Contrasting Inland-Coastal Aerosol Mixing States: An Entropy-Based Metric for CCN Activity

Ren, Jingye; Xu, Wei; Huang, Ru-Jin; Zhang, Fang; Wang, Ying; Chen, Lu; Ovadnevaite, Jurgita; Ceburnis, Darius; O’Dowd, Colin

doi:https://doi.org/10.5194/egusphere-2025-3284

Preprints

https://doi.org/10.5194/egusphere-2025-3284

Preprints

15 Jul 2025

| 15 Jul 2025

Contrasting Inland-Coastal Aerosol Mixing States: An Entropy-Based Metric for CCN Activity

Jingye Ren, Wei Xu, Ru-Jin Huang, Fang Zhang, Ying Wang, Lu Chen, Jurgita Ovadnevaite, Darius Ceburnis, and Colin O’Dowd

Abstract. Simplified assumptions of aerosol hygroscopic mixing states in modeling studies often introduce substantial uncertainties in estimating cloud condensation nuclei (CCN) concentrations and their climatic impacts. This study systematically investigates the contrasting relationships between mixing states and CCN activity by comparing ambient measurements from inland and coastal sites. We show distinct seasonal variations of the particles mixing state. In winter, externally mixed particles dominated both sites, with comparable mixing state indices (χ) of 0.38±0.12 and 0.39±0.09 respectively for coastal air masses and inland air. However, summer measurements showed pronounced differences: photochemical processes promoted significantly higher internal mixing in coastal aerosols (χ=0.69±0.19), whereas inland χ values only increased moderately to 0.47±0.12. A universal logarithmic correlation was identified between the critical diameter (D_cri) characterizing CCN activity and χ (D_cri = -32.15ln(χ)+84.71, Pearson r = -0.74), but with distinct decrement rates for coastal vs. inland aerosols. Our further quantitative analysis reveals a 0.1 increase in χ enhanced winter CCN concentrations (N_CCN) by 39–65 % under typical cloud supersaturations, whereas this effect diminished to ~9 % in summer. These results underscore that mixing states exert more pronounced control over N_CCN in diverse environments. Our work provides critical constraints for parameterizing fine aerosols CCN activity in climate models, thereby reducing uncertainties in aerosol–climate effect estimations.

Received: 09 Jul 2025 – Discussion started: 15 Jul 2025

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

Preprint (PDF, 1475 KB)

Supplement (1193 KB)

Download & links

Jingye Ren, Wei Xu, Ru-Jin Huang, Fang Zhang, Ying Wang, Lu Chen, Jurgita Ovadnevaite, Darius Ceburnis, and Colin O’Dowd

Status: final response (author comments only)

RC1:
'Comment on egusphere-2025-3284', Anonymous Referee #1, 09 Aug 2025
General comments:
This study aims to investigate aerosol mixing state characteristics at two sites—one inland and one coastal—and to examine their impacts on CCN activity. The comparisons and coastal measurements presented here indeed reveal some interesting features that could potentially offer new insights. However, the current manuscript contains numerous misleading statements and conclusions, as detailed in my major and specific comments.
The most important concern is that the key findings rest on conceptual flaws that could mislead less-experienced readers, such as students. Before explaining the reasons, the definition of “mixing state” must be clarified.
If the authors are quantifying the mixing state index for the entire aerosol population, the term refers to the distribution of aerosol mass among different particles of different sizes, under the condition of fixed bulk mass fractions of aerosol components. This implies that overall aerosol hygroscopicity is fixed (assuming the volume mixing rule applies).
If, instead, the authors are quantifying the mixing state index for aerosols of different diameters, the term refers to the distribution of aerosol mass among particles under fixed aerosol size distribution and size-resolved chemical composition.
In models, when both aerosol size distribution and size-resolved chemical composition are known, the mixing state refers to how aerosol mass is distributed across particles of the same size, corresponding to χ variations of different sizes under fixed size distributions and size-resolved compositions. Thus, when discussing the impact of mixing state, it should be done under fixed aerosol size distributions and bulk aerosol compositions (or ideally, fixed size-resolved compositions).
With this in mind, my following points are justified. Theoretically, CCN activity depends primarily on aerosol size distribution, with hygroscopicity playing a secondary role. Moreover, the role of mixing state—could be represented by the hygroscopicity distribution—is generally smaller than that of the overall hygroscopicity. In this study, the authors parameterize the critical activation diameter (D₍cri₎), which is determined by aerosol hygroscopicity and supersaturation, with the mixing state index χ. The reported relationship between D₍cri₎ and χ does not isolate the impact of mixing state; it simply reflects their co-variation. Higher χ often corresponds to higher internal mixing, which is mainly driven by secondary aerosol formation and thus higher hygroscopicity. Therefore, χ can correlate with D₍cri₎, however, without being its causal driver.
To genuinely examine the impact of mixing state on CCN activity, the authors should test scenarios with fixed aerosol size distribution and fixed aerosol composition (so that overall κ is unchanged). Variations in D₍cri₎ could then be explored under different combinations of κ (derived from composition measurements) and χ. A 2-D plot with κ on the x-axis and χ at ~150 nm on the y-axis, followed by analysis of D₍cri₎ variations, could yield more robust insights.
Statements such as:
“Mixing state impacts on N₍CCN₎ are most pronounced during winter in both environments, attributed to heightened winter D₍cri₎ sensitivity to χ: a 0.1 χ increase reduces D₍cri₎ by 5.2% (winter), boosting N₍CCN₎ by 39%, versus 2.4% D₍cri₎ reduction (summer) yielding only 6% N₍CCN₎ enhancement. Concomitantly, winter N₍CN₎–χ effects on N₍CCN₎ reach 65%, far exceeding summer responses.”
should be revised. Correlations only indicate co-variations and cannot be directly interpreted as causal impacts of χ on CCN.
In summary, the current findings do not provide critical constraints for parameterizing fine-aerosol CCN activity in climate models, nor do they reduce uncertainties in aerosol–climate effect estimates. The discussion does not bring genuinely new insights into CCN parameterization. I recommend major revisions, with a stronger focus on how secondary aerosol formation affects hygroscopicity distribution and thus χ. Section 3.1 and 3.2 analyses are preliminary. I suggest revising:
Section 3.2 → “Impacts of Primary Aerosol Emissions and Secondary Aerosol Formation on Aerosol Mixing State”

Section 3.3 → “Impact of Mixing State on CCN Activity” (with the role of mixing state isolated as described above)

Major Comments
Title and site representation – Can the inland–coastal contrast be represented solely by observations at two sites? Aerosol aging differs substantially across different continental and marine locations. Observations from two inland or two coastal sites could also yield contrasting characteristics. I Suggest consider a more neutral title, e.g., “Contrasting Aerosol Mixing States at Two Inland and Coastal Sites: An Entropy-Based Metric for CCN Activity”. In the text, I suggest using “IAP–winter” instead of “inland–winter” and “MH–winter” instead of “coastal–winter,” while placing broader inland–coastal implications in the conclusion.

Sampling period – The short observation period, especially the two-week inland summer dataset, weakens the robustness of the “contrasting” conclusion. For example, Beijing aerosol properties in June may differ significantly from those in August. The representativeness of these periods should be at least discussed.

Relationship between hygroscopicity distribution, mixing state, and CCN activity – This needs to be explained more clearly. A straightforward logical framework would help (on the basis of my comments on mixing state at the very beginning). Actually, for DMA-CCN measurements, the size-revied AR could be fitted using the formula proposed by Rose et al. (2008) which have three key parameters including the Da (critical activation diameter), MAF (Maximum Activation fraction) and the heterogeneity parameter σ. Among three parameters, σ is mostly affected by the mixing state, or so called heterogeneity, and MAF would also be affected by mixing state, especially mixing state of black carbon, while other hydrophobic components also matter (Tao et al., 2024). Authors should examine relationships between σ and χ, as well as MAF and χ. Those analysis and discussions would help and reflects impacts of mixing state on CCN activity, may be a relationship between σ and χ could be revealed.

Introduction – Coastal aerosols cannot be assumed to represent marine aerosols. The introduction could follow this logic: measurements in coastal regions can provide insights into marine aerosol properties, which differ markedly from continental aerosols and have distinct climate impacts.

κ-grouping method – The method of Yuan et al. (2023) is generally valid, but the grouping of κ=0.01 and κ=0.6 might still be improved. κ=0.01 corresponds to non-hygroscopic species such as external BC and nearly hydrophobic organic aerosols (mostly primary OA). The more hygroscopic group should contain mostly secondary organic and inorganic aerosols. At RH of 90%, κ of ammonium sulfate and nitrate is ~0.5, while SOA κ is clearly lower. Could the authors test κ_H settings for continental aerosols, as was done for marine aerosols? If κ=0.6 remains the choice, please justify.

Specific comments:
L31 make it clear, the mixing state index here is calculated based on what kind of measurements
L38-40, under what levels of supersaturations (concrete value), the supersaturations typical of clouds differ for different cloud types, even for certain kind of cloud, the supersaturation vary in a wide range depending on many factors.
L55, determine particle hygroscopicity distribution
L57 This statement is vague and misleading for researcher who does not understand quite well about aerosol activation. Rely on what parameters of inorganic components, size or hygroscopicity? Hygroscopicity of inorganic components of inorganic aerosols is close for ammonium nitrate and sulfate and fixed. Therefore, for internal case, still rely on aerosol size and hygroscopicity, with the hygroscopicity determined by organic aerosol hygroscopicity and organic aerosol fraction.
Therefore, for internal-mixed aerosols, the CCN activity rely on overall aerosol size, organic aerosol hygroscopicity and organic aerosol mass fraction.
For external-mixed aerosols, the CCN activity (activation fraction) also rely on overall aerosol size, organic aerosol hygroscopicity and organic aerosol mass fraction which determines the number fraction of organic aerosols.
Indeed, CCN activity depends highly on mixing state. However, for this case, CCN activity of both external- and internal-mixed aerosols depend on the organic aerosol hygroscopicity and organic aerosol mass fraction but in a different way.
Therefore, try to state in a clear way.
L66-69, reactions during night does not significantly impact on aging thus CCN activity of aerosols? for example, heterogenous nitrate formation or multiphase sulfate formation? Why authors directly refer to photochemical reactions
L71, Is this conclusion universally valid, if not, please revise as “might create unique mixing state”,
L72-74, again, night time reactions in coastal regions is not important?
L75-76, References here are not appropriate. References here demonstrate continental aerosols and marine aerosols (mostly marine coarse aerosols) have distinct climate impacts. Marine aerosols could not be represented by coastal aerosols. As authors stated, coastal aerosols are influenced by both continental and marine air mass.
L77 , both continental and coastal aerosols could only impact on regional cloud formation. Impacts of all types of aerosols on cloud formation through serve as CCN is regional.
L77-79, statement here should be weakened and might be misleading, aerosols at coastal regions only impact clouds near coast both on costal continental clouds and marine clouds. It could not impact on marine clouds in the vast ocean.
L84, Zhao et al. (2021) is an important and pioneer paper in China about this issue and should be cited, the history of the entropy philosophy should also be included, for example, Riemer and West (2013).
L144-147, CCN measurements conducted at what levels of supersaturations.
L226-227, this speculation was based on what evidence? Please include more discussions
L248-251, The limited measurements could suggest conclusions across seasons? Please explore more the relationships between diurnal variations of accumulation mode aerosol hygroscopicity distribution and aerosol chemical composition evolutions and put it into the context of existing literatures.
L249, I did not see pronounced diurnal cycles especially for Mace Head measurements, how to defined pronounced? Please use objective statement, for example “ diurnal variations of xx is shown in Fig.x”
L265, the “spread factor” is what? Please make it clear
L267-268, biogenic origin is inferred from what clue? At least add the reference
L276 to 278, this speculation is not convincing at all. The unimodal distribution peaked at gf of ~1.7, meaning that the kappa peaked at ~0.5, the secondary formation of organic aerosol as demonstrated by the cited reference Jimenez et al. would substantially reduce the aerosol hygroscopicity, which is contrast with the average distribution here. The results shown in Fig.3d also demonstrate that the mixing state index would decrease as PM increase with increase organic aerosol fraction, authors should plot the average Kappa distribution under different PM levels at the Mace Head, and authors might find that the unimodal distribution prevail only for small PM conditions in summer.
I guess that the unimodal distribution in summer is due to the marine air mass not because of the aging, the aging in summer would undermine the unimodal distribution.
L286, what kind of aerosol properties
L289, reference (Ren et al., 2018) is not appropriate, reference such as Dusek et al. (2006) serves better
L291 Figure captions of Figure 5 is not clear,
L294-296, speculation, how could you attribute to new particle formation? I did see the NPD characteristics embedded in the PNSD of IAP summer
L298, only mass fraction of water-soluble components? Depend on aerosol hygroscopicity which is determined by hygroscopicity and mass fractions of water-soluble components.
L299, typical of what types of cloud? revise “Using a typical cloud supersaturation of 0.2% as a case study” as “Using measurements at supersaturation of 0.2% as an example”.
L300 “decreases with increasing highly hygroscopic inorganic aerosol components (e.g., sulfate, nitrate)”, the key point is inorganic aerosol components with high hygroscopicity, not water soluble. Many water-soluble substances have small hygroscopicity due to high molecular weight and small van’t Hoff factor. By the way, water soluble might correspond to very small hygroscopicity (Chen et al., 2019).
L315, the dominant role of aerosol size on CCN activity should cite the paper of Dusek et al. (2006).

Chen, J., Lee, W.-C., Itoh, M., and Kuwata, M.: A Significant Portion of Water-Soluble Organic Matter in Fresh Biomass Burning Particles Does Not Contribute to Hygroscopic Growth: An Application of Polarity Segregation by 1-Octanol–Water Partitioning Method, Environmental science & technology, 53, 10034-10042, 10.1021/acs.est.9b01696, 2019.
Dusek, U., Frank, G. P., Hildebrandt, L., Curtius, J., Schneider, J., Walter, S., Chand, D., Drewnick, F., Hings, S., Jung, D., Borrmann, S., and Andreae, M. O.: Size Matters More Than Chemistry for Cloud-Nucleating Ability of Aerosol Particles, Science, 312, 1375-1378, 10.1126/science.1125261, 2006.
Riemer, N., and West, M.: Quantifying aerosol mixing state with entropy and diversity measures, Atmos. Chem. Phys., 13, 11423-11439, 10.5194/acp-13-11423-2013, 2013.
Rose, D., Gunthe, S. S., Mikhailov, E., Frank, G. P., Dusek, U., Andreae, M. O., and Pöschl, U.: Calibration and measurement uncertainties of a continuous-flow cloud condensation nuclei counter (DMT-CCNC): CCN activation of ammonium sulfate and sodium chloride aerosol particles in theory and experiment, Atmos. Chem. Phys., 8, 1153-1179, 10.5194/acp-8-1153-2008, 2008.
Tao, J., Luo, B., Xu, W., Zhao, G., Xu, H., Xue, B., Zhai, M., Xu, W., Zhao, H., Ren, S., Zhou, G., Liu, L., Kuang, Y., and Sun, Y.: Markedly different impacts of primary emissions and secondary aerosol formation on aerosol mixing states revealed by simultaneous measurements of CCNC, H(/V)TDMA, and SP2, Atmos. Chem. Phys., 24, 9131-9154, 10.5194/acp-24-9131-2024, 2024.
Zhao, G., Tan, T., Zhu, Y., Hu, M., and Zhao, C.: Method to quantify black carbon aerosol light absorption enhancement with a mixing state index, Atmos. Chem. Phys., 21, 18055-18063, 10.5194/acp-21-18055-2021, 2021.
Citation: https://doi.org/10.5194/egusphere-2025-3284-RC1
RC2: 'Comment on egusphere-2025-3284', Anonymous Referee #2, 16 Aug 2025

This manuscript, titled "Contrasting Inland-Coastal Aerosol Mixing States: An Entropy-Based Metric for CCN Activity", presents a systematic investigation of how aerosol mixing states—specifically, the degree of internal vs. external mixing—affect cloud condensation nuclei (CCN) activity in two contrasting environments: inland (urban Beijing) and coastal (Mace Head, Ireland). This work bridges a critical gap between aerosol microphysics and climate modeling by providing a quantitative framework to incorporate realistic mixing state effects on CCN activity. It underscores the need to move beyond binary mixing assumptions and adopt entropy-based metrics for more accurate climate projections, particularly in diverse and dynamic environments like urban and coastal regions. This paper merits publication after minor revisions. To further strengthen this manuscript, the following questions should be addressed:

Why were these two locations chosen? Are they representative for inland and coastal ? If possible, efforts should be made to collect more mixing state data from inland and coastal to support this discussion.

“But the current models lack regional-specific mixing state parameters and usually assume uniform mixing in both environments. This could lead to large uncertainties in predicting CCN concentrations, highlighting the need for site-specific observations.” I suggest discussing in detail this substantial uncertainty and its sources.

Does the surrogate choice (κNH = 0.01, κH = 0.6–0.8) fully capture the hygroscopic diversity of organics, especially oxygenated and fresh POA?

Do the short campaign windows (≈ 1 month per season) adequately represent inter-annual variability in air-mass type and photochemical intensity?

Is there evidence of κ-köhler non-ideality that would invalidate the single-parameter κ assumption at high S?

How about the results if the proposed parameterization being implemented in a sectional or modal aerosol model to quantify the reduction in CCN bias compared to default internal/external mixing assumptions?

What is the sensitivity of the Pearson correlation (r = –0.74) to random vs. systematic errors in Dcri? The relationship between Dcri and χ is heavily influenced by the chemical composition of particulate matter, as well as factors such as new particle formation, emission sources, and secondary reactions. Consequently, this relationship may exhibit significant variations across different regions. Is it possible for the author to further verify this using the results directly from chemical composition analysis?

“The inland atmospheric measurements were conducted for two periods from 16 November to 6 December and 29 May to 13 June” in which year？

Citation: https://doi.org/10.5194/egusphere-2025-3284-RC2

Jingye Ren, Wei Xu, Ru-Jin Huang, Fang Zhang, Ying Wang, Lu Chen, Jurgita Ovadnevaite, Darius Ceburnis, and Colin O’Dowd

Supplement

https://doi.org/10.5194/egusphere-2025-3284-supplement

Jingye Ren, Wei Xu, Ru-Jin Huang, Fang Zhang, Ying Wang, Lu Chen, Jurgita Ovadnevaite, Darius Ceburnis, and Colin O’Dowd

Viewed

Total article views: 398 (including HTML, PDF, and XML)

HTML	PDF	XML	Total	Supplement	BibTeX	EndNote
338	44	16	398	22	10	15

HTML: 338
PDF: 44
XML: 16
Total: 398
Supplement: 22
BibTeX: 10
EndNote: 15

Views and downloads (calculated since 15 Jul 2025)

Month	HTML	PDF	XML	Total
Jul 2025	91	30	13	134
Aug 2025	136	9	3	148
Sep 2025	111	5	0	116

Cumulative views and downloads (calculated since 15 Jul 2025)

Month	HTML	PDF	XML	Total
Jul 2025	91	30	13	134
Aug 2025	136	9	3	148
Sep 2025	111	5	0	116

Viewed (geographical distribution)

Total article views: 390 (including HTML, PDF, and XML) Thereof 390 with geography defined and 0 with unknown origin.

Country	#	Views	%

Latest update: 13 Sep 2025

Short summary

Impact of mixing state on cloud condensation nuclei (CCN) activity was incorporated in very limited modeling with typically simplified assumption. This study derived a mixing state index from hygroscopicity and systematically investigated its impacts on CCN activity in inland and coastal air. An entropy-based parameterization proposed here offers a novel approach to reduce model complexity in representing aerosol CCN activation, enabling more accurate simulations of aerosol CCN capacity.


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