer to information of concentration, d15N, and d18O of each sample at once.

Line 22, 261, and 488: The term “stratospheric NOx” sounds not proper. Nitrogen is transported from the stratosphere in the form of nitrate or nitric acid, not NOx. I would suggest rephrasing as “stratospheric nitrate”.

Line 68: “from” -> “form”.

Line 142: I presume “exceed 0 m s-1” is “exceed 1 m s-1”.

Line 156-195: Section 2.2.1 Isotopic analysis includes description of sea water sampling and its nitrite concentration measurement. To improve the construction, I suggest dividing section 2.2 to three small sub-sections as:

Line 157-170 -> 2.2.1 Concentration analysis

Line 171-190 -> 2.2.2 Isotopic analysis

Line 192-195 -> 2.2.3 Sea water sampling and nitrite concentration analysis

Line 168: “(Sect. 2.3)” should be matched with the appropriate section number.

Line 206: A sentence “During…Southen Ocean.” is not necessary.

Line 221: “ice edge transect (d)” and “northbound voyage (e)” may be reversed. I see Figure 1e shows ice edge transect voyage.

Line 335, Figure 4: The ranges of colorbars for d15N-NO3 and [NO2-] are consistent among three panels so one for each parameter is enough. I would suggest leaving three maps as is and placing two colorbars below the panel c.

Line 438: “δ¹⁸O-H₂O_(v) (= -13.6 ± 1.5‰)” is inconsistent with “(-13.9 ± 1.4‰)” in Line 420. Correct either.

Line 444: “d18O” -> “δ¹⁸O”

Citations that were not in the original manuscript:

Fibiger, D. L.; Hastings, M. G. (2016) First Measurements of the Nitrogen Isotopic Composition of NOx from Biomass Burning. Environ. Sci. Technol. 50 (21), 11569–11574. https://doi.org/10.1021/acs.est.6b03510.

Heidenreich JE III, Thiemens MH. (1986) A non-mass-dependent oxygen isotope effect in the production of ozone from molecular oxygen: the role of symmetry in isotope chemistry. J. Chem. Phys. 84:2129–36, https://doi.org/10.1063/1.450373

Miller, D. J.; Chai, J.; Guo, F.; Dell, C. J.; Karsten, H.; Hastings, M. G. (2018) Isotopic Composition of In Situ Soil NOx Emissions in Manure-Fertilized Cropland. Geophysical Research Letters, 45 (21), 12,058-12,066. https://doi.org/10.1029/2018GL079619.

Savarino, J.; Bhattacharya, S. K.; Morin, S.; Baroni, M.; Doussin, J.-F. (2008) The NO+O3 Reaction: A Triple Oxygen Isotope Perspective on the Reaction Dynamics and Atmospheric Implications for the Transfer of the Ozone Isotope Anomaly. J. Chem. Phys. 128 (19), 194303. https://doi.org/10.1063/1.2917581.

GENERAL COMMENTS

The authors present new shipborne measurements during winter and spring of the stable isotopes of nitrogen (d15N) and oxygen (d18O, D17O) isotopes in the coarse mode of atmospheric nitrate collected in the marine boundary layer (MBL) between South Africa and the marginal sea ice zone in Antarctica. d15N values are used to attribute primary sources of atmospheric nitrate: during spring/summer lightning, ocean (alkyl nitrates) and snowpack NOx emissions dominated at low, mid and high latitudes, respectively. During winter transport of NOx precursors such as PAN from lower latitudes as well as potentially stratospheric nitrate contribute mostly to the atmospheric nitrate background.

Using D17O and d18O values in an isotope end member mixing analysis the authors confirm the current understanding that oxidation during daytime is dominated by OH and during night time/ winter by O3. They speculate that a third end member emerging at sunrise in spring may be attributed to the onset of halogen chemistry and contribution to oxidation via peroxy radicals.

These are important new atmospheric data from the Southern Ocean MBL covering seasons which are notoriously under-sampled, and therefore should be published. However there are some weaknesses in data interpretation, some gaps in the cited literature as well as presentation of results can be improved. Major points:

- the introduction should expand on the nitrogen chemistry relevant for the oxygen and nitrogen isotope transfer, i.e. spell out key reactions of the relevant pathways:

Step1) NO,NO2 interconversion (fast) and Step2) NO2 oxidation to form nitrate (slower). This will help the reader to follow the arguments presented and assess key uncertainties and missing variables for future studies aiming at a quantitative isotope budget.

- halogen chemistry in step1) NO,NO2 interconversion and step2) NO2 oxidation to form nitrate with respective implications for the oxygen isotope transfer is currently not considered (Section 3.3) and not included in the oxygen isotope mixing model. However, halogens are important in the MBL particularly near/above sea ice or polar ice caps. There is evidence that halogen chemistry acts as a major NOx sink and source of nitrate via the production and subsequent hydrolysis of XNO3 species as observed in coastal Antarctica in summer (e.g. Bauguitte et all, 2012). Thus increases in D17O (or d18O) in nitrate may reflect increased oxidation by XO during step1 and step2 during daytime (mostly spring), possibly closely linked to local NOx emissions of NOx (e,g, Morin et al., 2012). This is because reaction of halogen radicals X (=Cl,Br,I) with ozone lead to the formation of XO

1) X + O3 --> XO + O2

followed by

2) XO + NO --> X + NO2 (e.g. at 2-3 pptv BrO small impact on D17O in NO2 and NO3-; Savarino, 2016)

3) XO + NO2 + M --> XNO3 + M, XNO3 + H2O --> HNO3 + HOX (efficient transfer of D17O of XO and NO2)

at halogen levels of only a few pptv there is considerable impact on NO/NO2 ratios (e.g. Savarino et al., 2016), NOx lifetime (Bauguitte et all, 2012; Frey et al., 2015) and impact on D17O/d18O in atmospheric nitrate. This needs to be mentioned and included in the discussion on latitudinal gradients of d18O/D17O(NO3-).

- related to the above: negative correlation between d15N and D17O observed in atmospheric nitrate during Arctic spring (Morin et al., 2012) and in inner Antarctica (e.g. Savarino et al., 2016) indicate that snowpack emissions result in enhanced D17O transfer to nitrate. Possible processes include reactions with XO near halogen sources (sea ice, open ocean) or HONO co-emitted with NOx from the snow pack contributing to the local OH budget (e.g. Legrand et al., 2014; Bond et al., 2023). Correlations between the reported D17O(d18O) and d15N especially during spring need to be analysed to discuss the impact of snow emissions and halogens on the isotope transfer. It seems to me that by overlaying Fig. 3 & 5 there is a noteable anti-correlation between d18O and d15N in the spring ice edge measurements.

SPECIFIC COMMENTS

L18 I think you mean "the dominating primary NOx sources"

L22-24 is the threshold  for when you think O3 oxidation dominates 60 or 70 permil? It does not make sense to have two threshold values or you have to explain why they are different in spring vs summer.

L26-27 not only HO2/RO2 but also oxidation by XO (see related comments)

L48 NOx emissions from snow are not considered a primary NO3- source, as this is recycled nitrate from atmospheric deposition (oceanic and lower latitude sources) and no3- produced in snow or coming from the sea ice surface/ ocean. Please clarify.

L61-62 or by halogens (see comment above)

L67-68 a few pptv of BrO are sufficient. Please expand following above comment.

L83 Note that using d15N in nitrate as a source tracer works only if any of the processes involved does not induce any significant isotopic fractionation. Please clarify.

L84-85 It is misleading especially for the modellers amongst the readers to cite only a single number for d15N in atmospheric nitrate originating from snow nitrate photolysis. In particular, d15N in the atmospheric nitrate above snow is not constant but changes after polar sunrise as photolytic recycling and isotope fractionation between snow and atmosphere proceed into summer, going from very negative values to near zero. Thus cite here a range of values observed in spring (when they are still strongly negative) at relevant polar locations were year-round observations are available (e.g. Wagenbach et al., 1998: Neumayer coastal Antarctica, 1986-92; Winton et al., 2020: Dome C East Antarctic Plateau 2009-15)

L88 what is the uncertainty (standard deviation) of this value?

L97 can serve as a proxy (see comment above on halogen chemistry)

L101-102 I strongly recommend to summarise in a table assumed isotope ratios for both d18 and D17O in the discussed end members (O3, OH, RO2/HO2, H2O etc), including respective uncertainties and references. Place it either here or later in section 3.3 when prevalent oxidation pathways are discussed.

L103-04 or XO; of course the O in XO originates from O3

L157-70 were the steps prior to freezing carried out on the ship right after filter exchange? please clarify.

L166 the blanc values are large compared to the ambient values. What are the N and O isotope ratios of the blancs? Were reported sample isotope ratios also corrected for the blanc contribution? This may actually have quite an impact on the reported values if the blanc comes from an isotopically very different pool.

L192-95 Considering the stability of NO2- in solution - When was NO2- measured? were samples frozen and kept in the dark? please clarify.

L197-204 Please provide also vertical information on the calculated back trajectories (this is output produced by default in your HYSPLIT runs), e.g. in the figures. Further below you discuss interaction with ocean/ sea ice/ snow surfaces, this applies only when the air mass arriving at the ship location spent time in the boundary layer.

L210, 213 interaction with sea ice. See previous comment.

L241 cite also other atmospheric nitrate observation in the relevant sector of coastal Antarctica: Halley 2004-05 (Wolff et al., 2008); Neumayer 1986-92 (Wagenbach et al., 1998)

L252 highest in early summer - supposedly due to the spring time depletion in stratospheric ozone. Please clarify.

L265-66 Please check vertical information of the corresponding trajectory to support this.

L272-73 Having a combined figure of all isotopes would make it easier to show this (see comment below).

L283 I am surprised, NOx atmospheric lifetimes are considerably shorter than for instance those of PAN (which in turn is admittedly stable at winter temperatures). How can NOx reach Antarctica from lower latitudes? Can you clarify?

L316 the triple stable isotopic composition ...

L320-25 Cite also relevant Antarctic observations e.g. Wagenbach et al., 1998: Neumayer coastal Antarctica, 1986-92; Winton et al., 2020: Dome C East Antarctic Plateau 2009-15

L328 refrence here also Frey et al., 2009; Erbland et al., 2013.

L332 I suppose there are no measurements of d15N in nitrate of the snowpack source? this is an important measurement gap to be addressed in the future.

L332-34 Please rephrase in light of the non-stationarity of the d15N in atmospheric nitrate from snow emissions (see comment above)

L343-44 limited influence from ... - the caveat is that this depends on the d15N in nitrate of the local snow source. higher d15N in atmospheric nitrate later in spring/ summer can also originate from snowpack emissions, when the source has become increasingly enriched (see comment above). Please balance your conclusion here.

L345-47 I don't understand this sentence. Please rephrase.

L347 oceanic RONO2 has been long proposed as an important net primary nitrate source to the Antarctic. This should be mentioned.

L374 reference here some of the earlier literature (Frey et al., 2009; Berhanu 2014, 2015)

L379-80 What is the expected time scale (or lifetime) of aerosol nitrate photolysis? If similar to snow nitrate (on the order of weeks), then it may not be relevant compared to the time scales of transport and deposition.

L390 qualitatively - this study is not a quantitative isotope budget

L406 higher d18O in spring possibly also due to oxidation by XO (see above)

L417 Water vapour is not an oxidant. I am a bit confused here - oxygen isotope transfer from oxidants: O3, OH (O source atmospheric H2O and O(1D) from O3 photolysis), HO2/RO2 (O source atmospheric O2) please clarify, also how is H2O(v) mixing here.

L414-31 This paragraph will greatly benefit from a table (see comment above on L101-102) to better follow your argument.

L459 increase control of O3 or XO ...

L492 potentially powerful, but complex (see previous comments); N & O stable isotope measurements of the regional sources (snow, sea ice) are required to achieve a more quantitative budget analysis. consider rephrasing

TECHNICAL CORRECTIONS

L19 emissions  ... originated from ...

L68 typo: from

L106 typo: atmospheric

L107 Antarctic tropospheric oxidation chemistry ...

L225 In Fig2 I cannot see the second highest winter value of 22 ng/m3, is it covered by other symbols?

L721 typo: atmospheric

Figures

Fig1: Label each subplot to help the reader navigate more easily, e.g. 1a. Winter-S 1d. Spring-N ... and include dates in the caption. There is a typo in the caption: ice edge transect should be (e) and N voyage (d)

Fig2,3 and 5: I strongly recommend to combine these figures including also Fig. S3. This will help to detect a lot more easily common features in [NO3-] and N & O isotope ratios. After all they are related.

To aid interpretation I also suggest to add a panel (or as a separate figure) showing air temperature, radiation (or solar elevation angel) and wind speed at the ship location.

Fig.4: Add labels to subplots, e.g. Spring-N ...; 4c: I suspect only the trajectories in bluish colours were within the atmospheric boundary layer above sea ice, whereas the ones with higher d15N (reddish colours) were likely higher up in the free troposphere. This is a point easily supported by including vertical AMBT info (see above).

REFERENCES

Bauguitte et al., Summertime NOx measurements during the CHABLIS campaign: can source and sink estimates unravel observed diurnal cycles?, Atmos. Chem. Phys., 12(2), pp 989--1002, doi:10.5194/acp-12-989-2012, 2012.

Berhanu et al., Laboratory study of nitrate photolysis in Antarctic snow. II. Isotopic effects and wavelength dependence, J. Chem. Phys., 140(24), doi:10.1063/1.4882899, 2014.

Berhanu et al., Isotopic effects of nitrate photochemistry in snow: a field study at Dome C, Antarctica, Atmos. Chem. Phys., 15(19), pp 11243--11256, doi:10.5194/acp-15-11243-2015, 2015.

Bond et al., 2023, Snowpack nitrate photolysis drives the summertime atmospheric nitrous acid (HONO) budget in coastal Antarctica, Atmos. Chem. Phys. Disc., doi:10.5194/acp-2022-845, 2023.

Erbland et al., Air--snow transfer of nitrate on the East Antarctic Plateau -- Part 1: Isotopic evidence for a photolytically driven dynamic equilibrium in summer, Atmos. Chem. Phys., 13, pp 6403-6419, doi:10.5194/acp-13-6403-2013, 2013.

Frey et al., 2009, Photolysis imprint in the nitrate stable isotope signal in snow and atmosphere of East Antarctica and implications for reactive nitrogen cycling, Atmos. Chem. Phys., doi:10.5194/acp-9-8681-2009, 2009.

Frey et al., Atmospheric nitrogen oxides (NO and NO2) at Dome C, East Antarctica, during the OPALE campaign, Atmos. Chem. Phys., 15(14), pp 7859--7875, doi:10.5194/acp-15-7859-2015, 2015.

Legrand et al., 2014, Large mixing ratios of atmospheric nitrous acid (HONO) at Concordia (East Antarctic Plateau) in summer: a strong source from surface snow?, Atmos. Chem. Phys., 14(18), pp 9963--9976, doi:10.5194/acp-14-9963-2014, 2014.

Morin et al., An isotopic view on the connection between photolytic emissions of NOx from the Arctic snowpack and its oxidation by reactive halogens, J. Geophys. Res., 117, doi:10.1029/2011JD016618, 2012.

Savarino et al., Oxygen isotope mass balance of atmospheric nitrate at Dome C, East Antarctica, during the OPALE campaign, Atmos. Chem. Phys., 16(4), pp 2659--2673, doi:10.5194/acp-16-2659-2016, 2016.

Wagenbach et al., Atmospheric near-surface nitrate at coastal Antarctic sites, J. Geophys. Res., 103(D9), pp 11007--11020, doi:10.1029/97JD03364, 1998.

Winton et al., Deposition, recycling, and archival of nitrate stable isotopes between the air--snow interface: comparison between Dronning Maud Land and Dome C, Antarctica, Atmos. Chem. Phys., 20(9), 5861--5885, doi:10.5194/acp-20-5861-2020, 2020.

Wolff et al., The interpretation of spikes and trends in concentration of nitrate in polar ice cores, based on evidence from snow and atmospheric measurements, Atmos. Chem. Phys., 8(18), pp 5627--5634, 2008.