the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 17 Oct 2023
Simultaneous organic aerosol source apportionment at two Antarctic sites reveals large-scale and eco-region specific components
David C. S. Beddows
Anna Jones
Thomas Lachlan-Cope
Matteo Rinaldi
Stefano Decesari
Francesco Manarini
Mara Russo
Karam Mansour
Roy M. Harrison
Andrea Mazzanti
Emilio Tagliavini
Manuel Dall'Osto
Abstract. Antarctica and the Southern Ocean are the most pristine areas of the globe and represent ideal places to investigate aerosol-climate interactions in an unperturbed atmosphere. In this study, we present PM1 (Particulate Matter < 1µm) source apportionment for two sample sets collected in parallel at two British Antarctic Survey (BAS) stations, namely Signy and Halley, during the austral summer 2018–2019. We find that Water Soluble Organic Matter (WSOM) is a major aerosol component at both sites (average 25–33 %). Remarkable differences between pelagic (open ocean) and sympagic (influenced by sea ice) air mass histories and related aerosol sources are found. The application of non-negative factor analysis techniques to H-NMR spectra of the samples allows the identification of five Organic Aerosol (OA) sources: two primary (POA) types, two secondary (SOA) types, and a fifth component of unclear origin possibly associated with the atmospheric ageing of primary emissions and dominating at Halley. Overall, the concentrations of primary and secondary organic aerosols are prevalently dictated by the emissions in sympagic and pelagic marine regions, with atmospheric circulation causing to establish marked latitudinal gradients only for some of such aerosol components. Our results strongly indicate that various sources and aerosols processes are controlling the Antarctic aerosol population, with the emissions from sympagic and pelagic ecosystems affecting the variability of submicron aerosol composition both in maritime areas as in inner Antarctic regions.
- Preprint
(2982 KB) - Metadata XML
-
Supplement
(2150 KB) - BibTeX
- EndNote
Marco Paglione et al.
Status: final response (author comments only)
-
RC1: 'Comment on egusphere-2023-2275', Anonymous Referee #1, 11 Nov 2023
Review for ‘Simultaneous organic aerosol source apportionment at two Antarctic sites reveals large-scale and eco-region specific components’ by Paglione et al., submitted for publication in Atmospheric Chemistry and Physics (ACP)
Paglione et al. describe the chemical composition of aerosol particles (PM1) collected at two Antarctic stations (Signy and Halley) during a field campaign from December 2018-March 2019. For the chemical analysis of the organic and inorganic compounds they applied several types of offline methods, such as H-NMR and ion chromatography. They observed differences in the chemical composition based on the dominance of either open ocean or sea ice in the air mass history. By applying non-negative matrix factorization, they identified five sources for organic aerosol relevant to this polar marine region.
Ambient aerosol particles contain a plethora of organic molecules, where a majority still eludes chemical characterization and quantification using classical instrumental approaches. The use of H-NMR appears to me unconventional, but promising with the large potential to enhance the knowledge about chemical compounds in aerosol particles. Even though, I am aware that the colleagues from Bologna are skilled and experienced in the use of NMR, I am very impressed, partially sceptic, with which confidence the authors attribute signals to individual compounds in the ‘forest’ of NMR signals in ambient samples, where potential interferences from other chemical compounds in the NMR spectra are very likely.
I believe that the scientific community could largely benefit from the findings of this study. A revised manuscript should be eventually published in ACP. However, I noticed large gaps in confirmability and thoroughness, which makes it hard to readers without profound knowledge in NMR analysis to understand (and get convinced of) the train of thoughts of the authors from their observations to their scientific conclusions. To enhance accessibility, I encourage the authors to conduct a meticulous revision, addressing these gaps and providing clarity in their reasoning from observations to scientific conclusions.
General concerns:
The presentation of chemical compounds in this study lacks clarity, resulting in a somewhat chaotic manuscript. Figures 2, 4, and 6 highlight various substances such as WSOM, ammonium, nitrate, ‘nSS other ions’, nSS sulfate, and sea salt. Figures 3 and 5 introduce amines, MSA, and organic functional groups, including unsubstituted aliphatic groups, polysubstituted aliphatic groups, and anomeric and vinylic groups. Figure S3 identifies lactic acid, low-molecular fatty acids, glucose, sucrose, ‘generic polysaccharides’ (What is that?), acidic sugars, neutral sugars, glycerol, MSA, TMA, and DMA. Figure S10 includes HMSA and 'acidic sugars (e.g., uronic acids) or sulfonate-esters'. Additionally, the written text discusses 'threitol' (Line 265), ‘oligomers (such as betaine)’ (SI, Line 60), ‘glycolipids and phospholipids’ (SI, Line 61), ‘lipopolysaccharides’ (SI, Line 72), and ‘lipids and polyols’ (SI, Line 74). To enhance the clarity and organization of this information, I suggest incorporating two comprehensive tables into the manuscript. The first table should summarize substances identified through chromatographic analyses, while the second should focus on substances identified through H-NMR analysis. The proposed H-NMR tables could include columns for the ‘name of the substance/functional group’, ‘chemical shifts used for identification’, ‘chemical shifts used for quantification’, ‘examples for molecules’, and ‘references’. This structured approach would greatly improve the accessibility and overall understanding of the diverse chemical compounds identified in this study.
Connected to the previous comment, the authors should elaborate how they conclude from certain signals of groups (e.g. unsubstituted aliphatic groups, polysubstituted aliphatic groups, and anomeric and vinylic groups) to specific substances, such as lipopolysaccharide, low-molecular fatty acids or polyols.
Please revise the use of the abbreviations throughout the manuscript: (I) The authors introduce many abbreviations and are partially not needed, since they don’t appear another time: e.g. ‘open ocean (OO)’ (Line 77). (II) The authors introduce other abbreviations repeatedly throughout the manuscript: (e.g. ‘methanesulfonic acid (MSA)’ (Line 46), ‘methane-sulfonic Acid (MSA)’ (Line 152), ‘methanesulphonic acid (MSA)’ (Line 290)). (III) The authors introduce abbreviations and do not use them consequently (E.g. Line 129: ‘factor of 2 was used to estimate the WSOM from organic carbon…’, ‘organic carbon’=WSOC?, or DMA, TMA versus dimethyl amine, trimethyl amine)
It is not clear to me, which data (from chromatographic and H-NMR analysis) eventually were included in the factor analysis. Which signals were used to receive these five Factors?
Given the significance of the 'sympagic' versus 'pelagic' environment, particularly emphasized in the Results and Discussion sections, providing a summary of the current understanding of their influence on chemical composition from latest literatures in the introduction would greatly enhance the comprehensibility of the manuscript.
Data availability: Atmospheric data from the Antarctic and Southern Ocean are sparse and hence precious to the scientific community. The authors have not published any raw data of atmospheric concentrations in the current version of the main manuscript or supplement, which makes it impossible to reproduce, reuse or compare their results. For a more transparent research and a reusability of field data for future projects, the authors are strongly encouraged to publish their atmospheric concentrations on a public repository, such as PANGAEA. The comment ‘Data are available from the authors on request’ (Lines 456-457) should not be accepted by scientific journals anymore.
Specific comments:
Line 21: You mention the ‘non-negative factor analysis’ in the abstract. However, this term does not appear anywhere in the manuscript anymore.
Line 81: Who is ‘we’? please add citation of reference.
Line 93: ‘It is becoming clear that in order to address important research questions in the polar regions it is essential measuring at multiple stations with a strong international scientific cooperation (Dall´Osto et al., 2019; Schmale et al., 2021).’ In general, I agree with the authors that international collaborations are essential for advancing in science. However, I don’t believe this sentence is suitable for finishing an introduction of a scientific paper. Instead, I recommend to replace this sentence with one that is related to your scientific findings or atmospheric implications.
Line 105: ‘[...] with mean annual air temperature of 3.5 C and annual precipitation ranging from 350 to 700 mm, primarily as summer rain.’ First, the authors should give this meteorological information for both stations (not only Signy). Second, it would be more interesting to give a meteorological overview just for the period relevant for the campaign, not for the entire year.
Line 107: How did you prewash and prebake the quartz fiber filters? Solvent? Temperature?
Line 119: ‘Aerosol offline measurements and H-NMR analysis’- Isn’t H-NMR one of the ‘offline measurements’?
Line 122-123: The authors measured organic acids, such as acetate, formate, oxalate. Where they all below limit of detection? Or why do they not appear as part of your WSOM discussion?
Lines 124-127: The authors give here information on the chromatographic analysis of inorganic ions. However, on which column and how did you analyze amines?
Line 209: ‘The meteorological conditions are not statistically different […]’ How did you test the statistical difference?
Line 265: Why do you mention ‘threitol’ as a possibility here? Is glycerol not so sure? Threitol is not discussed anywhere else within the manuscript, so you should elaborate it a bit. When glycerol and threitol are possible polyols, then why not arabitol or mannitol (tracers for fungi in aerosol particles)?
Line 266: Do you equate ‘low-molecular-weight fatty acids’ with ‘lipids’? I think this is deceptive and misleading for the readers to expand from a small subgroup to a big diverse class of molecules. Instead, I recommend to stick to the most correct terms possible.
Line 275: Based on which observation do you assume that this POA is ‘apparently being transported for thousands of kilometers and across the Antarctic continent.’?
Lines 461-465/ Author contributions: The abbreviation ‘M.R.’ could stand for both coauthors ‘Mara Russo’ and ‘Matteo Rinaldi’. Rethink your abbreviation system. Who is ‘D.S.C.B.’? (In the author list there is only a ‘David C.S. Beddows’. What were the contributions of the coauthors ‘Roy M. Harrison’ and ‘Thomas Lachlan-Cope’? They were not mentioned in this section.
Figure 1b: I guess the bluish colors represent sea ice. If so, please add a legend to the plots. Furthermore, I was wondering what about the role of shelf ice. Was this considered in the air mass history analysis and discussion in your study? In Figure 1b the Antarctic shelf ice regions are currently presented with the same coloring like open ocean, which might be misleading.
Figures 3 and 5 and 6b: Clarify the abbreviations denoting functional groups in the figure caption and ensure the usage adheres to proper English conventions for chemical terms (e.g., 'alif.' versus 'aliphatic').
Figure 6a: Use the same sequence for the substances in the plot as you did in Figures 2 and 4.
Table S1: Remove the last two sentences (‘Whilst the start and end […] fits the purpose of the work presented’) from the table caption. The column ‘Month of the study’ is redundant in regard of column 3,4,6 and 7. I recommend to remove column “Month of the study’. Instead applying a proper date-time format to the remaining columns (e.g. dd/mm/yyyy hh:min)
Table S2: Add Std to (S5;RH(%)). Should wind directions (WD) be averaged over sampling time considering that it is an angular dimension? I instead recommend to define four sectors (e.g. North: 315°-45°; East:45°-135°, South:135°-225°, West:225°-315°) to give in this table the percentage of time where wind came from which sector.
SI, Lines 60: How is betaine related to an oligomer?
Si, Line 77 and in other parts of the manuscript: ‘Dall’Osto et al. 2023, in prep.’. I would recommend to remove the year. When this manuscript is still in preparation and not submitted at least, it certainly won’t be citable for 2023 anymore.
Figure S3: You identify glucose and sucrose in Factor 1, which are known to be neutral sugars. In Factor 5 you attribute a complete different chemical shift to ‘neutral sugars’, which seem to be identical to ‘polysaccharides’ and ‘glycerol’ in Factor 1. It appears inconsistent. Is it possible that the signals currently assigned to glucose and sucrose could also be other monosaccharides, disaccharides or derivatives, such as fructose, arabinose, trehalose or levoglucosan? How does your assignment of substances (e.g. sugars) match with the findings of other groups? Are there publications on glucose, saccharose, ‘generic polysaccharides’ in the atmosphere of the Antarctic or Southern Ocean (maybe using other analytical tools)?
Figure S8: Why do you differentiate between ‘open ocean (<60°N)’ and ‘open ocean (>60°N)’. How would it impact the results of the measurements? Shouldn’t it be ‘°S’ instead of ‘°N’?
Figure S10: Eliminate the red underlining beneath 'Lac' (It appears the figure may have been copied from PowerPoint or Word).
Citation: https://doi.org/10.5194/egusphere-2023-2275-RC1 -
RC2: 'Comment on egusphere-2023-2275', Anonymous Referee #2, 27 Nov 2023
The authors present the analysis of filter data from two stations located at different latitudes in the Western Antarctic sector. They use ion chromatography and H-NMR to identify ionic species and water-soluble organic carbon functionalities and then apply positive matrix factorisation to the data to get insights into the sources of aerosols at those two sites. My major questions/comments are related to H-NMR and PMF. I am aware that authors from Bologna have previously used H-NMR for the analysis of aerosols, but this is not a method that is widely adopted in aerosol science and therefore, I think that this manuscript requires more information related to confidence in assigning certain functionalities (and especially compounds – e.g. lactic acid associated with Factor2) and also more on quantification (how is the intensity of various shifts converted to H amount and from that to water-soluble organic aerosol?) The authors are using WSOM (And later WSOA), when it should actually be WSOC. In relation to PMF I wonder if the small number of samples (n=22) is affecting the reliability of the method? I believe this should be briefly discussed in the paper. Have the authors considered downscaling and then including nss-SO4 and Na+ as a marker for seasalt (and possibly other ions) in the PMF? Or at least look at how some of the factors correlate with certain ionic compounds (e.g. POA pelagic with Na+).
The authors have not discussed MSA separately, but have grouped it into WSOM. As one of the most important compounds in marine environments, I think it should be reported separately. Looking at the MSA/nss-SO4 at those two sites would be valuable. There might be differences that can then be related to the origin of airmasses and atmospheric circulation. I also think that it would be valuable for the atmospheric community to include a supplement table containing ionic composition for all the samples.
There should be some consistency when reporting numbers. For average values the authors sometimes include error bars and in some cases they don’t. Please include error bars in all of your reported average values.
My other questions/comments are listed below.
Southern Ocean and Antarctica are very undersampled regions of the world and data coming from that part of the world is very valuable. However, I think that the authors should explore their data a bit more (e.g. MSA, MSA/nss-SO4 ratio) and provide more details related to their analytical and data treatment approaches. I would love to see this manuscript published in the ACP, but only after major revisions.
Specific comments/questions:
Line 20: “average 25-33%” seems a bit unusual. Is 25-33% minimum to maximum. It would be better to report it as average +/- ??. Also ~30% seems like a lot and I can see from reading further that this is due to including MSA in the WSOM. This is one of the most important compounds in the Antarctic environment and should be reported separately.
Line 26: latitudinal gradients for which aerosol components?
Line 26: the last sentence is quite vague. Is it possible to be more specific? E.g. what are the main differences between pelagic and sympagic emissions?
Line 30: instead “considered a window to the preindustrial atmospheric condition”, I suggest: considered to be a proxy for the preindustrial atmospheric conditions. Why is that so? I suggest adding a sentence providing an explanation of why is that region a proxy for the pre-industrial atmosphere?
Line 33: 106 instead of 106
Line 39: waves breaking…and bubble bursting
Line 42: if aerosol chemistry in southern high latitudes is described as “much more complex” then it would make sense to give more than one example confirming that complexity. (E.g. Involvement of iodine based compounds should be mentioned)
Line 42: no need for capital s in Southern
Line 53: there are studies presenting different results (i.e. MSA representing a significant portion of organic mass) and they should be mentioned in the introduction. E.g.:
Fossum, K.N., Ovadnevaite, J., Ceburnis, D. et al. Summertime Primary and Secondary Contributions to Southern Ocean Cloud Condensation Nuclei. Sci Rep 8, 13844 (2018). https://doi.org/10.1038/s41598-018-32047-4
Matteo Rinaldi, Marco Paglione, Stefano Decesari, Roy M. Harrison, David C.S. Beddows, Jurgita Ovadnevaite, Darius Ceburnis, Colin D. O’Dowd, Rafel Simó, and Manuel Dall’Osto, Environmental Science & Technology 2020 54 (13), 7807-7817
Jung, J., Hong, S.-B., Chen, M., Hur, J., Jiao, L., Lee, Y., Park, K., Hahm, D., Choi, J.-O., Yang, E. J., Park, J., Kim, T.-W., and Lee, S.: Characteristics of methanesulfonic acid, non-sea-salt sulfate and organic carbon aerosols over the Amundsen Sea, Antarctica, Atmos. Chem. Phys., 20, 5405–5424, https://doi.org/10.5194/acp-20-5405-2020, 2020.
Line 62: I suggest “airmass origin” instead of airmass type
Line 64: what do you mean by continental – Australian or Antarctic?
Line 65: latitudinal gradient in CCN was also reported in Humphries et al (2021), not only Sanches et al (2021). Please include that in the sentence.
Line 70: should be >60 S
Line 72: By analyzing simultaneous aerosol size distribution measurements at three sites, Lachlan-Cope et al (2020) showed..
Line 73: is more complex instead of are more complex
Line 73: it might not be understandable to every reader what the authors mean by “the simple sulfate-sea salt binary combination” and how does that reflect on aerosol concentrations and size distribution. I suggest revision of this part to make it clearer.
Line 83: Sentence starting with “Overall..” should be split into two separate sentences: Multiple eco-regions around Antarctica were found to act as distinct aerosol sources (Decesari et al., 2020; Rinaldi et al., 2020). However, the potential impact of the sea ice (sympagic) planktonic ecosystem on aerosol composition is frequently overlooked.
Line 84: The sentence starting with “Decesari et al..” needs to be revised to provide context. E.g. Decesari et al have investigated aerosol composition in…(where?) and found …
Line 105: why is annual temp and precipitation reported for Signy and not for Halley? Should be rerported for both, or none.
Line108: are the sampling sites influenced by station activities? How was ensured that aerosols coming from station combustion activities were not collected?
Line 123: what inorganic ions were measured? It should also be described here how sea salt was calculated.
Line 137: Should it be DCOOD instead of HCOOH?
Line 143: How are hydrogen concentrations calculated and how are they converted to organic carbon? This quantification should be described in a bit more detail.
Line 163: how does the low number of samples (22) affect the performance of factor analysis? Also, both methods should be briefly described in the Supplement.
Line 190: as already mentioned, it should be described in the methodology how sea salt was calculated; 55% +/- ??
Line 191: I am assuming MSA is included in WSOM? I suggest reporting it separately as it is a major DMS oxidation product.
Line198: some percentages include standard deviation/error, some don’t. please be consistent and include +/- with all of your reported values
Line 203: with respect
Line 207: Enriched by mSA
Line 209: You could colour Signy backtrajectories in Figure 1 b with lighter and darker red with each colour representing one of the two described periods. Or Fig6 backtrajectories could be coloured by a date.
Line 217: lower variability is likely due to the short sampling period
Line 221: add +/- to your percentages
Line 224: primary aerosol instead of primary sources.
Line 30: primary aerosols instead of primary sources.
Line 232: how far is Halley from the open ocean? Halley is a coastal site, so please specify here what do you mean by open ocean.
Line 233: contrary to Signy where alkyl-amines represent ??% of the PM1 mass.
Line 257: Please provide a sentence or two describing based on what was the 5 factor solution described as the most robust and informative.
Line 258: how are CWTs calculated? In Fig 8, what do the units for the colour legend represent? CWT should be described in the methodology and colour scale description can be given in Fig 8 caption.
Line 262: this is Water soluble fraction. Lipids are not water soluble. Could alkyl chains be coming from something else?
Line 278: what does “lac” stand for?
Line 280: how can you be sure that peaks at 1.35 and 4.21 ppm belong to lactic acid? Based on figure S4 Factor 2 seems to be dominated by MSA and DMA. It looks like factors and their corresponding NMR spectra have been mixed up.
Line 286: How are backtrajectories shorter? Backtrajectories look the same for all factors, they are just coloured differently. Backtrajectories coloured by the highest intensity seem to be coming south from Signy
Line 291: is there a difference in NMR shift between different alkyl-amines (i.e. can you tell wth certainty that 2.71 is coming from DMA)?
297: planetary boundary layer instead of PBL. Indeed, the majority of backtrajectories at Halley are coming from the Antarctic continent, but that alone is not sufficient to describe katabatic outflow. Humidity and temperature need to be considered here as well. How was it determined what percentage of time was spent below or above the boundary layer? That should be outlined in the methodology.
301: same as for DMA and MSA in the Factor 3, indicate TMA NMR shift here. DMA is also present here. To me it would make more sense to call this factor “alkyl-amines + MSA”
309: HC-O and H-C seem to be very low comparing to MSA and DMA
313: have never been
314: this is the first time that the acidic nature of aerosols in Halley has been mentioned. Where is this coming from and how has this been demonstrated? This cannot be assumed only based on high abundance of nss-SO4. To me it looks like factor 5 is simply a mix of Factor 1 and 2 and in line with that I think there is no enough evidence to hypothesise about the formation of organo-sulfates.
340: if is not clear to me what the authors mean by “the changing position of the atmospheric polar front? Do you mean the change in latitude? There are seasonal shifts of the atmospheric polar front but the campaign described in Humphries et al is too short for those changes. There are day-to-day variations in strength and slight variations in position of the atm. polar front, but Humphries et al do not investigate that, it only reports a noticeable change in atmospheric composition upon crossing the atm. polar front (i.e. crossing into the polar cell).
350: based on what has it been determined that 60% of the air masses have been linked to katabatic winds? Continental wind direction is not sufficient to classify winds as katabatic winds. If katabatic winds are going to be discussed here, then I would suggest to define them.
380: instead of “across latitudes” please put at both locations
389: can you please clarify what do you mean by breaking barriers? It is very reasonable to expect factor 3 to be present at both locations, regardless of atmospheric circulation. I suggest to avoid using the term “across latitudes” as it is suggesting that measurements were done at many different locations. Is it WSOA or WSOM?
398-400: as mentioned previously not every Antarctic continental flow is katabatic flow.
425: MSA should be reported separately and then report fractions of WSOM.
440: This study had a relatively low number of samples (especially at Halley) so it seems a bit far-fetched to make conclusions about “chemical segregation” and prevention for certain OA types appearing at the site
700: Figure 2: Please split WSOM into MSA and the rest of WSOM (e.g. MSA lighter green; the rest darker green).
Figure 9: why does Signy (whole) not have error bars included?
Supplement:
Line 30: for consistency, please use Positive Matrix Factorisation (PMF) throughout the text.
Citation: https://doi.org/10.5194/egusphere-2023-2275-RC2
Marco Paglione et al.
Marco Paglione et al.
Viewed
| HTML | XML | Total | Supplement | BibTeX | EndNote | |
|---|---|---|---|---|---|---|
| 194 | 84 | 8 | 286 | 23 | 5 | 9 |
- HTML: 194
- PDF: 84
- XML: 8
- Total: 286
- Supplement: 23
- BibTeX: 5
- EndNote: 9
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1