Schmale e et al. present a comprehensive and well written overview of the aerosol properties and trends at the Arctic sites over several decades. Arctic is changing rapidly, and this study is a good benchmark for further assessments of potential changes in aerosol properties in future climate and emission regimes.

I think the study can be published with some improvements suggested more specifically below. I would especially encourage the authors to make use of the 30 years of SIA observation (especially SO4, Na, Ca) at Zeppelin and not only use a dataset with much shorter time period. Pallas also have observations of SO4 since 1996 and Na,Ca,K since 2003, which the paper would benefit from using. All these data are available in EBAS.

More specific comments:

• Line 90. Maybe add references and numbers from the IPCC report recently published, which gives the must updated assessment of the impact of SLCF
• Line 96. Before the establishment of the Zeppelin observatory monitoring was conducted in Ny Ålesund. Thus observations at Svalbard foes back to the seventies.
• Line 160: The results from Collaud Coen et al (2020) should be added here.
• Line 166. Do not really see it is a contrast to Platt et al 2021. Different time periods and season contra annual trends. Maybe use another word than contrast.
• Line 182. Not sure if I understand this statement correctly. “no increased contribution during the fire season has been shown so far.” Are you talking about trend studies or influence by fires. Surely there are several examples of high episodes observed at the Arctic sites caused by emissions from fires (e.g. Stohl et al 2007)
• Line 327. Not clear how you have selected valid time series for trend analysis. With “at least five consecutive years” does that mean you can make a decadal trend with only five years? You need at least 7 years of data to use the Mann Kendall and Theil-Sen slope method.
• Line 353 contradicts a bit with what is written in line 357 (EBC vs SO4 seasonal trends respectively at Zeppelin compared to other sites). Sulfate is also easily washed out by rain and one would maybe expect similar behaviour. The differences may be due to different time periods used for comparing the seasonality between sites.
• Line 526. Even though the economic breakdown in Eastern Europe (and Soviet Union) results in a relatively steep decline in SO2 emissions in the nineties, the main reason for the large reductions in Europe are due to international protocols agreed upon under the UNECE LRTAP and EU directives, and in North America by the Clean Air Act.
• The increase in So4 during summer is not seen in the regular monitoring at Zeppelin for the period 1990-2019 nor for shorter time periods. See figure
  
  
• The dust tracer (Ca and Fe) measured are probably mainly water soluble (at least Ca2+) and may not be a good indicator of the real influence of dust.
• Surly using harmonized standard methods for aerosol optical properties and other aerosol observations are extremely important, and this is being developed. The AMAP (Arctic) sites should follow recommendations made by WMO/GAW. In addition, the observations should be made available and reported in a standardized protocol to international frameworks (GAW, AMAP, EMEP etc. See i.e. discussion by Laj et al 2020 : https://doi.org/10.5194/amt-13-4353-2020
• Figure 1 in *: spelling error. Sind should be since. Find the reasoning for why not the longest time series have not been chosen quite strange. Why not use the longest one available.
• Figure 2, in fig h) should that be red and not orange? Not Zeppelin data displayed

Table 1:

• It would have been useful to add for which time periods you are using data from for all the species
• Seems a bit strange to distinguish EBC from absorption coefficient since these are derived from the same instrument.
• Alert: Why are Absorption and scattering data taken from personal communication and not from EBAS. 2005-2019 available and also used by Collaud Coen et al. in their trend study
• Zeppelin: As mentioned above, I find it strange that you have not used the SO4,Na,Ca, from EBAS (and SO4 presented in Platt et al 2021), which has much longer time series than what is available by Sharma et al (2012). Further in Sharma et al, only MSA data is presented even though they probably collected many other species. The EBC data (or absorption rather) and scattering data are also available in EBAS as presented in Collaud Coen et al.

Review of “Pan-Arctic seasonal cycles and long-term trends of aerosol properties from ten observatories” by Schmale et al. in ACPD.

The manuscript presents a coherent analysis of an extensive Arctic dataset on aerosols and aerosol properties, focusing on uncovering their seasonal and decadal variability. Both natural and anthropogenic aerosol tracers are considered in the analysis and manuscript discussion, accurately highlighting the complexity of the issue. The manuscript describes nicely how different sources are important on different sides and at different times showing that the Arctic is not a homogenous area in this respect. The manuscript is well written, enjoyable to read and makes an important contribution for the scientific field.

Most of the data for the analysis are gathered from a publicly available nilu/ebas database. Although a fraction of data with supporting conclusions have been presented in previous literature, the systematic and comprehensive analysis and the new, previously unpublish data bring clearly new perspectives and conclusions. I recommend publishing the manuscript in ACP after modifications. I consider that the requested modifications do not require a major work and are therefore minor.

General comments:

My main general comment is about the description of the methodology. The manuscript concludes (action item 4) on the importance of standardized measurement and data practices. Was this met here? For example, was ebas level2 data always used? If so, were the data that were taken from personal communication treated (and measured) similarly? This should be mentioned in the text. Why Tiksi nephelometer was measuring at “ambient humidity”? How much was the RH? How the data flags (in ebas) were considered in data analysis? Which flags were included and which were omitted from final analysis? Would you please consider adding some relevant information on at least these, and other according to your consideration, in the methods section and data availability section (also see my comment regarding Table 1). In principle, information should be sufficient to repeat this study.

Another minor general comment I have on the discussion of primary aerosol sources (mainly sea spray and dust here). Their concentrations are not solely dependent on the changes in air flow routes and availability of the source (open sea water, open land areas) but the resuspension depends largely on the winds (speed). The seasonality of the winds over different sea areas around the Arctic could be mentioned somewhere in the text, and maybe some information added if there were notable changes in storms or other wind patterns in recent decades?

As a final general comment I add that from northern Scandinavia (e.g. Pallas) there are additional long-term data series on aerosol chemical composition available in ebas that the authors could consider adding in the analysis if they find that those could add to the conclusions.

Minor comments:

p3. l100. Although aerosol optical parameters in Pallas have been measured since 2000, aerosol number and ionic composition measurements in Pallas started earlier, in 1990s (data in ebas). Since this sentence refers to Arctic pollution monitoring maybe a better reference of the historical development of Pallas research activities and monitoring would be Lohila et al., 2015:

Lohila A., Penttilä T., Jortikka S., Aalto T., Anttila P., Asmi E., Aurela M., Hatakka J., Hellén H., Henttonen H., Hänninen P., Kilkki J., Kyllönen K., Laurila T., Lepistö A., Lihavainen H., Makkonen U., Paatero J., Rask M., Sutinen R., Tuovinen J.-P., Vuorenmaa J. & Viisanen Y. 2015: Preface to the special issue on integrated research of atmosphere, ecosystems and environment at Pallas. Boreal Env. Res. 20: 431–454. http://hdl.handle.net/10138/228278

p4. l119. Consider starting “the surface aerosol observations”, since it is not clear what “observations” are meant, especially after a long paragraph solely on eBC.

p5. l147. Tiksi (not Tiski)

p5. l154. Consider re-phrasing. It is not clear what is “the same database” used for “the study” of annual cycles. Assuming you mean: https://doi.org/10.21336/gen.1, would be clearer to write that directly.

p6. l183. “This is likely due to the fact that wildfires emit BC and OC further aloft into the atmosphere and in the absence of down-mixing to the boundary layer such aerosol layers are mostly not captured by surface observations.” Any reference to support this statement? Indeed, the fire emissions vertical profile vs. transport dynamics is an important (open) question, for which it is a quite strong statement to say this is known as a fact (e.g. Remy et al., 2017, ACP or Ke et al., 2021, JGR).

p7. l198-199: I would suggest to add “summertime nss-SO4”, because the trend studies, both winter and annual averages, do exist and have also been comprehensively summarized in this work which makes this sentence slightly confusing. In general, nss-SO4 is a complicated “tracer” since its sources are both natural and anthropogenic. This contradictory is also reflected in the article introduction indirectly.

p10. l268. Figure 1 does not show data.

p11. l287. Add the detection limit here e.g. in parenthesis.

p15. L311. What do you mean with “prior deseasonalization”?

p17. l362. Could you be more specific with what is meant by “ammonium’s statistical distribution”?

p22. l431. Even though it’s difficult to compare the bars in fig2 only, but for Tiksi it looks like the absorption is much higher in comparison to Barrow and Alert (10-fold) than the EBC is (3-fold), and not really 1:1? Check.

p28 l 504-506. First sentence says that emissions rise in the 2000s. Second sentence says that EBC concentrations do not show decline, likely due to short measurement period and variability. Why is it assumed that a decline should be seen, especially if emissions increase?

p29. l529. What is “this” pattern? The pattern observed in Eastern Arctic countries, or elsewhere, or Asia..?

p32. l631. “They” should be “they”

p33. l638-641. Unclear sentence, rephrase.

p34. l663. Interesting that here Pallas shows negative SAE trend in summer while previous studies (Collaud Coen et al 2013; Lihavainen et al., 2015) have shown this trend for winter. But as pointed out several times in the manuscript, the trends are also sensitive to the method and with such short time series should be interpreted with caution.

p41. l850. Might consider adding that the aerosol microphysical properties also provide data on aerosol quantities that are directly connected with their climatic impacts.

p40-41 action items / conclusions: Conclusions and action items very strongly seem to highlight the need for additional monitoring and analysis of natural Arctic aerosols. Why is this emphasized? Or do you mean there is a lack of data on the aerosol organics, more specifically? In my view, the results presented support the need for intensified aerosol monitoring around the Arctic (due to complexity and variability of the sources), importance of long-term efforts (for understanding the trends) and need for interdisciplinary collaboration (for complexity of the sources and processes).

Tables:

Table 1. Sometimes both an article and ebas are indicated as data sources. Why? If an article is provided where the data are used and measurements described more in detail, it should be done systematically for every dataset, or any. Was the article (e.g. in case of Gilardoni, Dutkiewich, Heslin-Rees, etc.) really the data source or a personal communication? Does ebas provide EBC values for absorption measurements (Alert, Tiksi)? Using what MAC? For Alert, Barrow and Tiksi, what measured wavelength was used to convert data to abs coef @550 nm? In addition, would you be able to provide a link to specific datasets in ebas and access date, instead of referring to a general database? I understand if this last point is adding too much data in table, then current form is ok.

Figures:

Figure 1. Missing Pallas station.

Figure 2. Tiksi EBC concentrations (1^st panel) appears to be about 3 times higher than in Alert and Barrow, however, the absorption is 10-fold? Is also Tiksi scattering multiplied by 0.1? Color shading for grey and black are not possible to separate from each others.

Figure 4 (and corresponding text). Is western Arctic including all the continental territory of USA and Canada? And does Eastern Arctic include all the Russia? Is that overlapping with Asia? Reference to AMAP, 2021 is given, but this report does not have a reference yet so a short explanation here could be given. In addition, should NO2 be NOx?