the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 23 Apr 2026
Measurement report: Chemical composition of submicron aerosol and cirrus and contrail ice residuals measured in the UTLS over Germany in winter 2018
Abstract. The knowledge of submicron aerosol composition in the upper troposphere and lower stratosphere (UTLS) and the contribution of aircraft exhaust on the cirrus and contrail formation is still limited due to sparse observations and snapshots not considering the evolution of these clouds. Airborne measurements of the aerosol chemical composition were conducted in the 2018 wintertime UTLS region over Germany. With the help of the hybrid mass spectrometer ERICA (ERC Instrument for the Chemical composition of Aerosols), the composition of background aerosol was analyzed as well as the composition of cloud residuals by applying a counterflow virtual impactor. We found that carbonaceous material plays an important role in the particulate matter in the wintertime UTLS over Germany, among which biomass burning (BB) material is the prevailing species. Complementary simulations of air mass history and synoptical analysis suggest that BB material results from wildfires, in particular the Thomas fire in Northern America. Besides the long-range transport of BB aerosol, the chemical composition of UTLS aerosol is driven by local meteorological conditions. Further, carbonaceous aerosol from aircraft exhaust including soot and engine oil contribute to the aerosol population in the size range below 200 nm. Aging contrails contain signatures of aircraft exhaust such as coated soot and engine oil among other biogenic organic compounds and are consistent with the enhancement of these compounds in aircraft exhaust plumes. Sea spray and mineral dust dominate cirrus residuals, implying the formation at a liquid state.
- Preprint
(10075 KB) - Metadata XML
-
Supplement
(1181 KB) - BibTeX
- EndNote
Status: open (until 04 Jun 2026)
- RC1: 'Comment on egusphere-2026-2161', Anonymous Referee #1, 19 May 2026 reply
-
RC2: 'Comment on egusphere-2026-2161', Anonymous Referee #2, 25 May 2026
reply
This manuscript presents detailed in-situ chemical measurements of atmospheric aerosols and cirrus and contrail ice residual particles sampled in the upper troposphere and lower stratosphere. These are important and novel mass spectroscopy measurements that are also very challenging to conduct, so the authors are commended for this. I also appreciate the authors' calculations of number fraction uncertainties, which are critical for interpretations throughout. Major revisions to the manuscript are needed to consider and incorporate critical missing literature (see below), which will add greater depth to the interpretation of the important data presented herein. Additional recommendations for revisions are described below.
A key revision need is to do a thorough literature search for other single-particle measurements in the UTLS. PALMS measurements of aerosols and cloud residual particles in the UTLS are very pertinent, and most of those prior studies are missing from the manuscript. Reading these papers and connecting them to this work will improve the current manuscript. From my literature search, some key missing literature includes (but likely not limited to): Murphy et al. 2021, ACP, “Radiative and chemical implications of the size and composition of aerosol particles in the existing or modified global stratosphere”; Katich et al 2023, Science, “Pyrocumulunimbus affect average stratospheric aerosol composition”; Hu et al 2019, ACP, “Long-range-transported Canadian smoke plumes in the lower stratosphere over northern France”; Sharpe et al. 2026, Environ. Sci: Atmos., “Chemical imaging of individual stratospheric particles sampled over North America”; Yu et al 2026, GRL, “Radiative forcing from anthropogenic sulfur and organic emissions reaching the stratosphere; Juurikkala et al 2025, ACP, “Prior heterogeneous ice nucleation events shape homogeneous freezing during the evolution of synoptic cirrus”; Froyd et al 2010, ACP, “Aerosols that form subvisible cirrus at the tropical tropopause”; Froyd et al 2009, ACP, “Aerosol composition of the tropical upper troposphere”; Jost et al 2004, GRL, “In-situ observations of mid-latitude forest fire plumes deep in the stratosphere”; Shen et al 2025, Nat. Geosc., “Stratospheric aerosol perturbation by tropospheric biomass burning and deep convection”; Lawler et al 2025, JGR, “Composition and stratospheric fate of aerosol particles originating in the Polar Vortez”; Ullrich et al 2019, J. Atmos. Sci, “Comparison of modeled and measured INP composition in a cirrus cloud”.
Similar to the above statement, the authors state (L85): “the chemical composition of contrails has not been provided yet.” From a quick literature search, it is clear that this is false. Some key missing literature includes (but likely not limited to): Twohy and Gandrud 1998, GRL, “Electron microscope analysis of residual particles from aircraft contrails”; Petzold et al 1998, Atmos. Res., “Elemental composition and morphology of ice-crystal residual particles in cirrus clouds and contrails”.
Since ERICA includes both single-particle and bulk analysis, the discussion of data from each part of the instrument needs to be very clear. Throughout the manuscript, the laser ablation data are discussed in terms of “particle fraction” or similar phrasing. I recommend that the authors make sure to add “number”, i.e. “particle number fraction”, so that it’s not accidentally confused as particle mass fraction. For someone familiar with single-particle mass spec, the switching back and forth between the types of data will likely be much easier to discern than someone not familiar with single-particle data; therefore, it’s important for the text to be very clear for all readers.
Section 3: It is critical that the various particle types are described chemically before the discussion that starts on L226. For example, L226-229 discuss fractional contributions of 4? particle types, but the reader has no reference for the chemical composition of these particle types, which is critical for the reader to understand the rest of the results. These particle types are shown in Figures S3-S11 and defined in terms of ion markers in Tables S2-S3, but there is no dedicated text section defining the observed particle types and explaining each type’s composition. At least one paragraph describing the particle types should be added at the beginning of Section 3, with a text section added to the SI. (I note that the BB particles are chemically described later in Section 3.1, which is useful.)
I am confused by the definition of “Processed” for sea spray and mineral dust. I would have thought this meant that they had undergone reactions in the atmosphere to gain nitrate and sulfate, for example. Instead Table S2 defines the difference between these processed types and the non-processed types to mainly be the presence of metals, which confuses me. For “Processed sea spray”, Cornwell et al. is cited, but they observed metals in fresh sea spray aerosol. No reference is given for the markers for “Processed mineral dust”. Similarly, for Figure S2a, are you sure this isn’t a dust mass spectrum? Also, what makes the mass spectra in Figure S4 more processed than Figure S3?
Table 1 lists two cloud probes, but I didn’t see that data incorporated into the Results (except for a statement about possible ice shattering?). These size-resolved cloud particle data would likely be useful for helping interpret the cloud residual particle data and especially for L428-437, where water vapor saturation conditions are discussed in detail. The cloud probe data would also likely be useful for interpretation of Figure 15.
The conclusions is a long summary of the results and discussion. I recommend adding some implications throughout the conclusions section and also connections to prior UTLS aerosol and cirrus and contrail residual particle measurements for further support and context.
Additional Comments:
- Abstract, L13-14: Why does mineral dust as a cirrus residual imply liquid formation? I’m surprised by this.
- Add references for the sentences on: L31, L52, L80, L241
- L48: Clarify that, unless you are referring to new particle formation, “secondary aerosol particles” aren’t being formed; rather secondary aerosol is forming on existing particles.
- L117: Remove “American”
- Figure 1: Should the x-axis be “GPS Latitude”?
- L154: This states 14 individual types, but I only count 13 in the list.
- Are there updates to the in prep Anderson et al. and Clemen et al. manuscripts?
- L191: Fix typo “the their”
- L202: Is the “number concentration of aerosol particles” here referring to interstitial aerosol or cloud residuals behind the CVI? What does “enhanced” mean quantitatively?
- L214-217: Please state here which flights corresponded to warm vs cold periods.
- Section 3.1 title: According to Figure 3, sulfate appears to be equal or greater than organics in mass, and the single-particle data show that these species are internally mixed. Therefore, would “carbonaceous-sulfate” particles be more appropriate than “carbonaceous” in the header here? Similarly, L476 in the Conclusions refers only to carbonaceous compounds when sulfate is a common internally mixed aerosol species.
- L234: Do you mean to imply that meteoric material is internally mixed with carbonaceous aerosol?
- L245: It would be useful to state the concentration ranges in parentheses here for context.
- L261: What is meant by “higher signals of BB material”?
- L282: Does PF mean particle fraction? I did not see it defined in the text. Reduce use of unnecessary abbreviations to make the text easier to read.
- L289: I recommend using the phrasing “low Arctic” instead of “polar regions” (here and elsewhere in the text), as I do not see any trajectories going to Antarctica as well.
- Figure 7 caption: I recommend referring to the box in Figure 6 to define the source region (at least I assume this is what you did?).
- L333: Is this referring to an enhancement in the “meteoric” particle fraction? Please clarify.
- L377-378, 409-412, 459-462, and elsewhere: Add uncertainties to the percentages discussed in the text to enable evaluation of significance. Also, for 378, it is not clear what “major differences” are being referred to here.
- Figure 12: Are these two size distributions within their variance and uncertainty? Please show this. Also, could this figure go in the SI?
- Section 3.3: What altitudes were clouds sampled at? Please state here.
- L405-406: It is stated that it is “obvious” that the BB particles “participated in cloud processing”, but it isn’t clear here how this conclusion was made.
- L408-409: These states about particle coating amount and ice-active sites seem speculative. Figure 14a is cited here, but it doesn’t provide this information as implied, especially since cloud phase also isn’t separated here (ice and mixed-phase combined).
- L448-449: Please refer to where this is shown.
- Figure 15: What is the uncertainty on the RH measurement? Could this figure be moved to the SI?
- L473: Aren’t number fractions with uncertainties calculated quantitative though?
- L476: BB is a source, not a species.
- L481-482: Were sulfate and nitrate mass concentrations also higher during these periods? I don’t remember this being discussed, although I could have missed it.
- L530: Where is the size distribution of cloud residuals? What percentage of cloud residuals were >170 nm?
- Is there an update on the availability of the ERICA data?
- Figure S12 and Table S2: There are prior studies of diesel lubricating oil that show similar single-particle mass spectra that could further support your source attribution (e.g., Spencer et al 2006, Atmos. Environ.).
Citation: https://doi.org/10.5194/egusphere-2026-2161-RC2 -
RC3: 'Comment on egusphere-2026-2161', Anonymous Referee #3, 27 May 2026
reply
This study reports the measurements of aerosol composition within the UTLS over Germany during the winter of 2018. The findings highlight a substantial contribution from biomass burning with North American wildfire origin to the wintertime UTLS aerosol in this region. The characterization of aerosol composition within the aircraft contrails shows no significant difference from ambient background aerosol, mainly due to the instrument detection limit in small particles. The cloud residual composition reveals that mineral dust and sea spray are the primary constituents of these residuals.
The presented measurements are undoubtedly important, but the manuscript requires some revisions before considered for publication. For example, many figure captions are vague and lack explanations of the data, the methods, and/or the axis labels. For example, what is “source fraction” in Figure 7, and how does this relate to the data presented in the figure? Also, there is no explanation about the particle type/composition indicated in the Figures. An example is that it is unclear what "processed OC" and "processed dust" are, and how they differ from just "OC/ dust"? Additionally, the reported size distribution ranges are inconsistent and do not align with the instrumental detection limits described, which confuses. There are also some acronyms used in the text without definition (i.e., ExTL, CPRs).
Below are some detailed comments:
- The reported size ranges of aerosol data are inconsistent. The size limit for ERICA-LAMS is 170 nm to 3.2 μm, for ERICA-AMS is 80 nm to 2 μm, and importantly, the aerosol size of the aircraft inlet (scoop inlet) is only up to 1 μm. However, the aerosol data reported here often include sizes beyond 1 μm. For example, Figure 4 shows the size distribution of ERICA-LAMS identified particle types in the range of 90 nm to >3 um, without any information about how they extended the lower size limit of ERICA-LAMS from 170 nm to 90 nm. The size range of BB type 1 aerosol is reported as 150 nm to 500 nm, which is also below the lower limit of the ERICA-LAMS. There is no explanation of how the size distribution reported here is processed/selected/compared. The author needs to be clearer here about their data processing of the size.
- The differentiation between BB-type 1 and BB-type 2 is unclear. It was explained in the manuscript that the two types are different in size and altitude distribution, and they are very similar in chemical signatures. However, it’s vague how the authors identify them and differentiate these two types in the first place. Are they identified by the MS signal of C2H- and CNO-? Also, Figure 9 shows a comparison of organic mass concentration obtained by ERICA-AMS and particle fractions of BB Type 2 and EC/soot by ERICA-LAMS. Might it be worth including the BB type 1here? I wonder what the 3D dimension of BB type 1 looks like.
- Figure 12: The label of the Y-axis: shouldn’t the unit for dN/dlogd be "number per cm3"? It is unclear to me what "normalized with respect to the measurement period " means. Also, the author didn't specify here whether the size distribution is ambient or STP.
- In the contrail section, the aerosol composition measured in the exhaust is very similar to that in the background atmosphere (full size range). What are the corresponding CO2 levels of the measured “background aerosol”? Figure 13b shows that the background aerosol composition contains engine oil comparable to that in the exhaust. If the CO2 at this sampled event is high, can that still be considered “background atmosphere”? Also, Dp<200nm is really 170nm<Dp<200nm, right?
- “We found that carbonaceous material plays an important role in the particulate matter in the wintertime UTLS over Germany, among which biomass burning (BB) material is the prevailing species.”Is this a unique UTLS signature just for 2018, since wildfire influence is significant this year? How does this year compare with other years?
- Line 23: seems like a single sentence talking about the gas phase in the middle of the stratosphere aerosol discussion without context.
- Line 68: “ultra-fine metal pieces” in the exhaust? Or does it mean nanoparticles containing metals? might need some clarifications and a reference here.
Citation: https://doi.org/10.5194/egusphere-2026-2161-RC3
Data sets
ND-MAX data M. Yang-Martin https://science-data.larc.nasa.gov/aero-fp/projects/
Viewed
| HTML | XML | Total | Supplement | BibTeX | EndNote | |
|---|---|---|---|---|---|---|
| 246 | 121 | 19 | 386 | 58 | 20 | 19 |
- HTML: 246
- PDF: 121
- XML: 19
- Total: 386
- Supplement: 58
- BibTeX: 20
- EndNote: 19
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1
This study sets out to present analyzes of background aerosols in the tropopause region and cirrus cloud residuals over Germany using the ERICA hybrid mass spectrometer.
The manuscript is not ready for publication and will require some major revision.
The manuscript contains material that merits to be published, but the presentation needs some additional work. Especially, an analysis showing the relevance of the observed residual properties to the research objectives is required. The manuscript lists several cloud probes in Table 1, which should be used to compare the integral number density of ice crystals from these probes with the integral number density of the CVI residuals (It is not clear how the different cloud probes are used in this study other than one is used to determine ice crystal event). I suspect that this comparison alone, between cloud-probes and CVI, will trigger an analysis about the relative contribution of particles larger than 170 nm to any climate related property of the cirrus cloud. In general, the results would benefit from being framed as detailed case studies rather than as a broad climatological assessment of the UTLS region over wintertime Germany.
Specific comments
P1 Abstract Observations using aircraft are snapshots, but there are more snapshots than the manuscript suggests. Especially with respect to the ice crystal section. The statement “below 200 nm” is misleading since the detection limit is (80 nm). Similarly, the statement that sea spray and minerals dominate cirrus residuals is misleading as this only concerns the accumulation mode. If all residuals were accounted for, the picture will look different.
P2 Introduction
The introduction could benefit from being shorter and less in the style of a textbook chapter and more targeted to the context of the two main research objectives. That is: “…the contribution of aerosol particles to cloud formation .. “(P2, L27) and “…the contribution of biomass burning…”(P2, L64). This would also allow the authors to site more relevant literature if focusing on introducing the research objectives rather than a generic big-picture of airmass transport and cirrus properties.
P3, L84 After contrails (ref)
P5, L133 How much time remained after chasing contrails for background and cirrus observations at these three flight levels? How much of that time was in-cloud data?
P5, L149 How much time is represented by the 2269 spectra and how much of this is background cirrus clouds? Does this include RF7 or not?
P7, L169 Is particle classification made for six of the eight research flights?
P8, L178 How was “presence of cloud” defined?
P8, section 2.4 and S3 It is difficult to follow how this is done. Could a simple decision tree be used? The comment about chattering crystals and resulting aerosols is not clear. Is this outside or inside the CVI? How is this a second indicator of cirrus? Do you make a distinction between “natural cirrus” and “cirrus”? The supplement section did not help very much. Why is an aerosol and ice particle event and background CO2 classified as cirrus and not a non-aerosol and ice particle and background CO2 event also classified as cirrus? Are contrails never formed at temperatures above -38C? How much of the flights were conducted at temperatures below -38C? A worrying statement is that the background concentration of ice particle number concentration (ambient?) was determined to be 0.35 cm^3. What instrument was this? If this is the CVI it is an indication that the counterflow is not balanced. One of the first steps in a flight is to check the integrity of the instrument in cloud free air (by for instance reaching the stratosphere) and make sure the background is zero. Is this base line value representative for cloud-free air? A background concentration of 0.35 cm^3 is a rather dense cirrus and if this was the background, what concentration levels where the events classified as cirrus (time series?)?
P10, L220 The statement about the UTLS region over wintertime Germany is too general for the dataset available, which is part of three flights. There are several instances when this type of generalization is made and I think the whole approach should be turned to case studies. Especially as a major effort on the discussion is put on the fire in North America, which was exceptional in size and time, and ended shortly after the flights. Hence, three flights later in the winter would have likely given a different characteristic result.
P10, L225- When giving fractions, please include what the fraction is based on, total or carbonaceous.
P11, L230 Figure 3 and elsewhere state if the concentrations are given as ambient or STP. I found that the CPC is STP.
P11, L246-You state Type I and II are mainly differing in the occurrence and size, but how do you actually classify them? I understand that they show similarities and differences once classified, but how do you deal with both types occurring at the same time?
P12, 275 Figure 6 It looks like most trajectories shown arrive at altitudes below 4 km. What does this mean for the UTLS over Germany?
P12, L276 Figure 7 What is source fraction, how is it calculated? What is the significance level of the scatter plot? Is it the different y-scales or is it potentially different relations to source fraction I should focus on?
P12, L281 “We have indications…” is this Figure 7?
P13. L286 How were these three airmasses selected? Would different instances of BB1 or BB3 air masses (i.e. at some other flight) show the same relations with respect to the types as in Figure 8, or are these the only instances with these air masses during the campaign?
P13, L295 “Indeed,….” What do you mean?
P14, L219 “In conclusion,…” Do you exclude the contribution of particles injected in the ITCZ region?
P16, section 3.2 That exhaust is diluted with time (distance) is of course expected, but it is nevertheless important to understand how the measurements were conducted. Because it is not trivial to chase exhaust plumes because of the dynamics of the plume. Was the plume visible, if so, how would a contrail affect the observations by the scoop inlet? Did the chasing platform fly at the same altitude or lower to intercept the plume (considering the descent of a plume)? The analysis would benefit from using CO2 as tracer and use it normalize the other variables to something like emission indices. This would make the discussion about the distance behind the emitting aircraft more interesting.
P20, L395 In balance with the discussion about BB, it would be appropriate to include some references on the topic how representative the 200 nm exhaust particles are for the impact of the UTLS by A/C exhaust.
P21, L415 There are old CVI measurements that support that residual particle in the accumulation mode is non-volatile (i.e. SS and MD). However, it is not what controls the number density of ice crystals even at cold temperatures. As pointed out above, some categoric statements should be softened based on the limitation of the analytical techniques.
P21, L428 The relation between cirrus properties and the ambient relative humidity is more an indication about where in the life cycle the cloud is rather than about ascending or descending airmasses. Cloud forms in ice supersaturated environments and spends most time around 100%. As it dissipates it can remain for some time, but below 80% it is only the largest crystals that sedimented into dry air that survives long enough to be detected by sufficient probability. This discussion would benefit by being put into context using the suite of papers that came out from the INCA experiment.
P24, L454 Over Germany aircraft emissions are not only visible in the direct exhaust plume but will show up in what looks like natural cirrus as well. As the authors pointed out the dilution will bring the ambient concentrations down to near background values in a rather short time period, but the signature will still be possible to observe in the residual properties. Were the observed contrails embedded in cirrus clouds, or where they conducted in cloud-free air? The discussion about residual particles and contrails would benefit by comparing it with previous CVI measurements. For instance, the suite of papers related to the AEROCONTRAIL experiment.
P24 Conclusions This section could be significantly reduced, focused and better organized. Currently points arrive a little random and some sentences are hard to understand. Limit the conclusions to key results, bullets or according to the structure in the main text. Alternatively, vertical distributions, BB transport, or CVI residuals.