the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 22 Jul 2025
Aviation soot is unlikely to impact natural cirrus clouds
Abstract. The impact of aviation soot on natural cirrus clouds is considered the most uncertain among the climate impacts of the aviation sector. In this study, a global aerosol-climate model equipped with a cirrus parametrisation is applied to quantify the impact of aviation soot on natural cirrus clouds and its resulting climate effect. For the first time, the cirrus parametrisation in the model is driven by novel laboratory measurements specifically targeting the ice nucleation ability of aviation soot, thus enabling an experimentally-constrained estimate of the aviation-soot cirrus effect. The results indicate no statistically significant impact of aviation soot on natural cirrus clouds, with an effective radiative forcing of −6.9±29.8 mW m−2 (95 % confidence interval). Sensitivity simulations conducted to investigate the role of other ice nucleating particles (INPs) competing with aviation soot for ice supersaturation in the cirrus regime (soot from sources other than aviation, mineral dust and ammonium sulphate) further show that the impact of aviation soot remains statistically insignificant also when the impact of these other INPs on cirrus is reduced in the model. Acknowledging that the complexity of the soot cirrus interaction is associated with uncertainties, the model results supported by dedicated laboratory measurements suggest that the climate impact due to the aviation soot cirrus effect is likely negligible with no statistical significance.
- Preprint
(589 KB) - Metadata XML
- BibTeX
- EndNote
Status: final response (author comments only)
- RC1: 'Comment on egusphere-2025-2589', Anonymous Referee #1, 02 Aug 2025
-
RC2: 'Comment on egusphere-2025-2589', Anonymous Referee #4, 14 Aug 2025
General Comments:
The authors present a study investigating the impact of aviation soot on natural cirrus clouds using a global aerosol-climate model constrained by novel laboratory measurements of aviation soot’s ice nucleation ability. The research addresses a critical uncertainty in aviation’s climate impacts, and the integration of experimental data to parameterize model simulations represents a valuable advancement. However, several aspects of the methodology, result interpretation, and discussion require further clarification and expansion. The overall quality of the manuscript could be significantly improved with revisions. I recommend the editor reconsider the manuscript after a major revision.
Specific Comments:
Section 2.1: The description of contrail processing simulations for sampled aviation soot lacks details on the experimental conditions (e.g., temperature, pressure, duration) and how these parameters mimic real atmospheric contrail formation.
- Section 2.2: The model simulations (Table 1) involve scaling factors for mineral dust INPs (e.g., NOBGSOOT+DUST5/10) to correct for a positive bias in upper tropospheric dust concentrations. However, the basis for choosing factors of 5 and 10 is not explicitly justified. Please provide quantitative evidence (e.g., comparison with observational data) supporting these specific scaling values.
- Section 3: The statistical significance assessment using a paired-sample t-test is mentioned, but details on the sample size (e.g., number of model years, spatial grid points) and how internal model variability was quantified are lacking. This is critical for evaluating the robustness of the “non-significant” conclusion, please expand on the statistical methodology.
- Fig. 4: The relative share of ice crystals from aviation soot remains at 0.04% across REF and NOBGSOOT simulations, but the underlying mechanism for this stability is not explained. Given the removal of background soot in NOBGSOOT, why does aviation soot not increase its contribution? A more detailed microphysical explanation is needed.
- Section 3.3: The discussion of ammonium sulfate as an INP (NOBGSOOT+AMSU) notes that its effect is transient due to efflorescence/deliquescence, but there is no quantitative analysis of its temporal variability. How does this transience affect the model’s ability to capture its competition with aviation soot?
- The comparison with prior studies (e.g., NCAR CAM model results showing large ERF vs. ECHAM4/CAM5/CESM2 with non-significant effects) is cursory. The authors should explicitly discuss why their experimental constraints resolve this discrepancy.
- The conclusion suggests future focus on aviation-aerosol interactions with low-level clouds, but the manuscript does not address the limitations of the current model in simulating such interactions. How might the model’s cirrus parameterization differ from that of low-level clouds, and does this affect the transferability of the methodology?
- Section 2.2: The model update restricting aviation soot INP activity to the insoluble mode (vs. Righi et al., 2021) is justified by laboratory data on coating-free soot. However, the fraction of ambient aviation soot that remains uncoated in the upper troposphere is not quantified. Please provide estimates of this fraction to support the model’s assumption.
Citation: https://doi.org/10.5194/egusphere-2025-2589-RC2 -
RC3: 'Comment on egusphere-2025-2589', Anonymous Referee #3, 28 Aug 2025
General Comments:
For decades scientists have been conducting research to determine the role of elemental carbon as soot as a potential ice nucleating particle (INP). The effectiveness of aviation soot as an INP is a critical factor for determining human impacts on climate. By utilizing soot measurements from aircraft engines commonly used in the airline industry, using standard jet fuel, and then subjecting these measurements to contrail-processing simulated with custom-designed ice nucleation chambers, the ice nucleating ability of these soot measurements was quantified and then parameterized in the MADE3 submodel within the EMAC global chemistry model. This appears to be a sound method for evaluating the impact of aviation soot on cirrus clouds and their radiative properties. Importantly, only when the contrail-processed aviation soot particles are free of any coating (e.g., sulfuric acid and/or organics) do they act as potential INPs (nucleating ice at RHi below the threshold for homogeneous ice nucleation, henceforth hom). This was true for these aviation soot measurements, and their corresponding INP results were applied in the model to the insoluble soot mode of the aerosol microphysical scheme.
However, the finding that aviation soot is unlikely to impact natural cirrus clouds challenges a growing body of evidence that argues the contrary. That is, Groß et al. (2023, ACP) combined airborne lidar and in situ ice cloud measurements from the ML Cirrus campaign to show that the particle lidar depolarization ratio (PLDR) is higher in regions that may be more affected by air traffic, and that these regions also exhibit lower cirrus cloud ice particle number concentrations and larger effective diameters. Relatively high values of the PLDR are associated with relatively complex ice particle shapes, and ice particle shapes tend to become more complex with increasing ice particle size and ice supersaturation (Bailey and Hallett, 2009, JAS). These observations are consistent with what is predicted by the negative Twomey effect (Kärcher and Lohmann, 2003, JGR) and cirrus cloud thinning (Mitchell and Finnegan, 2009; ERL). Previously, other studies have supported this reasoning, perhaps beginning with Urbanek et al. (2018, GRL), followed by Li and Groß (2021, ACP) who used CALIPSO lidar measurements to reveal a reduction in cirrus cloud PLDR and cirrus cloud coverage (by 17% to 30%) over Europe during March and April of 2020 coinciding with the COVID-19 pandemic (relative to March and April during six other years). Aviation over Europe was reduced by at least 80% from 25 March 2020 through 30 April 2020, presumably accompanied by a similar reduction in aviation particulate emissions (e.g., soot and metallic particles). This could reduce the contribution of heterogeneous ice nucleation (i.e., het) and promote hom in cirrus clouds. Conversely, Lin et al. (2025, Sci. Adv.) found roughly a 20% increase in cirrus cloud coverage in regions affected by volcanic ash sedimenting from the lower stratosphere, showing that higher INP concentrations may enhance cirrus cloud coverage (similar to the findings in Sporre et al., 2022, GRL). More corroborating evidence is given in Li and Groß (2022, ACP).
Given the extensive evidence above arguing that civil aviation does indeed affect cirrus cloud microphysics consistent with that expected from an increase in INP concentration, a conundrum is apparent between this body of evidence and the findings of this study by Righi et al. that concludes that aviation soot will not affect the microphysics of cirrus clouds significantly. At a minimum, the authors should mention this body of evidence and the conundrum it presents. It would be most helpful if the authors could also suggest an alternative explanation (other than soot) for this body of evidence. Perhaps one possibility is that metallic particles are also emitted by civil aviation and that these particles are effective INPs that can affect cirrus cloud microphysics. Cziczo et al. (2013, Science) used the residual matter from sublimated ice crystals and clear-sky aerosol composition measurements to discriminate between het and hom processes in cirrus clouds. When het was dominant, mineral dust and metallic species were greatly enriched in the ice crystal residuals (relative to clear-sky aerosol composition). Unfortunately, the contribution of metallic species alone was not indicated, with dust and metallic species lumped together as a single category. Nonetheless, the study promotes the question of what the source of these metallic INP may be and whether they may originate from the jet engines of civil aviation.
Specific Comments:
Lines 24-25: If there is a reference for this statement, please provide one. Otherwise, the statement is obvious enough to overlook this need.
Lines 29 – 33: It may be of interest that orographic gravity waves (OGWs) are not represented in the NCAR climate models (Lyu et al., 2023, JGR). OGWs appear to contribute substantially to cirrus coverage, promoting hom due to their higher updrafts (Barahona et al., 2017, Nature; Gryspeerdt et al., 2018, ACP; Mitchell et al., 2018, ACP).
Technical Comments:
Line 142: Fig. 3f => Fig. 3e?
Figure 2 caption: (ERF => ERF?
Citation: https://doi.org/10.5194/egusphere-2025-2589-RC3 -
RC4: 'Comment on egusphere-2025-2589', Anonymous Referee #2, 02 Sep 2025
Review of "Aviation soot is unlikely to impact natural cirrus clouds" by Mattia Righi et al."This manuscript is a nice summary of recent modeling work based on new observational analysis of aviation soot and it's potential to impact cirrus clouds. This is a nice treatment and should be publishable with minor revisions as I note below. I do think some important additional work should be done before publication: some map figures illustrating regional effects would help to understand what is noise and what is signal. See detailed comments below.
Page 5, L112: is Scrit fixed in the model for each aerosol type? Shouldn’t Scrit vary for each aerosol type as a function of temperature?
Page 6, L125: have you stated how long the simulations are? Free running or nudged. I think you need a bit more detail.
Page 6, L129: See comment above. How do you pick a small signal for ∆ERF out of the noise?
Page 6, L140: To understand whether this is noise or not, it would be useful to show maps of ∆ERF (SW and LW) for one or more of the experiments: is there an expected pattern to ∆ERF that looks like aviation, or is it really noise?
Page 7, L164: how does no change in ICNC in the NOBGSOOT lead to the same change in SW? You explain why there is no change to LW, but then why is there still a SW change?
Page 10, L208: as noted, it would be interesting to look at some maps of these quantities, at least for the baseline case with background aerosols. Are the changes significant in flight regions?
Citation: https://doi.org/10.5194/egusphere-2025-2589-RC4
Data sets
Model simulation data used in "Aviation soot is unlikely to impact natural cirrus clouds" (Righi et al., 2025) Mattia Righi https://doi.org/10.5281/zenodo.15495975
Viewed
| HTML | XML | Total | BibTeX | EndNote | |
|---|---|---|---|---|---|
| 569 | 44 | 17 | 630 | 9 | 12 |
- HTML: 569
- PDF: 44
- XML: 17
- Total: 630
- BibTeX: 9
- EndNote: 12
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1
My general impression is that this paper nicely describes the work done, but that the model may be insufficient to actually give an accurate assessment or approximation of how aircraft emissions of soot actually affect cirrus clouds. This criticism is based mainly on atmospheric observations of two aspects: 1) cirrus clouds in regions heavily travelled by aircraft compared to regions with less aircraft travel (as noted in work by Urbanek, Gross and their colleagues), and 2) observations of aerosols in the atmosphere. Unfortunately, these latter observations are not specifically comparing regions heavily travelled by aircraft to regions with less aircraft travel, but they do show that aerosols that act as INP in the free atmosphere is made up of soot mixed with other species and is primarily coated by organics (Lata et al., 2021, China et al., JGR, 2017). This leads me to think that the observations of ice number concentrations by Urbanek and colleagues may be the result of mixing with organics. I read the papers described in the current pre-print, and the model apparently does not include coagulation or condensation of organics in the free atmosphere on the aircraft soot. Thus, the response in the model to emissions of aircraft soot may not be the response expected in the atmosphere. Admittedly, if the INP activity seen in the atmospheric observations is due to the coating by organics, then there may be a question of whether to assign the INP activity to the organics or to the primary soot emissions, but in any case, the soot from aircraft emissions may act as an INP after coating by organics (even though idealized laboratory experiments do not show this).
Given the above, I think some discussion of the possible drawbacks in the current study is needed, whereas now, the paper reads as if this is the final answer regarding the effect of aircraft soot in the atmosphere.
Also:
2. I am curious whether, based on the methodology and results of this study, the cirrus ERF induced by non-aircraft soot emissions would also be evaluated as insignificant. Similarly, would the cirrus ERF attributed to dust alone as an INP be significant, given that the model settings assign dust far superior ice nucleation capabilities compared to other INPs?
3. In the ERF attributed to aircraft soot in this study, how much originates from the suppression of homogeneous ice nucleation, and how much from the enhancement of heterogeneous ice nucleation? Could the authors provide these results and separately verify whether both components are statistically insignificant?
Other comments where better clarification is needed include:
Line 58-59: statement is not necessarily true in the atmosphere.
Line 77-78: Statement is too strong and not necessarily true for aircraft soot in the atmosphere.
Line 82-84: Here, I believe, this really depends on the organics that were tested and whether they agree with the mixtures expected in the atmosphere.
Line 107-108: This is a key assumption that may not be true in the atmosphere. Also, would there not be coating of sulfate and organics that takes place within the aircraft plume?
Line 125-133: What are the length of the simulations? Longer simulations could reduce the uncertainty. Also, the simulations appear to me to be nudged towards observed (or ECMWF) temperatures and velocities (based on the magnitude of uncertainty reported here, in comparison to the Righi, et al 2020 paper). You need to state this, if true, or say these are free-standing model predictions.
Line 150: I don’t believe this statement is true, since the ice nucleation properties, as measured in the laboratory, may not represent their properties in the atmosphere.
Line 155-156: aged soot can act as good INP in the atmosphere, as shown by Lata et al.
Line 213: this experimentally informed soot-cirrus impact does not account for atmospheric observations.