the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 17 Sep 2026
Cloud processing signatures in boreal forest aerosol revealed by long-term in-situ observations
Abstract. Chemical cloud processing has the potential to substantially modify atmospheric aerosol properties and influence air pollutant lifetimes, yet these processes remain challenging to empirically quantify from observations. Previous observational evidence have largely relied on short-term aircraft campaigns, providing only episodic snapshots of cloud–aerosol interactions. Here, we demonstrate that continuous ground-based observations at the SMEAR II boreal forest station enable the identification and characterization of cloud-processed aerosols over multi-year timescales. We show that bimodal aerosol number size distributions exhibiting a pronounced local minimum between the Aitken and accumulation modes –commonly referred to as the Hoppel minimum– are closely linked to the presence of clouds over the boreal forest. These bimodal distributions serve as efficient proxies for cloud-processed aerosols, allowing their physicochemical properties to be examined using long-term in-situ data. This approach provides a novel framework for studying cloud processing effects as an important, long-term complement to aircraft-based observations. Our analysis reveals that cloud processing leads to a clear enrichment of sulfate aerosol mass in the accumulation mode, while impacts on organic aerosol mass concentration and chemical composition remain insignificant. The observed impacts of cloud processing on aerosol composition are consistent with previous estimates derived from trajectory-based analyses, while extending them to a statistically robust, long-term observational context.
Competing interests: Some authors are members of the editorial board of Atmospheric Chemistry and Physics.
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.- Preprint
(3874 KB) - Metadata XML
-
Supplement
(2023 KB) - BibTeX
- EndNote
Status: open (until 29 Oct 2026)
- RC1: 'Comment on egusphere-2026-5082', Anonymous Referee #1, 11 Oct 2026 reply
-
RC2: 'Comment on egusphere-2026-5082', Anonymous Referee #2, 18 Oct 2026
reply
General comments:
Heikkinen et al. use long-term observations from a boreal forest site in Finland to investigate cloud-processing signatures, using Hoppel minima in PNSD as a proxy for cloud-processed aerosol populations. The study provides evidence that cloud processing may be associated with sulfate enrichment in the accumulation mode and discusses possible mechanisms underlying changes in OA. The long-term dataset allows the authors to examine these relationships across different seasons and meteorological conditions, providing a useful framework for investigating cloud-processing effects. However, several methodological issues and uncertainties in the interpretation of aerosol chemical composition need to be addressed. I therefore recommend major revision.Specific comments:
1. The study is based on eight years of observational data from 2012–2020, so the dataset is large. However, after applying k-means clustering, nearly half of the PNSD_Hoppel data were removed. The remaining data were then divided into PNSD_Hoppel and PNSD_Other. Is the final sample size still representative? Do the seasonal distribution, chemical composition, or meteorological characteristics of the remaining samples still support the results and speculations of this paper? The authors stated that k = 7 gives the best clustering performance. Please show or justify this conclusion. Also, what is the basis for choosing a silhouette threshold of 0.8?2. The description of the Wilcoxon rank sum test is relatively short and brief. If the comparison is based on hourly data, adjacent hours may be affected by the same air mass and thus be correlated. Given that the study uses a criterion of p < 0.05 (95% confidence interval) and compares multiple chemical composition variables, is this sufficient to support the conclusions about significance levels? Are there other auxiliary methods or parameters that can demonstrate the reliability of choosing this method?
3. In this study, chemical composition is obtained from ACSM, and the authors note that sulfate and OA (considered to be mainly SOA present as OOA) are the main components of the aerosol. I am curious about their specific mass fractions, or the range over which they vary, and also the mass fractions of nitrate, ammonium, chloride, and BC. According to Fig. 6, the sulfate fraction can reach 30–40%. The forest can emit abundant BVOCs. Have the authors considered the possibility that the total sulfate includes organosulfates? For example, SOA from isoprene may contain organosulfates. At the same time, nitrate also has strong hygroscopic ability; could it also be affected by cloud processes? The authors also mention organic nitrates. It is recommended to strengthen the discussion of uncertainties in chemical composition in the current analysis. In addition, the current study directly uses the model results of size-resolved chemical composition obtained by Ranjan et al. (2025). Please provide the uncertainty of the model and its sources. Moreover, Line 223 clearly states that the model results cover the years 2016-2020. Is there a mismatch in time span with the 2012–2020 period of this study? Did the authors extend the model simulation to obtain data for 2012-2020?
4. Other detailed issues: 1) DMPS and monoterpene (PTR-MS) measurements were made at a height of 8 m, while other gaseous pollutants such as NOx and O3 were measured at 67.2 m, and the measurement height for ACSM is unknown. Is it reasonable to use gaseous pollutants with such large sampling height differences in the analysis in this study? How much uncertainty might this introduce? 2) Line 279 mentions using RH > 94% as a threshold. What is the supporting evidence or literature source for this threshold?
Minor comments:
1. The Introduction is a bit too long. Please shorten it appropriately.2. Many self-citations indicating methods can be omitted in the text, especially in figure captions, such as “(see Sect. 2.1.5)” in Fig. 1.
3. The numbers in Figs. 11–14 are small and difficult to read, especially the gray numbers. Please optimize them.
Citation: https://doi.org/10.5194/egusphere-2026-5082-RC2
Viewed
| HTML | XML | Total | Supplement | BibTeX | EndNote | |
|---|---|---|---|---|---|---|
| 494 | 202 | 70 | 766 | 82 | 64 | 62 |
- HTML: 494
- PDF: 202
- XML: 70
- Total: 766
- Supplement: 82
- BibTeX: 64
- EndNote: 62
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1
By analyzing long-term observational data from 2012–2020 at the SMEAR II station in the boreal forests of Finland, the study found that bimodal particle size distributions with a Hoppel minimum can serve as a robust microphysical proxy for cloud-processed aerosols. Cloud processing significantly enhanced the sulfate mass fraction and concentration in the accumulation mode, but produced no detectable net enhancement in the mass or oxidation degree (O:C) of organic aerosols. The core innovation lies in being the first to use the microphysical fingerprint of long-term ground-based size distributions (rather than short-term aircraft campaigns or single chemical tracers) to systematically identify and statistically characterize the cloud-processed aerosol population. However, the paper still has some issues that need to be addressed before it can be accepted.
（1）The Introduction covers sulfate cloud chemistry and the monoterpene/isoprene aqSOA controversy fairly comprehensively, the evidence regarding SOA formation during cloud processes is summarized insufficiently, and the observations from major alpine sites worldwide are not covered.
（2）The Results and Discussion section contains quite a lot of redundant information, which interferes with the core content of the paper. The first few result figures of the paper (Figures 2–4) sequentially present: time series and seasonal variation of bimodal frequency, average PNSD spectra in different seasons, and comparisons of meteorological conditions (T, RH, BLH, WS) between the two PNSD classes. These are only supplementary pieces of information, and they do not require extensive descriptions.
（3）A considerable portion of the paper's Discussion is devoted to explaining the mechanisms of sulfate enhancement (O₃/H₂O₂ oxidation, metal catalysis, etc.). While this part is reasonable, it is already well supported by the literature and is not a new contribution of this paper. Yet the question that truly deserves in-depth discussion—why, over such a long observation period and at such a typical boreal forest site, is there no detectable cloud-processing signal in organic aerosols?—receives only limited space and remains at the qualitative level of "may be offset." Does averaging all the data mask some signals?
（4）The evidence for "no significant enhancement" of organic aerosols is relatively weak: ACSM measures only non-refractory PM₁; the f44→O:C approach works for highly oxidized organics but is insensitive to semi-volatile species, nitrogen-containing organics, and aqueous-phase products with low f44. Even if cloud-water SOA forms at monoterpene-dominated sites, it may be simultaneously offset by reduced low-temperature sources, boundary layer dilution, declining O₃/OH, and wet deposition. This paper does not use aqueous-phase markers such as HMS, oxalic/glyoxylic acid, isoprene products, or organosulfur compounds as direct evidence of "reaction occurred but was offset"; concluding "no net enhancement" based solely on O:C similarity is somewhat unreliable.
(5) The ACSM measures only non-refractory PM₁ components, but particles after cloud processing may grow into the supermicron range, causing some mass to be underestimated. The paper does not discuss the systematic impact of this size cutoff on the comparison of sulfate and OA concentrations.