Increase in precipitation scavenging contributes to long-term reductions of black carbon in the Arctic

Heslin-Rees, Dominic; Tunved, Peter; Ström, Johan; Cremer, Roxana; Zieger, Paul; Riipinen, Ilona; Ekman, Annica; Eleftheriadis, Konstantinos; Krejci, Radovan

doi:https://doi.org/10.5194/egusphere-2023-940

Preprints

https://doi.org/10.5194/egusphere-2023-940

Preprints

12 May 2023

| 12 May 2023

Increase in precipitation scavenging contributes to long-term reductions of black carbon in the Arctic

Dominic Heslin-Rees, Peter Tunved, Johan Ström, Roxana Cremer, Paul Zieger, Ilona Riipinen, Annica Ekman, Konstantinos Eleftheriadis, and Radovan Krejci

Abstract. Black carbon (BC), the most efficient atmospheric aerosol for absorbing light in the visible spectrum, exerts a warming effect on a region undergoing unprecedented climatic changes. Here, BC is studied indirectly using filter-based methods to ascertain aerosol light absorption parameters. We investigated long-term changes using a harmonised 21-year data set of light absorption measurements, in conjunction with air mass source analysis. The measurements were performed at Zeppelin Observatory (ZEP), Svalbard, from 2002 to 2022. We report a statistically significant (s.s.) decreasing long-term trend for the light absorption coefficient, measured at the site for the entirety of the data set. However, the last 7 years, 2016–2022, showed a slightly increasing s.s. trend for the haze season. In addition, we observed an increasing trend in the single scattering albedo from 2002 to 2022. Five distinct source regions were identified; the trends involving air masses from the five regions showed decreasing absorption coefficients, except for the air masses influenced by emissions from Eurasia. We show that the changes in the occurrences of each transport pathway cannot explain the reductions in the absorption coefficient observed at the Zeppelin station; an increase in contributions of air masses from more marine regions, with lower absorption coefficients, is compensated by the influence from high-emission regions. Along with aerosol optical properties, we also show an increasing trend in accumulated surface precipitation experienced by air masses en route to the Zeppelin Observatory. We argue that rainfall, as a sink of aerosol, plays a role in the long-term trends in the absorption coefficient, explaining approximately a quarter of the overall trend. A decreasing trend in the scavenging ratio further suggests an increase in the aerosol removal processes. We note that there is an increasing potential influence from active forest fires, particularly in the last few summers (i.e. 2015–2022). Active fires have been shown to have a significant impact on the mean seasonal absorption coefficient especially during northern hemispheric summer. However, no noticeable alteration in annual long-term trends can be observed.

Received: 08 May 2023 – Discussion started: 12 May 2023

Download & links

Preprint (PDF, 7182 KB)

Supplement (27599 KB)

Download & links

Preprint (7182 KB)
Metadata XML
Supplement (27599 KB)
BibTeX
EndNote

Dominic Heslin-Rees et al.

Status: final response (author comments only)

RC1:
'Comment on egusphere-2023-940', Anonymous Referee #1, 19 Jun 2023

SUMMARY
Heslin-Rees and co-authors present one of the longest time series of aerosol absorption coefficient in the Arctic region and correlate the trend with the origin of airmasses, precipitation and biomass burning events. The topic of the manuscript is well within the scope of ACP and is of main interest to the aerosol scientific community. The presentation and structure are, however, critical to the explanation of methodology and communication of scientific founding and its discussion. The conclusions are linear and clear and can be used as a guideline to restructure the full manuscript. Overall, I suggest the authors reworking the structure of their text (reduce the length of the manuscript, merge the result and discussion section, reduce supplementary and remove all non-essential data or information), and resubmit the manuscript for a second round of review. In the current state, the scientific message is not delivered properly while the quality of the data treatment is hard to ass, which is a pity, considering the rarity of the dataset.

MAJOR COMMENTS
Method section
I suggest restructuring this section. Aerosol, trend analysis and back trajectory and mixed together, making the methodology section not coherent I thus suggest using subsections in this order: 1) absorption measurement and data treatment including subsections on technical description, data correction, data harmonisation and trend analysis, and scavenging ratio; 2) Aerosol origin and transport including subsections on transport model, clustering method, season definition and forest fire identification.
Results and discussion section
First, the result section is full of not-needed subsections. The manuscript is already quite long with many sections and subsections, try to reduce their number by merging subsections together. The result section reads mostly as a list of results without the proper scientific context, which is confined to the discussion section. In the latter, many results already presented in Section 3 are repeated, making Section 4 unnecessarily long and redundant. I suggest complementing the results with the data interpretation and discussion in the results section. Section 4 could be, then, shortened to a minimum, or completely removed.
Supplement
Supplementary material is important to explain specific technicalities or show interesting, but not essential results. In this specific work, the supplementary is 16 pages long with 17 figures. In some cases (harmonization of data), critical information is included in the supplementary but not in the main text. So, the reader is forced to constantly check the supplementary material in order to understand the content of the manuscript. This back-and-forth disrupts the reading. Please find a way, not trivial, to avoid iterative references to the supplementary.

SPECIFIC COMMENTS
The title is a bit misleading. The manuscript present absorption data, not BC data or eBC data.
L3: specify the region
L15-16: These sentences are a bit unclear, what the authors mean with “explaining approximately a quarter of the overall trend” and “scavenging ratio” ?
L17-19: I would move the forest fire discussion up in the text, together with the “source identification” part.
L25-29: please provide some references for the climatic impacts of BC.
L35-36: absorption is poorly explained here, but I also think it does not belong to introduction but rather methodology.
L55-74: mix various topics (measuring challenge, long term dataset, cleaner combustion, emission reduction, aerosol sinks) in a relatively confused order. It is hard for the reader to follow the discussion and understand the message of this subsection.
L75-87: it is very important to define the goals of this work in the introduction. However, objectives and goals being are repeated several times. I suggest reducing the first part and focus on the bullet list, which is short and easy to understand.
L94-95: please the grammar of the phrase.
L136-137: please remove these statements. Focus on your study and not the history of Zeppelin, which is impressive but not of primary interest here.
L161-162: remove sentences, not usefull.
S3.1: You have only one subsection in section 3.1. no need for subsection here
F1: Although the text is clear, Figure 1 is hard to read and the trend described in the text cannot be observed in Figure 1. I suggest reducing the vertical scale range (even if it cuts the errors bars) and, maybe, split in three panels each of them showing the seasonal temporal series. There is a bit too much red, so that is hard to identify what you refers to.
L275: information on SSA is superficial here. I would remove this sentence.
S3.2: No need for subsection here
S3.5: No need for subsection here
I stop here the specific comments, since the manuscript need a deep revision of its structure.

Citation: https://doi.org/10.5194/egusphere-2023-940-RC1
- AC1: 'Reply on RC1', Dominic Heslin-Rees, 25 Sep 2023
  
  We thank the reviewer for their positive and constructive comments. We have modified our manuscript based on their
  
  suggestions. We believe that the comments received by the reviewer have greatly improved both the text and the science.
  
  Please find our detailed reply attached.
  
  Citation: https://doi.org/10.5194/egusphere-2023-940-AC1
RC2:
'Comment on egusphere-2023-940', Anonymous Referee #2, 27 Jun 2023

Review for manuscript
"Increase in precipitation scavenging contributes to long-term reductions of black carbon in the Arctic" by Heslin-Rees et al.

General Comments
The authors investigated mechanisms for changes in the absorption coefficient (as a proxy for black carbon) at the Zeppelin Observatory, Svalbard over a twenty-year period. Out of three potential main mechanisms, they identified and quantified as the most dominant mechanisms a decline in the absorption coefficient based on precipitation and source emissions changes.
The paper is well in scope of ACP and presents an interesting, novel and long timeseries.
Overall, the manuscript is relatively well written, especially the introduction and the conclusion. The authors have convincing arguments for their hypothesis, although data is only presented as figures. However, there are a number of issues that are unclear and need to be addressed before the study should be published. The manuscript lacks a bit of clarity with regards to the methodology used, it is in parts unspecific and there is quite some redundancy in terms of information presented in the different chapters, as well as the Supplementary Material. Especially the conversion of the PSAP values from one wavelength to another with the assumption of an Absorbing Ångström Exponent of unity. That information is only presented in the Supplementary Material, along with the remaining chapter on data harmonization, which is quite crucial to the study and should be in the method section of the main manuscript.

Main Manuscript
Specific comments (Numbers refer to Line numbers in the original manuscript PDFs)
18-19: How do these two sentences go together? Could the last sentence be clarified further?
37: "Significant" is not very specific. Can you put a number on that claim? FYI, chapters 7 onwards of the AMAP 2021 report have very recently been published. In chapter 8 they put the 2015 Total effective radiative forcing of Arctic BC at 0.96±1.21 vs Global of 0.08±1.08. Additionally, my personal preference would be to try to use this word in terms of «statistical significance» only.
73: «why models overestimate observed BC (Sharma et al., 2013)» This is not universally true. Sharma et al. used only one specific model. If you look at Fig. 7.25 of AMAP 2021, you will see that the picture looks different for every model. And this is true for all SLCFs.
82: «climatically relevant time period» I understand that 20 years of observational data of that kind is a big accomplishment, but in terms of climatically relevant data the (WMO, NASA, NOAA, etc) guidelines are usually 30 years.
93: What does «not so affected» mean? This is too unspecific.
104: “both slightly heated” Be more specific regarding the temperature of the whole-air inlet. This information is also not in Platt et al. 2022.
119: “detection limits for a given sample rate.” It is not clear to me what you mean with sample rate. Is this the flow? If so, could you give a LOD value for a typicall flow of 1 L/min? And what is the measurement frequency?
139: “regularly calibrated” What is the frequency of calibrations? and is the calibration process described somewhere?
148: Could you give more details? Were the planes at different altitudes? Why 9 starting points?
166: “as if the air parcel had travelled near the surface for the entire time period» Couldn't this effect your conclusions? What if the air parcel wasn't near the surface? What is the weighted average of parcels above or within the precipitate? Would in-cloud scavenging change your conclusions?
168-170: Please elaborate on why you made this choice. Your title says “black carbon” after all.
174-175: I think it would be useful to be more specific and mention which property of "aerosol loading" you mean and give a range of the values as well as the LOD in your text here.
177: What are the consequences of low signal to noise in summer? Is it zero values or higher uncertainty, i.e. error bars? Move this information here.
191: Regarding FigS2: It is not clear what EBAS is. From which instrument does this data come? The description in the SM is insufficient. Additionally, the sentence on line 67 does not make much sense: "Despite, the EBAS data set and the data processed from the manual PSAP"
198: Regarding FigS4: I don't see AE31 data in that plot (unlike the claim in the figure caption). The old PSAP data is off the chart. The caption of the figure could be slitghtly improved as well. E.g. description of lines, etc.
248-249: How does the lower time resolution affect uncertainties around the scavenging ratio? And do you know how does the normalization with CO influence conclusions regarding sources of different combustion types? Smouldering, more common in solid fuel combustion, usually leads to more CO as compared to more efficient combustion. This would lead to a bigger correction from e.g. biomass burning. Close to a source a CO/CO2 ratio (modified combustion efficiency) could be used to infere the type of combustion. I don't think for your receptor site, long distance from a source, this would still be true.
252: “fourth version of the GFED” Did you also use small fires? The potential number of small boreal fires could make a difference.
265 “negative trend in σap. The long-term trend, based on seasonal medians, is approximately -0.004 (-0.0063 to -0.0016) Mm−1yr−1” For one, I wonder how this changes with an updated AAE (see my comments for the Supplementary Material). Secondly, could you please give consistent significant numbers and decimals? Additionally, please mention explicitly what the values in brackets mean.
268: “correcting for autocorrelation” Mention how you correct for autocorrelation.
270-271: “The full-time series is probably not best explained using a single trend,” Isn't that quite normal for a long time series and exactly the reason why climate normal timeseries should be 30 years or longer?
272-273: “For the period after 2016, only the AHZ displays a positive s.s. (p=0.003) trend of 0.01 Mm−1yr−1 based on daily medians.” This partly contradicts the previous sentences. Could you rephrase this?
315-316: “then only perturbed based on the respective changes to the frequency of each cluster,” I don't understand this explanation. Each cluster has a different frequency. At some point you must have summed up the contribution of the different clusters if you calculated the "resepective changes ... to each cluster". This info is later repeated in the discussion but doesn’t make it clearer.
349: “9 years of data, 2002 - 2010,” What is the reason that you could not use the full twenty year data?
350: “back trajectories” Are these the same 10 day BTs used for your previous analysis? And is this the ERA5 precipitation data? I think it would be good if there would be a bit more information in this section or in the Methods section.
351-352: “For the majority of transportation clusters and seasons, there is an exponential decline in σap as ATP increases.” I don't know. This seems like a stretch. I can see a clear exponential decline in 4 of your subplots. Some might have initially an exponential decline but the y-scale such that it appears quasi-linear. In clusters 4 and 5 there is even an increase at higher ATP.
357: “The hourly ATP averages are converted to their respective σap values using the closest corresponding value for each σapATP relation” It is not cristal clear what you did here. The ATP data is hourly, but what time resolution does the σap have?
358 “see Fig. S17” The time resolution is not visible in this figure.
363: “Section 11 in the supplement” That "section" is literally one sentence and not very informing.
364: “calculated seasonal medians of σap is approximately -0.001 Mm−1yr−1” What is the uncertainty in this number? This trend is so tiny, that it might only be clearly visible on a log scale.
368-369: “The back trajectories arriving at ZEP have experienced an increasing influence from forest fires. The number of active forest fires each back trajectory traversed over has increased since 2002, with the most notable shift occurring after 2015” Can you back this up with data or a reference?
373: “not shown here” Why not? You come back to it in the Discussion and it seems relevant to your argument.
“Figure 7.” What happened in 2013? The year without a summer.
“Figure 7.” “Note that the Arctic Haze season (AHZ) of 2006 is significantly smaller compared with the MODIS fire count” Is this data now only MODIS, or also from other data? And did you remove exreme biomass burning events or not? It’s not clear to me.
Figure 7.”“Note that the y-axis is displayed on a log scale» Even on a log scale, the trend is barely visible.
377-378: “Removing the data points which correspond to the most extreme number of active forest fires (i.e. defined as the 99th percentile of a running 15-day average), reduces the seasonal means, however, has a negligible effect on the seasonal median.” Without showing us the data there is little meaning in these words. Does the figure show data where most extremes have been removed or not? How much are the means reduced? How much is negligible?
387: “increasing trend during the Arctic Haze season of 0.001 Mm−1yr−1” Again, this reads like a contradiction. You have discovered a counter trend in a longer trend. If you would look at even shorter periods, you would find trends in every direction you like. But when is a trend really a trend?
389-390: “we can conclude that the amount of scattering in relation to the absorption has increased” I've now read this several times, but am not entirely sure what you are trying to communicate.
390 “Fig. S3” All I can see in this Fig is a decrease and then stagnation in σap.
415-416: “The increase in this region southeast of Europe could also be the result of increases in contributions from sources further south” It is not clear what you mean here. You are still speaking about the emission inventory. The increase of emissions in one area does not influence another area. Emissions are "grid-bound".
426: “has been explained” Please point us to peer-reviewed publications where this is the case.
427: “further aloft” I don't think this is an entirely valid point. Do you mean chimneys? There are a lot of different types of wildfires, and they are not all crown fires, unless we get massive fires like currently in Canada. There is also a lot of burning shrub and peat, very close to the ground.
427-428: “In this study, it was shown that the potential influence from BB events has increased (see Figs. 7)” Again, I think you don not have very strong evidence. It is a hypothesis of yours. Although a reasonable one. But your evidence does not fully prove your claim.
433: “report the trends using seasonal medians” Why does the axis of Fig 7 say "mean"?
482-489: I specifically liked this part. It shows a fair and balanced view of all the involved issues.
496: “decline” Looks more like stagnation to me.

Technical corrections
2: Add "online" or "in-situ" to filter-based
8: Does the haze season increase in time or magnitude?
22: has undergone -> is undergoing
44: “major source of BC” You might want to cite (Stohl et al., 2013) here
57: There is a question mark and your two Sinha ea 2017 references are the same paper. It is a nice study but only covers one site. Here, I think AMAP 2021 would be a good (additional) reference.
73-74: “pronounced difference between trends in atmospheric and BC ice core measurements” I think this is a very valid point and it would make it more clear if you could specify the time horizon you are referring to.
173: mention that Eq 2 is in the SM
206: “number of active numbers” I assume this should be active fires?
270: “3PW” I think it would be easier to understand/read if you just write it out. You only use this abbrevaition once, besides it's introduction.
276: Regarding FigS8 “decrease in σap.” A decrease (or any trend) of this value is really difficult to see in the figure. Could you add trend lines?
284: “, the trend is calculated for each and every grid cell on an annual resolution” Are all these trends s.s. ?
335: “quite some” Please be more specific.
338: “Short-term” Define what this means.
353: “decline” doesn't hurt to write that you mean decline in σap, to be more clear.
355: “exhibit the lowest amounts of wet removal strengths;”) Shouldn't this be the highest amount of wet removal strength?
360: “The proportion of the trend is approximately the same” That is not very specific information. Almost useless.
363: “Fig. 7” change to Fig 6
375: “it is clear” Not really it isn’t.
375: “potentially influential» On what? Be more specific (BMS).
379: “no impact on the long-term trend” Of what? BMS
383: “distinctly lower values” Please remind us again from where to where the values went. It's really hard to read any numbers of the chart of figure 1, i.e. where was the LMS value in 2003 vs 2022?
385: “anthropogenic influence” I guess you could call a decrease in haze and other pollution, partly due to in increase in wet scavenging, anthropogenic influence, since we influence everything. I just think it reads funny
426: “a signature” what exactly do you mean with this? BMS
431: “just 3% of the data” By time or CWT or what measure? BMS
434: “when» did you mean to write "if" or "why"?
436: “potential” prossibility?
497: “most recent years” BMS
504: “there has been a shift in the sign of the trend.” When? BMS
505: “collocated” How do you collocate a property that has a fixed place with something that is moving?

Supplementary Information
General Comments
The figure captions are really insufficient and many figures are poorly described.

Specific comments
35:“For periods in which nephelometer measurements were invalid or not present,” (p. 2) Show this in Fig S1
49: “σap is acquired from EBAS” (p. 3) based on which instruments?
56-57: “Only values greater than 0 Mm−1 were considered valid. This is despite the stated detection limit by the manufacturers, for a 30-minute time resolution, being approximately <0.13 Mm−1” (p. 3) Why would you do this? At least explain your rational.
75-76: “assuming an Absorbing Ångström Exponent (ÅAE) of 1 (i.e. pure EC) (see Eq.2). Examining the ÅAE from the Aethalometer data this was considered a fair assumption.” (p. 3) (Liu et al., 2018) have shown that an AAE of 1 is not an appropriate assumption. You should either back up your claim that "Examining the ÅAE from the Aethalometer data this was considered a fair assumption." by showing us the data and statistics, or consider using size distribution data (SP2) to find the geometric mean diameter (GMD) and an appropriate AAE, e.g. analog to (Tunved et al., 2021)
FigureS4 caption: “Aethalometer (A31) data” (p. 6) Where is this data?
92: “zeroing mode” (p. 7) Can you explain this for people not familiar with the TSI?
FigureS7 caption: “at least 5 year’s worth of long-term data is required to perform any reliable trend analysis” (p. 8) Says who? Can you give a reference and could you also give the absolute trend?
“Figure S14.” (p. 12) caption missing

Technical corrections
51: “1000 l” (p. 3) In this font the "1" looks like the "l". I suggest to use capital L, always.
88: “need to: reference as to why the TSI is better)” (p. 7) Please enlighten us.
91: “additional problems” (p. 7) Additional to what exactly?
FigureS6 caption: “red” (p. 8) blue?
FigureS6 caption: “The number of active fires at any given time has been calculated and coloured according to the number of fires present in that grid” (p. 8) This is not explained very clearly, the color bar label is also unclear.
140 “short-term perturbations in” (p. 13) what is short-term
FigureS15 caption: “S” (p. 13) undefined
FigureS16 caption: “short-term” (p. 14) unlcear what this means
159: “despite” (p. 14) Despite?

Recommendation
I recommend publication after the herein suggested major revisions to this manuscript.

References
Liu, C., Chung, C. E., Yin, Y., and Schnaiter, M.: The absorption Ångström exponent of black carbon: from numerical aspects, Atmospheric Chemistry and Physics, 18, 6259–6273, https://doi.org/10.5194/acp-18-6259-2018, 2018.
Stohl, A., Klimont, Z., Eckhardt, S., Kupiainen, K., Shevchenko, V. P., Kopeikin, V. M., and Novigatsky, A. N.: Black carbon in the Arctic: the underestimated role of gas flaring and residential combustion emissions, Atmospheric Chemistry and Physics, 13, 8833–8855, https://doi.org/10.5194/acp-13-8833-2013, 2013.
Tunved, P., Cremer, R. S., Zieger, P., and Ström, J.: Using correlations between observed equivalent black carbon and aerosol size distribution to derive size resolved BC mass concentration: a method applied on long-term observations performed at Zeppelin station, Ny-Ålesund, Svalbard, 73, 1933775, https://doi.org/10.1080/16000889.2021.1933775, 2021.

Citation: https://doi.org/10.5194/egusphere-2023-940-RC2
- AC2: 'Reply on RC2', Dominic Heslin-Rees, 26 Sep 2023
  
  We thank the reviewer for their positive and constructive comments. We have modified our manuscript based on their suggestions. We believe that the comments received by the reviewer have greatly improved both the text and the science. Please find our detailed reply attached.
  
  Citation: https://doi.org/10.5194/egusphere-2023-940-AC2

Dominic Heslin-Rees et al.

Supplement

https://doi.org/10.5194/egusphere-2023-940-supplement

Dominic Heslin-Rees et al.

Viewed

Total article views: 419 (including HTML, PDF, and XML)

HTML	PDF	XML	Total	Supplement	BibTeX	EndNote
276	128	15	419	51	6	11

HTML: 276
PDF: 128
XML: 15
Total: 419
Supplement: 51
BibTeX: 6
EndNote: 11

Views and downloads (calculated since 12 May 2023)

Month	HTML	PDF	XML	Total
May 2023	132	65	5	202
Jun 2023	73	25	5	103
Jul 2023	27	13	3	43
Aug 2023	27	10	0	37
Sep 2023	17	15	2	34

Cumulative views and downloads (calculated since 12 May 2023)

Month	HTML	PDF	XML	Total
May 2023	132	65	5	202
Jun 2023	73	25	5	103
Jul 2023	27	13	3	43
Aug 2023	27	10	0	37
Sep 2023	17	15	2	34

Viewed (geographical distribution)

Total article views: 425 (including HTML, PDF, and XML) Thereof 425 with geography defined and 0 with unknown origin.

Country	#	Views	%

Latest update: 26 Sep 2023

Short summary

Light-absorbing atmospheric particles (e.g. black carbon (BC)) exert a warming effect on the Arctic climate. We show that the amount of particle light absorption decreased from 2002 to 2022. We conclude that in addition to reductions in emissions of BC, wet removal plays a role in the long-term reduction of BC in the Arctic, given the increase in surface precipitation experienced by air masses arriving at the site. The potential impact of biomass-burning events is shown to have increased.


Total:	0
HTML:	0
PDF:	0
XML:	0