the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
Identification of Cloud Droplets, Ice Crystals, and Aerosols Using Ground-based Fog and Aerosol Spectrometer: Chamber Experiments and Field Observations
Abstract. Simultaneously identifying micron-scale aerosols, cloud droplets, and ice crystals within mixed-phase clouds is essential for advancing cloud microphysical research, yet it presents substantial technical challenges. This study proposes a novel phase identification method based on the Ground-based Fog and Aerosol Spectrometer (GFAS; DMT Inc.), which is rigorously validated through cloud chamber experiments and in-situ observations. Cloud chamber experiments conducted at −14 °C, −10 °C, and −6 °C yielded distributions of the polarization ratio (δ = (S−P)/(S+P)) and total scattering counts (TotalCounts = S+P) for highly non-spherical ice crystals. Based on these measurements, quantitative thresholds were established to classify three distinct particle regimes: ice crystals (δ > −0.33, TotalCounts < 1.68 × 10²), water droplets (δ < −0.82, high TotalCounts), and a transition zone (−0.82 ≤ δ ≤ −0.33) corresponding to weakly non-spherical ice or aerosols. Applied to ground-based observations at Mt. Liupan in Northwest China from January to April 2024, the threshold method was applied to a complete in-cloud glaciation cycle. The GFAS effectively distinguishes highly non-spherical ice crystals from pure water droplets, while the TotalCounts threshold achieves a separation efficiency exceeding 98 % in distinguishing pure ice crystals from aerosols. Although aerosol δ values primarily fall within the transition zone (−0.82 and −0.33) and overlap with weakly non-spherical ice crystals, preliminary discrimination remains achievable using a joint constraint of δ and TotalCounts. This method provides a temperature-independent diagnostic tool for quantifying the glaciation degree in mixed-phase clouds, offering critical observational support for understanding cloud microphysical processes.
- Preprint
(2193 KB) - Metadata XML
-
Supplement
(792 KB) - BibTeX
- EndNote
Status: final response (author comments only)
-
RC1: 'Comment on egusphere-2026-6076', Darrel Baumgardner, 02 Nov 2026
-
CC1: 'Reply on RC1', Yuchen Jiang, 03 Nov 2026
Response to Reviewer Comments
We thank the reviewer for the thorough and constructive
criticism. We acknowledge that the polarization ratio (PR)
has important limitations for particles smaller than 5 μm,
as demonstrated by the theoretical calculations in
Haberstock et al. (2026) and confirmed by our own Mie
calculations. We have made the following major revisions:1. ANGLES CORRECTED. The backscatter collection angles
have been corrected from 183.5-192° to 168-175°
following Haberstock et al. (2026).2. SIZE RESTRICTION ADDED. All PR-based phase
classification is now restricted to particles with
equivalent optical diameters > 5 μm (conservatively
> 10 μm where noted). The chamber-derived thresholds
are recalculated using only the > 5 μm subset.3. P/S GAIN CALIBRATION. We have applied the scale
factor (SF) calibration procedure described in the
GFAS Operator's Manual Appendix A.5 to convert raw
A/D counts to physical scattering cross-sections
before computing PR.4. TERMINOLOGY. "TotalCounts" has been renamed to
"P+S" throughout.5. NEW REFERENCES. Haberstock et al. (2026), Lucke et
al. (2023), and both Neuberger et al. papers are now
cited and discussed.6. FIGURE 1 REVISED to match the correct optical
configuration.Citation: https://doi.org/10.5194/egusphere-2026-6076-CC1 -
AC1: 'Reply on CC1', Yuan Wang, 03 Nov 2026
Author statement: This is not our final response to the reviewer' comments. We will carefully address each comment raised by the reviewers and submit a detailed point-by-point response together with the revised manuscript in due course.
Citation: https://doi.org/10.5194/egusphere-2026-6076-AC1
-
AC1: 'Reply on CC1', Yuan Wang, 03 Nov 2026
-
CC1: 'Reply on RC1', Yuchen Jiang, 03 Nov 2026
-
RC2: 'Comment on egusphere-2026-6076', Anonymous Referee #2, 05 Nov 2026
The authors present their first measurements of the Ground-based Fog and Aerosol Spectrometer (GFAS), characterizing its polarization and backscattering signals using cloud-chamber experiments with mixed-phase and fully glaciated clouds, as well as two in-situ events at Mt. Liupan in Northwest China. Based on the chamber measurements, they establish thresholds in the polarization ratio, δ = (S−P)/(S+P), and total backward scattering signal, S+P, to distinguish between ice, mixed/aerosol, and water regimes. These thresholds are subsequently applied to identify particle phases and aerosol populations in the field observations.
The differentiation between liquid droplets, ice crystals, and aerosols using polarization-enhanced scattering measurements is an important topic, and the combination of laboratory and field measurements has considerable potential. However, I have several fundamental concerns regarding the interpretation of the measurements and the validity of the proposed classification scheme. In particular, the absence of a liquid-water reference measurement, the strong size dependence of the measured signals, and the instrument-specific nature of the proposed thresholds make the current phase classification insufficiently supported. Consequently, the conclusions regarding cloud glaciation and the identification of partially frozen particles remain speculative.
I recommend rejection, but I would encourage the authors to submit a new manuscript once these very fundamental issues have been addressed. In particular, the authors need to provide reliable reference measurements of liquid droplets, a comparison with theory, and more detail on the calibration, experimental set-up, data acquisition and data processing. The interpretation of the results also needs to be revised. My detailed comments are given below.
General comments
- Measurements of liquid water droplets as a baseline is missing in the cloud chamber measurements. The assumption that water always has a polarization ratio close to -1 is incorrect. In (Haberstock et al., 2026) scattering cross section for the opening angles of the GFAS (Fig. 2 and 3) was modelled and measured and it is visible that water shows a size dependent signal decreasing with droplet size but never reaching values below -0.75. The reason why the signal is so low for the potential droplets you are seeing is because the detector signal response is unique in each detector and therefore the polarization ratio measured for water also depends on the individual instrument (see also point 5). This assumption appears several times in the text and will not be mentioned every single time in the following review. Consequently, the interpretation of particles with low δ as liquid droplets is not sufficiently supported and large particles cannot simply be classified as liquid based on their polarization ratio. A liquid water reference is necessary across the entire GFAS size range.
- Both S+P and δ are size dependent. The ice and aerosol signal thresholds are established with size distributions at a mode below 5 µm but applied for the entire size range. Larger particles produce stronger scattering signal and therefore also S+P will increase. Thresholds will need to be established size dependent. Considering that most ice crystals in clouds will be much larger, this threshold will not hold. This affects both, chamber and field observations.
- Detector sensitivities differ between individual instruments. Therefore, the S and the P signal, and their ratio are instrument dependent. Two different GFAS will yield different backward scattering values for the same particle population. Therefore, these thresholds cannot be generally applicable. This should be clearly pointed out by the authors. The instrument specific nature of S+P is discussed in the manuscript but the same consideration needs to be addressed for δ as well. In order to define thresholds that can be used by other researchers with a similar measurement setup I recommend calibrating the backward channels.
- Changes in the measured particle population do not necessarily indicate a phase transition. At a mountain the cloud is likely passing by, leading to measurements of different parts of the cloud that have a different ice/aerosol load and a different size distribution. Changes in wind direction, cloud dynamics, updraft velocity, and particle population could explain the observations. Additional meteorological measurements are needed to distinguish temporal phase evolution from spatial variability. While the observed changes of δ are intriguing to investigate, they alone are not sufficient evidence for glaciation.
- The intermediate classification range is interpreted as mixed-phase or partially frozen particles and then later on also as aerosols. However, the presence of particles within this range does not demonstrate that individual particles are undergoing a liquid-to-ice transition. The intermediate signals could come from measuring different sizes, shapes, or particles. The photo of the holographic measurements also no not establish the presence of partially frozen particles. They show two ice crystals. Independent verification is needed before making claims about incomplete freezing or individual particle phase transitions. This is particularly important because the authors subsequently use the classification to calculate IWC and LWC.
Specific comments
Introduction
L16ff. Is there a reason for referring to this parameter as the polarization ratio rather than the degree of linear polarization, given its definition?
L18ff. Why is the term TotalCounts used when S+P represents the total backward-scattered detector signal? I suggest using a term that more accurately describes the measured quantity. This also applies to L156ff.
L50. What specific physical or optical signatures of aerosols are you referring to?
L55. I would remove the term "fine" when referring to particles below 50 µm, as fine particles in aerosol science generally refer to particles below 2.5 µm. The same applies to L272.
L77ff. Is there a reference supporting the statement that small, weakly non-spherical ice crystals cannot be unambiguously distinguished from background aerosols using polarization-enhanced scattering measurements?
Experimental setup and measurements
L87ff. The opening angles of the GFAS are 3.8–20° in the forward direction and 168–176.5° in the backward direction. It appears that different specifications may have been taken from another instrument manual.
L108ff. The cloud chamber set-up needs to be described in much more details (incl. adding respective references, technical details, dimensions, etc) given the crucial role within this study? In particular, please clarify whether the GFAS was placed directly inside the chamber or connected via a sampling line, and whether potential wall effects could influence the measurements. How long were measurements conducted at each temperature? The supplementary figure does not provide sufficient information to understand the experimental setup.
L117. You state that Snomax induced complete glaciation at −10 °C and mixed-phase conditions at −6 °C. Is this supported by independent measurements or a reference demonstrating the expected phase state under these conditions? Also, aged Snomax does not nucleate as well (https://doi.org/10.3390/mi12030296). How was the Snowmax stored and handled? Which batch was used?
L120. Which holographic fog droplet spectrometer was used, and where are the corresponding results presented? The single image in the supplementary information provides little information about the individual experiments. Please also specify the instrument's measurement size range.
L125. I suggest changing the section title to "Field campaign at Mt. Liupan".
L129. Please provide a reference supporting the described climatological conditions, particularly the frequency of fog events and the occurrence of supercooled fog and orographically forced mixed-phase clouds.
Instrument description and methodology
L140. Rather than describing the GFAS as a "multi-functional particle spectrometer", I suggest specifying that it is a single-particle light-scattering spectrometer.
L145. The description of the two backward-scattering detectors as being positioned at a nominal scattering angle of 90° is potentially misleading. The 90° angle refers to the beam splitter redirecting the scattered light towards the detectors, rather than the actual scattering angle. I recommend removing the reference to 90° and describing the beam splitter instead. I would also remove "sideways scattering", as the instrument measures backward scattering.
L149. I would be careful about referring to the definition used by Glen and Brooks as the conventional CAS-DPOL definition. They use different formulations in their 2013 and 2014 publications, namely S/(S+P) and S/P, respectively.
L150. The degree of depolarization is not defined. Please clarify its definition and distinguish it from the polarization ratio introduced earlier.
L153. What is meant by "GFAS-specific formulation"? Please clarify what makes this formulation specific to the instrument.
L156ff. Several studies have used the total backward-scattering signal to constrain particle optical properties (Baumgardner et al., 1996; Glen and Brooks, 2013, 2014). These should be cited here. Also, why was the more established forward-to-backward scattering ratio not considered? The total backward-scattering signal is strongly size dependent, which complicates its interpretation using fixed thresholds.
L167. Please provide references supporting the statement that the described power-law diameter–mass relationship is an established parameterization.
L169. The equation IWC = 𝛼𝐷𝛽 describes the mass of an individual ice crystal rather than the ice water content. I suggest writing m(D) = 𝛼𝐷𝛽 and reserving IWC for the integrated mass over the ice particle size distribution.
L178. The reference to resolution limits below 50 µm is unclear in the context of GFAS measurements, as the instrument does not measure particle size through imaging resolution. Furthermore, this statement appears to concern measurement uncertainty and would be more appropriate in the discussion than in the methods section.
Results and Discussion
Cloud chamber results
Figures 2 and 3.
- The background shading makes it difficult to distinguish where actual measurements took place. I suggest removing it (also Fig. 4).
- The jet colormap is not perceptually uniform and is not colorblind-friendly. I recommend using a perceptually uniform alternative.
- Please clarify the definitions of relative frequency and conditional frequency.
- How do you explain the observation that all ice crystals are smaller than the droplets? Given the high supersaturation, would freezing and subsequent growth not be expected? Is this really physically feasible? If so, please provide references.
L184ff. How was complete glaciation of the chamber independently verified? The interpretation of the distributions appears premature. The narrow distributions in δ and S+P could partly result from the limited particle size range, as the particles are predominantly below 5 µm. Furthermore, the larger-signal tails in Fig. 2a and c show a decrease in δ similar to the mixed-phase case in Fig. 2e. The distributions in panels b, d, and f also appear remarkably similar. I would therefore be cautious about interpreting these distributions as evidence of a clear separation between liquid droplets and ice crystals.
L188ff. How were the threshold values selected? The percentages reported in different sections of the manuscript appear inconsistent. Please clarify the selection criteria and apply them consistently.
Additional suggestion: To better understand the instrument response, I recommend plotting two-dimensional histograms of the S and P channels against forward scattering. In some GFAS versions, the S channel appears to be dominated by noise for small particles below approximately 5 µm, with values around 150–200 ADC counts, before increasing with forward scattering. This should be investigated to ensure that the measured polarization ratio is not primarily determined by the difference between detector noise and the P-channel signal. A liquid-water reference measurement would be particularly useful here.
L226. The high concentration of INPs introduced through Snomax limits how representative these observations are of natural mixed-phase cloud conditions. INPs are generally much less abundant and active in the atmosphere. Revise the claim that these measurements provide representative polarization characteristics of natural mixed-phase clouds.
L236–239. Please provide a reference supporting this interpretation. You mention that each temperature condition was maintained for 30 minutes. How does this compare with the expected freezing timescale of a 10 µm droplet once freezing has been initiated?
L243. The interpretation of larger particles as liquid droplets and smaller particles as ice crystals is not sufficiently supported. Please consider the size dependence of δ and compare the polarization distributions for different particle sizes with those expected for liquid droplets.
Additionally, S/P is introduced here without prior definition or explanation of its physical meaning. I also do not see sufficient evidence that surface-tension oscillations explain the observed changes in δ. Please provide references supporting the proposed influence of surface-tension and aerodynamic perturbations on the polarization of scattered light by water droplets.
L266. What do you mean by "representative" mixed-phase cloud? Were multiple measurements under comparable conditions conducted?
Field observations: mixed-phase cloud event
Figure 3d. The interpretation of this event raises several questions:
- How can you distinguish a phase transition from changes in the sampled cloud volume or a moving cloud with spatially varying microphysical properties?
- The particles classified as liquid do not appear to exhibit polarization values comparable to those observed for water in the chamber.
- Why does δ decrease towards more mixed-phase/liquid-like values towards the end of the event, despite the steadily decreasing temperature?
- I recommend extending the time series to include measurements before and after the cloud event to better characterize the cloud boundaries and the background conditions.
L271. Why is the size range shown as 2–50 µm when the GFAS measurement range is 0.5–50 µm?
L273. The spectral evolution appears more complex than the described continuous broadening. The distribution broadens until approximately 15:00, narrows afterwards, and shows a temporary increase in larger particles around 17:00 before broadening again towards 18:00. Although the small-particle mode remains relatively constant in size, its number concentration increases around 17:00, coinciding with the maximum in δ. Given the size dependence of δ, I recommend examining the polarization distributions of the two modes separately.
L277. Please specify the exact time referred to as the "near peak of glaciation" to allow comparison with the other figures.
L278. How do you distinguish interstitial aerosols from activated particles? Furthermore, the ice crystals measured in the chamber were within a similar size range. How can you ensure that not some of the particles classified as ice in the chamber has actually been interstitial (non-activated) aerosols? Please specify the Snomax concentration used in the chamber.
L279. The MVD does not reach 12 µm and actually decreases when the proposed glaciation begins. It only increases shortly before 17:00. I therefore find the interpretation that newly generated ice crystals contributed to larger particle sizes difficult to follow. The curves also do not fully align, and the statement in L289 that the MVD doubled does not appear accurate. Please clarify the interpretation and quantify the changes.
L287. Why is a decrease in δ interpreted as cloud dissipation? This could equally reflect a change in the sampled cloud volume or the passage of a cloud edge.
L292ff. I find the interpretations of Fig. 4 difficult to support, as they rely on the phase classification discussed above. However, there are several interesting features that deserve further investigation:
- The distributions in panels a, c, and e follow remarkably similar trajectories to those observed in the previous experiments. This raises the question of whether the observed differences primarily reflect particle size rather than phase.
- In panels b, d, and f, there is an abrupt change around 6 µm, where the otherwise smooth polarization distribution appears to shift. What causes this feature?
- Below 10 µm, the δ distribution in panels b and d remains relatively constant around −0.7 before shifting towards higher values with increasing diameter.
- Around 15 µm, two distinct clusters with different polarization values appear at similar particle sizes. What could explain these populations, particularly if only the smaller particles are interpreted as ice crystals?
L334. How is the "mixed" classification accounted for when calculating IWC and LWC? Since approximately 80% of the particles fall into this category, excluding or assigning them to either phase could substantially bias the calculated water contents. Furthermore, if the distinction between liquid and ice is based on a fixed δ threshold, the size distributions suggest that only droplets larger than approximately 12 µm would be classified as liquid. This needs to be addressed and the implications for the calculated IWC and LWC discussed.
Aerosol event
Figure 6.
- The distribution similar to those in Figs. 2 and 4 (a, c, e), with the main differences being the concentrations in different parts of the distributions. How does this support the proposed separation between different particle types?
- What causes the abrupt change in polarization values around 6 µm in panel d, including values reaching approximately 0.8?
- A higher refractive index in spherical particles can lead to stronger backscattering compared with water. This could be relevant when interpreting the observed distributions.
- Including relative humidity (e.g., in panel a) could help interpret the influence of hygroscopic growth on particle size and optical properties
- Please clarify the difference between particle percentage and conditional frequency in panels b and d.
L346. Rather than representing a general aerosol baseline, this appears to be an example of an event with elevated aerosol concentrations. Were other cloud-free periods characterized by similar distributions? If the intention is to establish an aerosol baseline, the authors should combine statistics from multiple cloud-free periods rather than relying on a single event.
L352. The peak in δ appears to be around −0.82 rather than −0.6, and therefore extends close to the proposed threshold of −0.33. Please check the reported value.
L370. The statement that the combined polarization and TotalCounts thresholds effectively separate aerosols from ice crystals is too strong. If the aerosol event is used to establish the aerosol regime, could the particles classified as "mixed" in the chamber and cloud-event measurements also be interpreted as aerosols? More importantly, does this distinction remain valid when a cloud is present? The current results do not demonstrate a reliable separation of aerosols, liquid droplets, and ice crystals within a cloud.
L377. The stated range of δ appears to be below −0.7 to approximately −0.3, rather than −0.7 to −0.2.
Conclusions
The entire conclusions section needs to be revised. Currently it reads like a mixture of results and discussions. I recommend that the authors rewrite it and focus on their main findings and clearly separate on what has been found and what remains hypothetical.
L385. The statement about the synchronous discrimination of aerosols, supercooled liquid droplets, and ice crystals is ungrounded and needs to be revised, as aerosols have not been independently distinguished within the cloud event.
L414–430. This section reads more like a discussion of the limitations and implications of the results. I suggest moving it to the discussion section. The main methodological limitations should also be introduced earlier in the manuscript rather than repeated in the conclusion.
Technical corrections and typos
L17: Add "backward" before "scattering counts".
L109: Clarify what is meant by "calibrated particle microphysical datasets".
L257: The definition of TotalCounts is repeated.
Fig. 1: Please cite the original source if adapted from the instrument manual. The P-polarized detector also appears to be missing its polarization filter. I recommend labelling the collection and collimating lenses as well as the beam splitters.
Fig. 3a: What do the horizontal lines represent?
L344: Visibility appears to be well above 10 km, closer to 20 km.
L383: The conclusion starts with "However," which seems inappropriate as an opening statement.
References
Baumgardner, D. et al. (1996) ‘Refractive indices of aerosols in the upper troposphere and lower stratosphere’, Geophysical Research Letters, 23(7), pp. 749–752. Available at: https://doi.org/10.1029/96GL00707.
Glen, A. and Brooks, S.D. (2013) ‘A new method for measuring optical scattering properties of atmospherically relevant dusts using the Cloud and Aerosol Spectrometer with Polarization (CASPOL)’, Atmospheric Chemistry and Physics, 13(3), pp. 1345–1356. Available at: https://doi.org/10.5194/acp-13-1345-2013.
Glen, A. and Brooks, S.D. (2014) ‘Single Particle Measurements of the Optical Properties of Small Ice Crystals and Heterogeneous Ice Nuclei’, Aerosol Science and Technology, 48(11), pp. 1123–1132. Available at: https://doi.org/10.1080/02786826.2014.963023.
Haberstock, L. et al. (2026) ‘The Ground-based Fog and Aerosol Spectrometer–Technical description and first applications’, EGUsphere, 2026, pp. 1–38.
Citation: https://doi.org/10.5194/egusphere-2026-6076-RC2
Viewed
| HTML | XML | Total | Supplement | BibTeX | EndNote | |
|---|---|---|---|---|---|---|
| 439 | 157 | 47 | 643 | 84 | 33 | 43 |
- HTML: 439
- PDF: 157
- XML: 47
- Total: 643
- Supplement: 84
- BibTeX: 33
- EndNote: 43
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1
General comments
The primary message from this study, which the authors want the readers to take from this manuscript, is that the polarization ratio (PR) is a sensitive indicator of asphericity and can be used to differentiate ice crystals from supercooled liquid water. This is a worthy objective and one that I had looked forward to approving of when I started reviewing this manuscript. Unfortunately, none of the cloud chamber results that are shown in this study can be used to make this case. Theoretically, and as shown by results in the paper that the authors should have referenced, but didn't, Haberstock et al (2026), the PR is sensitive to shape, but can be used to do shape discrimination only for equivalent optical diameters larger than at least 5 um, and more significantly above 10 um. Had the authors calculated the theoretical PR for spherical particles, using Mie theory, they would have discovered several important things about this ratio, using the backscattering collection angles of the GFAS, which are incorrectly stated in the manuscript:
1) For a system collecting light over the angles of 168-175o (not 183.5-192o as stated in the manuscript), the theoretical PR varies from -0.2 to -0.6.
2) The P and S scattering cross sections below 5 um oscillate so much that at times the PR is positive, not negative.
3) In order to apply PR to distinguish the aspherical particles, the P and S measurements have to be scaled to their theoretical values because the amplification of these signals in the instrument are different.
All of the results that were presented from the cloud chamber were from ice crystals << 5 um, hence their high PR values cannot be used as a metric to separate liquid from ice.
Additional comments.
1) The authors need to read the Haberstock et al paper (see below) so that they understand all of the uncertainties that have to be taken into account before they start analyzing their data.Had they read that paper they would have realized that their Figure 1 is not correct, nor are either the forward or backward scattering collection angles.
2) The correct reference for the BCPD is Lucke et al (2023), not Freer et al.
3) Two papers by Neuberger et al. have already been published with the GFAS measurements, see references below.
4) Referring to the sum of the P and S intensities as "TotalCounts" is confusing because the term "counts" is usually thought of as the number of events, but here the authors appear to be referring to the analog to digital counts when the detector outputs are digitized by the analog to digital converters. I suggest a different terminology, e.g. P+S of sumP_S.
A lot of work will be necessary before this study is ready to be published.
Needed references:
Haberstock, L., Baumgardner, D., Hughes, D. D., Legrand, J., Neuberger, A., Zhang, C., Kanji, Z. A., Maier, C., Riipinen, I., Krejci, R., and Zieger, P.: The Ground-based Fog and Aerosol Spectrometer – Technical description and first applications, EGUsphere [preprint], https://doi.org/10.5194/egusphere-2026-3790, 2026.
Lucke, J.R., Jurkat, T., Baumgardner, D., Kalinka, F. et al., “Characterization of Atmospheric Icing Conditions during the HALO-(AC)3 Campaign with the Nevzorov Probe and the Backscatter Cloud Probe with Polarization Detection,” SAE Technical Paper 2023-01-1485, 2023, doi:10.4271/2023-01-1485
Neuberger, Almuth, Stefano Decesari, Andreas Aktypis, Hendrik Andersen, Darrel Baumgardner, Federico Bianchi, Maurizio Busetto et al. "From Molecules to Droplets: The Fog and Aerosol Interaction Research Italy (FAIRARI) 2021/22 Campaign." Bulletin of the American Meteorological Society 106, no. 1 (2025): E23-E50.
Neuberger, A., Ranjan, R., Ding, H., Mattsson, F., Haberstock, L., Baumgardner, D., Decesari, S., Ekman, A. M. L., Hughes, D. D., Mohr, C., Paglione, M., Riipinen, I., Rinaldi, M., and Zieger, P.: Importance of hydrated aerosol particles for aerosol-fog relationships in the Italian Po Valley, EGUsphere [preprint], https://doi.org/10.5194/egusphere-2025-5419, 2025.