the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 12 Dec 2025
Anthropogenic Modulation of Dust-Dominated Ice Nucleation in an Urban Dryland City of China
Abstract. Ice-nucleating particles (INPs) are crucial for cloud formation and precipitation, yet their variability and influencing factors in urban dryland regions remain poorly understood. While natural dust is recognized as the dominant INP source, the extent to which anthropogenic pollution modulates INP abundance remains insufficiently quantified. Here, we present online observations of INPs (−15 to −35 °C), together with co-located aerosol size distribution and chemical composition in Lanzhou from winter 2024 to spring 2025. We show that long-range dust transport boosts INP concentrations by × 15 at −30 °C. Elevated secondary inorganic aerosol during pollution was enhanced and negatively correlated with INP activity (R = −0.71). We further refine a two-parameter scheme (1–2.5 µm aerosol diameter and temperature) that reproduces 83 % of observations within a factor of 5. These findings underscore the need to include local aerosol heterogeneity and dust-pollution interactions in INPs parameterizations for more accurate regional climate simulations.
- Preprint
(2364 KB) - Metadata XML
-
Supplement
(1539 KB) - BibTeX
- EndNote
Status: open (until 11 Feb 2026)
- RC1: 'Comment on egusphere-2025-5998', Anonymous Referee #1, 07 Jan 2026 reply
-
RC2: 'Comment on egusphere-2025-5998', Anonymous Referee #2, 30 Jan 2026
reply
Chen et al. conducted field measurements using an online automated CFDC instrument from winter 2024 – spring 2025 in Lanzhou, China. Combining INP measurements, ambient monitoring measurements, and air-mass trajectories they determined that dust transport was associated with higher concentrations of immersion INP. Compositional analysis revealed that polluted urban environments are negatively correlated with INP concentrations, with implications for dust-pollution interactions where urban plumes can suppress INP concentrations. This study fits within the scope of ACP. However, there are several scientific issues that need to be addressed in a detailed revision. First, the authors discuss the inclusion of a new CFDC instrument, but details are not provided on the specifics of the instrument and a comparison to existing instruments not provided. The calibration detailed is also lacking, where only two compounds used to validate immersion freezing. Furthermore, the limitations of the compositional analysis provided needs to be included and reorganized. The manuscript would also benefit from reorganization, as methodologies are frequently described in the results. After a careful revision, this manuscript should be reconsidered for publication as a research article.
Title should be specific you are testing immersion freezing. Maybe replace ice nucleation with Immersion Freezing?
Line 14 Elevated secondary inorganic aerosol associated with urban pollution was enhanced in the winter and…
Line 21 – While true, it can depend on region. Provide a citation to strengthen the statement that INP are a very low constituent of background INP.
Line 35 – 38. This is misleading as currently stated and can be improved. The role of BBA is highly debatable within the INP community and arguably not largely supported by experiments. You even reference the study by (Chen et al., 2025)describing this discrepancy is likely due to co-lofted materials such as dust rather than a systematic underestimation of BBA particles themselves. Numerous compositional aircraft ice residual analyses also not corroborate this finding either, as cloud residuals are frequently depleted in carbonaceous BB. From the paper cited, BBA was also binned into categories (residential, agricultural, natural, and others) rather than a lumped BBA.
Line 39 – 42 Metallic particles from industrial sources in urban environments have also been identified as an INP source. You later measure this so indicating their importance prior would be beneficial.
Line 83 Please include AGL height to your latitude and longitude coordinates.
Line 87 A critical description or citation of the aerosol inlet used is missing from the methodology. For any field measurement studies, this description is essential. At the very least, an existing citation be provided with a D50.
Fig 1. Contrast of text, trajectories, and background colors make this figure difficult to read, particularly panel A. Perhaps reducing the opacity of the background? You can also label the CFDC as CFDC-IAS to be more specific as this figure is introduced before the instrumentation is described.
Line 87-97 In your overview of sampling methods conducted, you are unspecific in instrumentation for everything but the Aethalometer and heavy metal analyzer. Be consistent in how you are describing instrumentation for the overview. This can inadvertently highlight certain instruments as the “highlight(s)” of the campaign. Arguably, list the variables measured and reference the appropriate sections of the manuscript where they are discussed in more detail.
Line 92 Are there any pertinent citations on the operation of these monitoring stations?
Line 94-95 Measurement units for compositional aspects? Are these arbitrary ion intensities?
Line 99 – 100. The SMPS measures the electrical mobility diameter. What shape factor, particle density, and other corrections are you using to get an equivalent Stokes diameter?
Line 105 I would add a sentence that you are converting diameters to merge the SMPS and APS measurements together.
Line 106 Averaging isn’t what allows you to merge the data sets, the equivalent diameter conversion in combination with averaging is.
Section 2.3
Major Revision: If this is a new CFDC instrument as you indicate, you need to include a much more thorough description of the instrument. At the minimum it should match a short version of existing descriptions of CFDCs (Garimella et al., 2016; Kulkarni and Kok, 2012; Lacher et al., 2017; Rogers et al., 2001; Stetzer et al., 2008). If this is a previously used instrument, you can cite that study but as you state it is new.
Line 115 Design of chamber? Relationship between flow, vapor pressure, and temperature? Steady state profiles? Length of main chamber and evaporative section? Aerosol inlet? This all needs to be addressed if it’s a new instrument.
Line 120 Is there a reason your sheath to sample flow ratio is 5? That seems low in comparison to other instruments and could lead to significant lamina spreading and particle losses (Garimella et al., 2017).
Line 125 State the freezing modes of your calibrations. You also need to be consistent with your units for water saturation. Supplement uses SSw and methods use RHw. I would also include the onset conditions you achieved for these calibrants and provide a brief comparison to literature.
Line 125 I would demonstrate that your freezing onsets for your calibrants are consistent with literature values across a broader range of conditions. Your calibration also seems limited to temperatures in the range of -20 to -30 °C but conducted experiments from −15°C to −35°C? Ideally you should plot your onsets of standards on a temperature vs RHice or RHliq and include respective lines for water saturation, homogeneous freezing etc. For new CFDC’s, obtaining a homogeneous freezing point is required to validate functionality and can be provided with ammonium hydrogen sulfate or ammonium nitrate which should only undergo homogeneous freezing. For a new instrument, presenting this summary of calibrations in the article is important.
Line 127 It may be better to just formally refer to aerosol transmission as D50 throughout the article and briefly explain it before the first use (inlet preferably). Cut size may not be intuitive to some readers.
Line 144 There isn’t just results in this section as you also include interpretation. It should be labeled Results and Discussion to make this clear.
Fig. 2 - There is a lot of information here and some of it is redundant when considering Fig. 3. I would only select subplots for the monthly variations and seasonal differences and related aerosol characteristics be moved.
Line 155 – 178 Your results/conclusion appear to depend on the assumption that PM10 and water-soluble Ca2+ are definitive markers of mineral dust, with HYSPLIT corroborating this result. While I appreciate the inclusion of composition measurements, I would be very clear in explaining that your chromatography ion concentrations are the soluble fraction in figures and further results. You stated measurements were also collected using a XRF heavy-metal analyzer, but I do not see those presented until later and only in the context of PM2.5. Was that instrument measuring PM2.5 only? This is a more convincing technique to indicate mineral dust aerosol rather than IC which only gives information on surface water soluble material. Al, Fe, and Si would be much stronger markers than water soluble Ca. As you indicated later, the IC aspect is important for aged mineral dust but to define the presence of dust using IC isn’t appropriate.
Line 162 I am surprised your BC is not correlated with K+ as it is a strong biomass burning marker.
Line 180 Clearly define the clean, moderate, PM2.5 high, and dust categories including any thresholds you considered. This would be better placed in the methods early on so the reader expects to see these categories and knows how they are defined. There is also no definition of “high” provided. What is the mass threshold being used? If you associate this with specific ions (particularly nitrate and sulfate), then perhaps categorized as Urban?
Line 186 Here you define dust events without water soluble Ca. Be consistent throughout the article on how you are defining dust.
Line 191 Which ions and specify these are water soluble ions
Fig 3. Should have the heavy metal analysis included.
Line 214 This is vague, if the atmosphere is clean wouldn’t the low INP concentration be due to the low concentration of aerosol in general not the ice nucleating activity?
Line 217 The HYSPLIT analysis should be its own section.
Line 230 Satellite imagery rather than elevation would be more convincing in showing dust sources. You could also shade/circle desert regions for the source trajectories. Include in the HYSPLIT section.
Fig. 5 I agree with Reviewer 1, the volume site density isn’t previously described and is confusing here.
Line 245 Given your compositional measurements, I would include these here to strengthen the argument this is dust.
Line 247 Define SNA before using the acronym.
Lines 253 – 280 Definitions of compositional methods should be in the methods section, not the results. These are also supported by citations which would strengthen the argument earlier rather than later. I also see more inclusions of the heavy metal analysis
Line 296 This sentence is confusing as it is currently written.
Fig. 9 There are a lot of fits on each of these plots, particularly the top row. The correlations between these are also very low. I wonder if this figure may be better placed in the supplement or only include the figures you reference in Lines 325 – 335.
Lines 358 – 365 This reads more like introductory material.
Lines 384 …study provides… measurements that can improve parameterizations of aerosol-ice-cloud interactions in North-West China with implications for other dryland cities where urban air interacts with transported dust.
Citation: https://doi.org/10.5194/egusphere-2025-5998-RC2
Viewed
| HTML | XML | Total | Supplement | BibTeX | EndNote | |
|---|---|---|---|---|---|---|
| 219 | 104 | 26 | 349 | 50 | 21 | 28 |
- HTML: 219
- PDF: 104
- XML: 26
- Total: 349
- Supplement: 50
- BibTeX: 21
- EndNote: 28
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1
Chen et al. report an observational study on Ice-Nucleating Particles (INPs) conducted in Lanzhou, a semi-arid inland city in Northwest China, from winter 2024 to spring 2025. The research addresses a critical gap in understanding INP variability in urban dryland regions, where natural dust and anthropogenic pollution interact. The core finding is that episodic, long-range transport of mineral dust is the primary driver of INP bursts, while persistent urban pollution plays a much smaller and potentially suppressive role. The study topic matches the journal scope. The manuscript is a bit too lengthy for the given content, and the overall quality and clarity of figures and writing should be improved for its consideration for publication in ACP. If the data scarcity and providing the long-term observational data are the main motivations of this study (L65, L69, L366-369), this manuscript might be more suitable to be a measurement report rather than a research article.
Title: Add “during Winter and Spring” at the end to clarify that the result is representative for specific seasons.
L9 …cloud and precipitation modulation, yet…
L10 Consider changing from influencing factors to sources or properties. “Influencing factors ” sounds awkward.
L10-11 the dominant INP source --> substantial INPs
L11 the extent to…INP abundance remains --> interactions between dust and anthropogenic pollutants, and how they alter INP abundance remain
L15-16 Avoid using the word “two-parameter scheme” in the abstract. Generalize to something like INP parameterization based on aerosol size and freezing temperature or anything similar.
L19 Ice“-”nucleating
L19 …ice crystal formation on water-insoluble aerosol surface by…
L21 10^-3 to 10_-5 numbers seem misleading. INP concentration depends on freezing temperatures, and could be an order of 10^-6 or lower. The authors may consider deleting the parentheses.
L23-24 What INP abundance ranges and properties are crucial for precipitation formation then? The authors may consider explaining a bit more in detail for readers.
L31 dominant --> abundant
L44-45 This reviewer does not understand this sentence. Break it down to two sentences and add sufficient explanation for each reference.
L53 particle diameter --> particle larger than 0.5 micron diameter
L54-56 How about empirical parameterizations, such as the Phillips parameterization (https://doi.org/10.1175/2007JAS2546.1)? The current discussion of INP parameterizations seems superficial and irrelevant to the study topic (interaction of anthropogenic and dust). More in-depth discussion incorporating previous INP parameterizations would be meaningful for readers.
L106-108 This reviewer respectfully disagrees that time-averaging the data enhances the accuracy of the size distribution. It just offers a time-averaged representation, which smooths out the size distribution spectra as some pulsive data points get merged.
L113-115 Please add an appropriate reference here. The citation is missing.
Sect. 2 The authors need to describe the time resolution of each aerosol measurement data and how they synchronized those data (time averaging, right?) for the correlation analyses presented later in the manuscript.
L117 …in the upper growth region of simulated cloud particles.
L118-119 The authors should offer a reference justifying 3 micron threshold size for ice crystals.
L123 This reviewer disagrees. There are many papers reporting the gap/offset between online INP measurement techniques (e.g., CFDC) and offline ones at a certain freezing temperature range. Often, INP concentrations measured by online techniques reads higher than offline ones. This discrepancy can stem from different detection limits of detectable INP concentration for various techniques. What’s the detection limit of CFDC that the authors utilized for this study? From Fig. 10, this reviewer guesses it’s ~0.15 INP sL^-1? Please clarify this in the manuscript. A proper discussion of CFDC’s INP detection limit in the manuscript might be beneficial for readers.
L129 What is the temperature ramping interval from -15 dC to -35 dC? Were the authors be able to observe homogeneous freezing at -35 dC?
L141-143 Reference missing – if it’s a commonly applied approach, some refs should be offered.
Sect. 3 (L217-228) Discussion of the impact of precipitation and gusty wind on INP suppression/abundance is missing. Airmass trajectory is important to trace the airmass source and path, but local-synoptic scale meteorological conditions could also be key. Dust in mid-latitude is tied to convective cloud and precipitation formation (https://doi.org/10.1016/j.atmosenv.2008.09.069). Often, the same convective system and associated surface winds can induce dust resuspension from surface sediments.
L156-158 Too many things are discussed in a single sentence. The authors might consider explaining this thoroughly over several sentences.
L167-168 It is very hard to see the relation between dust, PM10, and Ca ion concentrations in Fig. S4. Dust time series data are not even seen in Fig. S4. A better presentation needs to be offered; otherwise, the statement here is not convincing. Normalized concentration time series for Panels (c) and (d) might be a better representation to make the authors’ point.
L179 Define aerosol conditions.
L186-188 Offer a reference for dust event characterization here. Not just in SI.
Fig. 5 Why is volume site density offered? What is its significance to surface site density? To the reviewer’s knowledge, surface site density is more relevant to INP as IN active sites are presumably on the surface of water-insoluble particles. Why volume matters? Please clarify. Otherwise, the reviewer suggests removing the volume site density discussion. The current manuscript seemingly does not offer the significance of the volume site density.
L287-295 Sounds speculative and superficial.
Fig. 8 looks very busy with many fits. Some can be moved to SI. A better presentation with only crucial info should be offered. The reviewer also wishes to see the comparison of surface site density parameterizations from this study to previous studies as a function of freezing Ts. Perhaps the surface site density comparison can be offered in another panel.
Fig. 10 vertical lines at 0.15 sL^-1 on x-axis and horizontal ones ~1 sL^-1 on y-axis look like an artifact. It’s probably due to the detection limit of observation and prediction. Regardless, why the minimum INP concentration values are different between observation and prediction? A relevant discussion seems to be missing. The authors may consider calculating time average INP concentrations so that artifact-looking straight lines would disappear from the figure.
L358-359 This does not fit in the conclusion. Sounds like introductory info.
L366-369 The authors might consider summarizing IN efficiency (surface site density) besides INP conc. here.
L370 What are urban-inland INP bursts? Define it well earlier.