Continuous new particle formation in a Mediterranean coastal environment: Insights from atmospheric ions behaviour analysis

Kalivitis, Nikos; Markoulakis, Spyridon-Emmanouil; Kalkavouras, Panayiotis; Kerminen, Veli-Matti; Kulmala, Markku T.; Kanakidou, Maria

doi:10.5194/ar-2026-16

Preprints

https://doi.org/10.5194/ar-2026-16

Preprints

22 Apr 2026

| 22 Apr 2026

Status: this preprint is currently under review for the journal AR.

Continuous new particle formation in a Mediterranean coastal environment: Insights from atmospheric ions behaviour analysis

Nikos Kalivitis, Spyridon-Emmanouil Markoulakis, Panayiotis Kalkavouras, Veli-Matti Kerminen, Markku T. Kulmala, and Maria Kanakidou

Abstract. Atmospheric new particle formation (NPF) is crucial for aerosol number concentration and for studying the production processes of secondary aerosol particles. NPF events are commonly classified based on visible particle growth, and their occurrence frequency is often underestimated. Recent methodologies propose that "quiet" NPF (QNPF) events, which are not traditionally classified as NPF events, can contribute significantly to particle number concentrations. This study presents three-years (June 2020–May 2023) of observations of ion and particle size distributions performed at the Finokalia environmental research station in Crete, Greece, using a Neutral cluster and Air Ion Spectrometer (NAIS) and a Mobility Particle Size Spectrometer (MPSS). By analysing the observed ion number size distributions and applying a nanoparticle ranking analysis method, this study reveals that QNPF events are frequent and contribute significantly to particle formation and growth in the Eastern Mediterranean. Negatively charged intermediate size ions are found to be reliable indicators of particle formation, including both classical NPF and QNPF episodes. Our analysis indicates continuous particle formation even on days traditionally classified as "non-event" days, providing fundamentally new understanding of NPF processes in the region.

Received: 02 Apr 2026 – Discussion started: 22 Apr 2026

Competing interests: At least one of the (co-)authors is a member of the editorial board of Aerosol Research.

Publisher's note: Copernicus Publications remains neutral with regard to jurisdictional claims made in the text, published maps, institutional affiliations, or any other geographical representation in this paper. While Copernicus Publications makes every effort to include appropriate place names, the final responsibility lies with the authors. Views expressed in the text are those of the authors and do not necessarily reflect the views of the publisher.

Download & links

Nikos Kalivitis, Spyridon-Emmanouil Markoulakis, Panayiotis Kalkavouras, Veli-Matti Kerminen, Markku T. Kulmala, and Maria Kanakidou

Status: final response (author comments only)

CC1: 'Comment on ar-2026-16', Ülkü A. Şahin, 06 Jul 2026

This manuscript presents three years of observations of ion and particle size distributions at the Finokalia station in Crete, Greece, using a Neutral Cluster and Air Ion Spectrometer (NAIS) and a Mobility Particle Size Spectrometer (MPSS) to determine the NPF events in the Eastern Mediterranean. The authors applied the newly proposed method to gain fresh insights into NPF events in the marine background of the Eastern Mediterranean and examined the role of ions in identifying and classifying NPF events. The topic is timely and relevant, as NPF determining methodology and NPF occurrence in megacities located at the crossroads of Europe, Asia, and the Mediterranean remain poorly documented, and scientists try to understand the NPF mechanism. Overall, the study is well designed, clearly written, and methodologically sound. The manuscript is well within the scope of Aerosol Research Discussion after the following minor revision.
The abstract should be revised to include the article’s original findings.
The introductory text should include more up-to-date sources, and the regional NPF mechanism should be discussed in greater detail in the paragraph on line 45.
"..of the.." in Table 1 caption should be removed.
In line 145: Why did the authors select different time intervals for active and background in the four seasons? Is it possible to formulate the ranking analysis method to understand clearly?
What is the methodology of defining the cotegory of "undefined NPF"? 21% percent is need to discuss, because higher level comparing the literature.
Between lines 155-165: The results of recent studies carried out in the region should be included in the discussion in this section. The region’s unique status in terms of NPF formation should be discussed.
The seasonalty is very important on forming NPF events. So the seasonal level of NPF frequency should be added in Table 2.
It is not necessary to write "..by Dal Maso et al. (2005)" in the figure 1 caption.
Line 183: "A secondary maximum is found in October (48%), probably due to the onset of the wet period and the decrease in CS". Is there any evidence to support this finding? Such as rainfall data and also evidence that wet period increases NPF.
Section 3.1 needs further discussion with regional NPF events through the seasonality etc.
The conclusion may include the limitations and further study suggestions.

Citation: https://doi.org/10.5194/ar-2026-16-CC1
RC1:
'Comment on ar-2026-16', Anonymous Referee #1, 24 Jul 2026
The authors report insights from atmospheric ions behaviour analysis stemming from observations of three years of ion and particle size distributions at the Finokalia station in Crete, Greece, using a Neutral Cluster and Air Ion Spectrometer (NAIS) and a Mobility Particle Size Spectrometer (MPSS). With this approach, the authors very nicely conclude new methods to study and observe NPF events. In addition, they suggest a fundamental change in how NPF events should be classified. Hence, the new scientific findings and well conducted study is very important for the aerosol community. The text is well written and describes the main findings in an understandable way with valuable new data. So, I recommend that this manuscript should be published in Aerosol Research.
Nonetheless, I propose some minor revisions related to the result section in order to facilitate the readability of the presented result. This can be done by revising and connect some of the Figures to visually and schematically improve the storyline of the manuscript. I would recommend to also link the text better to the Figure description. Please see my detailed comments below:
abstract
line 24: sentence stop is missing.

introduction
line 34: the authors did not introduce 'CCN' yet

line 44: ok, here the authors outline what frequently means on Crete. maybe add this number to the abstract

line 46: what is meant by 'undefined days'? - the authors define it only on line 52, I see. Please try to move this above and maybe in this way it is possible to write it in one section. No line break on line 44...

Line 52: please use AE for 'fulfil' --> fulfill

Line 54: is growth rate usually abbreviated? It looks a bit weird to me. Also, later, the authors don't use it anymore (e.g. line 75). Hence, I would omit it.

Line 61: use 'that' or 'diffusing' instead of which. In this way you don't need a comma

line 67: please rephrase sentence to: 'Intermediate ions, which are usually present at low concentrations, can experience pronounced increases under certain atmospheric conditions, for example during heavy precipitation events or NPF episodes (Tammet et al., 2014; Leino et al., 2016).'

line 71: please rephrase. For instance: '... non-event days, which are termed...'

line 76: please explain in more detail what you mean by 'relatively similar', it reads very vague like that.

Line 80: please double-check if 'eastern' is capitalized or not. The authors have been using both ways throughout the text.

Last section of 'Introduction': may the authors extend the discussion and summary of your study a bit more? It is still not so easy to grasp what new results will be shown in the main text. Elaborate it in more detail.

material and methods
line 85: shouldn't it be 'Ions'?

line 87: please write out 'ACTRIS'

line 98: missing comma before 'which'

line 100: the authors need to specify 'DMA'

line 112: please either write 5 min or five minutes

first paragraph in section 2.3 starting on line 114: it could be helpful for the reader if the authors summarize the NPF classifications in a table or schematic.

line 123: what is 'di' and 'dj'?

line 124: I don't think the authors used abbreviation 'GMD' again, so please delete

results
line 145: Delete one 'of the' in the caption of Table 1.

line 166: Caption of Figure 1: please mention again for how many years the data is representing. And one additional idea: could the authors integrate Table 2 somehow schematically into Figure 1? This would facilitate the readability a lot. If so, the authors would not need Table 2 anymore.

line 189: Are the authors aware of any reason for the differences reported in Pikridas et al., 2012? Please elaborate a bit more on that.

line 195: Why not asking this question already in the last section of the Introduction? Or even in the abstract? This would help to guide the reader through the presented research questions, which is now still a bit missing.

line 203: 'a' instead of 'an' NPF event

line 224: please rephrase the sentence starting with 'These results...' or make two sentences. Right now it is hard to follow.

line 227: delete the comma after NPF

line 229: please introduce Figure 4 first in one sentence before talking about it as the authors are doing right now. The reader has not seen Figure 4 yet.

line 236: Caption of Figure 3: please here again declare, which time period is represented in the wind rose diagram.

line 254: Figure 4: Is there no uncertainty for the Radon concentration from the measurements?

line 264: delete comma after 0.8 mm

line 269: delete comma after 'averaged'

line 270: delete comma after 'days'

line 271: please rephrase the sentence starting with 'Normalizing...'

line 274: delete comma after 'observed'

line 289: Figure 5: please be consistent with the hyphen between non-events as you do in the text and caption. Also, please label the right y-axis for all the plots and color bars (left is number concentration?) even though it is only values without units. You could also enlarge all the subpanels for better visibility and if possible put the subpanel labeling (a,b,...) into the panels. In this case the left y-axis label 'diameter' is better visible. Now it looks as if the panel labels correspond to the x-labelling. In addition, the small difference in average growth rate as the authors mention on line 274 is not visible in the panels. Can you mark this somehow?

line 293: caption of Figure 5: a comma is missing before 'respectively', the same on line 295

line 304: please use consistently Oxford spelling as you do here: 'characterize'. But at some place you used 's' instead of 'z'. I copy here the copernicus English guidelines: '- Variety of English: all standard varieties of English are accepted in order to retain the author’s voice. However, the variety should be consistent within each article. For an international readership, Oxford spelling using -z- variants instead of -s- (e.g. characterize) is often utilized. When using Oxford spellings, please do so consistently. For example, if "characterize" is spelled as such, then the -z- variant should be used for all such words throughout the article. The use (or lack thereof) of the Oxford (serial) comma should also be consistent. For more information, please see the Wikipedia pages about Oxford spelling and Oxford (serial) commas.' Another example: 'Normalised' vs 'normalized' in caption of Figure 7. The authors used once 's' and once 'z' in one sentence.

Figure 6: ok, I see that the color bar must be ion number size distribution. Please adjust in Figure 5. And please be consistent with using Time [h] or Time_[UTC] [h] in the x-axis label

line 329: Figure 8. please delete the Figure title. This information should be apparent from the caption. Also for the legend, please be consistent with labeling: non-event instead of noevent and there is a wrong Tab before CLASS II. What are the vertical grey lines? Also the x-axis should be linear, right? The second tick (10% ?) and the second last (90% ?) look offset.

Figure 6 and 7: I don't find it straight forward to see the midday formation in the 25-30% interval as you mention on line 318. May you precise this and maybe add a line or pattern into the 25, 30 panel? The authors could also try to emphasize on that one by enhancing this panel and point it out more clearly. This would really improve readability and the storyline. In this case you could also better link Figure 8 to Figure 7. You could clearly note in the Figure 8 at which percentile rankings the subpanels in Figure 7 correspond to. This would also make the result section more concise. So, please restructure it accordingly.

Figure 9: the dots are not represented in the legend, please adjust.

conclusion: may the authors come back in more detail to your stated research question on line 195? like mention more precisely what the authors found and state the numbers for the sizes and polarities being optimal for identifying a NPF event.
Citation: https://doi.org/10.5194/ar-2026-16-RC1
RC2:
'Comment on ar-2026-16', Anonymous Referee #2, 13 Aug 2026
This paper describes the analysis of three years of NAIS data from the Finokolia site in the eastern Mediterranean. The goal of the analysis is to use size distributions of intermediate ions to identify new particle formation. That is a worthy goal and would be a valuable addition to the literature. However, the paper is very confusing. It is difficult to follow the logic or to understand what the final conclusions are. It needs to be revised before publication. Specific comments follow.
Lines 70-79. Please add a bit more discussion about how QNPF is identified and interpreted in previous work. If I understand correctly, it is based on averaging PNSDs over many non-event days and seeing a signature of particle growth in the normalized data. I am not sure it makes sense to talk about “QNPF events” or a “typical QNPF day” if it is only observed in averages over many days. It is also the case that while the growth rates are similar, the formation rates are much lower for QNPF than for NPF events (e.g., in Kulmala et al., 2022). Also, please explain what the “radical transformation in NPF studies” means. Please update this section to more accurately reflect the previous literature. This will provide a better context for the current results.

Section 3.2. It is not clear why the authors are plotting 1.87 to 2.88 nm in Figure 2 when the ranking method is based on 2 to 5 nm ions and the intermediate size is 1.6 to 7.5 nm. Please explain why these particular sizes were chosen. Why average over all available measurement days? Wouldn’t it be more to the point to show the average for NPF days and non-NPF days to show that they are different and therefore these sizes can be used to identify NPF? What is meant by the 95% confidence interval and how is it related to the statement earlier that the measurement of 3 nm negative ions was 80% low compared to a reference instrument? Why not show the standard deviation which would show the variation in the dataset?

Lines 221 to 234: Why 2-3 nm ions for this analysis? Why include south but not north in the wind rose analysis? North has slightly more of a contribution than south. Why not just discuss west and southwest? What is meant by locally produced small ions and how would the authors identify them? I am not clear on what this figure and discussion adds to the main points of the paper.

Section 3.3. In Figure 5, normalizing the concentrations gives a hint of a growth curve for the average of all non-NPF days but the data is very noisy. It looks reasonable that the growth rate is similar between NPF and non-NPF, since the shape of the curve is similar, but it seems unlikely that the formation rate is higher for non-NPF. In fact, Figure 9 shows that the starred points (Class I NPF) have a much higher average formation rate than the dark circles (lower rank and more likely to be non-NPF). The authors state that the formation rate is only calculated for Class I events (line 125). How are they getting a formation rate for non-NPF days? Later, they state that the formation rate is only calculated for 200 of the 775 days (line 337). For what fraction of non-NPF days can they calculate a formation rate? If it is only a small fraction of non-NPF days, then they might be biasing the formation rate high. The authors need to assign error bars in Table 3 and perhaps tone down the discussion.

Section 3.4. Does NPF occur continuously (line 304)? The ranking data suggests not since there is no indication of midday formation below the 25^th percentile in Figure 7. In fact, the ranking data suggests that calling all non-event days QNPF is not correct since a large fraction exhibit no particle formation. Same comment about the conclusions - lumping all non-event days into NPF is not consistent with the ranking data. Instead, the ranking data provides an opportunity to refine the QNPF concept. The authors should revise the discussion to reflect what the ranking data reveals.

Could the authors make a bar graph for monthly non-NPF and NPF using below and above the 25^th percentile as the cutoff? Does it show the expected seasonal cycle? This would be a useful confirmation that the ranking method reflects the atmosphere.

Minor comments:
Please use consistent terminology throughout the paper, e.g. either non-event or non-NPF. Or maybe non-event means non-NPF plus undefined? It is not clear.

Section 2.2 is missing information. Lines 95-96: What is the difference between the inlet tubing and the sampling lines? Lines 101-102: What is a reasonable amount? Line 103: Presumably negative particle mode? Line 104: What is an acceptable range?

Please define the size for nucleation mode (lines 115-116). Is this observed in the particle size distribution or the ion size distribution? Please define the size of the nucleation range (line 122).

Figure 1. These colors are difficult to distinguish here and even more difficult to distinguish in Figure 8. Please use more highly contrasting colors in Figure 1 and then use the same colors for the same classifications in Figure 8. I would label the months with names. Since the data runs from June to May, it is not clear if Month 1 is June or January.

Please use the same timebase for all figures. Some are labeled UTC and some are not. If they are all UTC, then label them that way.

Figure 9. Caption says daily formation rate but y-axis of graph is per second. Please use the same units for formation rates throughout the paper.

Issues with language:
This paper needs to be carefully edited for clarity and proper English usage. Here are a few examples of poor usage. Words like “really” (line 65) and “actually” (line 72) do not belong in a scientific journal article. “Cluster’s” (line 32) is awkward when “cluster” is sufficient. “afresh” (line 32) is misspelled. “In the island” (line 46) should be “On the island.” Word missing in “extensively approach” (line 114). Delete one of “used” or “combined” (line 144). There is no need to state a major research question in quotes (line 195), especially since the authors do not answer the question. The abstract and introduction should make it clear what the authors want to learn.
Citation: https://doi.org/10.5194/ar-2026-16-RC2

Nikos Kalivitis, Spyridon-Emmanouil Markoulakis, Panayiotis Kalkavouras, Veli-Matti Kerminen, Markku T. Kulmala, and Maria Kanakidou

Viewed

Total article views: 424 (including HTML, PDF, and XML)

HTML	PDF	XML	Total	BibTeX	EndNote
277	115	32	424	29	41

HTML: 277
PDF: 115
XML: 32
Total: 424
BibTeX: 29
EndNote: 41

Views and downloads (calculated since 22 Apr 2026)

Month	HTML	PDF	XML	Total
Apr 2026	83	45	5	133
May 2026	106	24	10	140
Jun 2026	26	7	2	35
Jul 2026	51	38	12	101
Aug 2026	11	1	3	15

Cumulative views and downloads (calculated since 22 Apr 2026)

Month	HTML	PDF	XML	Total
Apr 2026	83	45	5	133
May 2026	106	24	10	140
Jun 2026	26	7	2	35
Jul 2026	51	38	12	101
Aug 2026	11	1	3	15

Viewed (geographical distribution)

Total article views: 378 (including HTML, PDF, and XML) Thereof 378 with geography defined and 0 with unknown origin.

Country	#	Views	%

Latest update: 20 Aug 2026

Short summary

Atmospheric new particle formation (NPF) is crucial for aerosol number concentration. Recent methodologies propose that “quiet” NPF (QNPF) events, which are not traditionally classified as NPF, can contribute significantly to aerosol number. This study presents three-years of observations of ion and particle size distributions at the Finokalia station in Eastern Mediterranean. It shows that QNPF events are frequent providing fundamentally new understanding of NPF processes in the region.


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