the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 05 Mar 2025
Contrail models lacking post-fallstreak behavior could underpredict lifetime optical depth
Abstract. Proposed optimized contrail avoidance schemes rely on being able to robustly predict which contrails cause the most climate warming. However, it has not yet been shown that different contrail models agree sufficiently to support the targeting of individual contrails by climate impact. To address this, we compare the most widespread contrail model, CoCiP, to a higher-fidelity contrail model, APCEMM, under parametrized meteorological conditions. The results show that the lifetime optical depth (a proxy for climate impact) in APCEMM is 3.8 times that in CoCiP, and that the models have opposite sensitivities of their lifetime optical depth to relative humidity. We argue that these differences are due to the differing representations of the distribution of ice particles in space and in size across the contrails. The use of a monodisperse ice particle size distribution in a Gaussian plume means that CoCiP models the contrail exclusively as an accelerating, falling mass – a fallstreak. The use of a spatially gridded and size-resolved aerosol scheme allows APCEMM to represent the separation of the precipitation plume from the contrail core, hence modelling behavior past the initial fallstreak phase. This behavior is consistent with prior large eddy simulation studies, and it accounts for 92 % of the aggregate APCEMM lifetime optical depth. This suggests that fallstreak-only simulation may underestimate contrail climate impact. While a strategy avoiding all contrail formation is still expected to yield a reduction in climate impact, implementing optimized strategies requires more research to establish confidence in model predictions.
- Preprint
(1420 KB) - Metadata XML
- BibTeX
- EndNote
Status: final response (author comments only)
-
RC1: 'Comment on egusphere-2025-278', Anonymous Referee #1, 20 Mar 2025
Review of
Contrail models lacking post-fallstreak behavior could underpredict lifetime optical depth
(egusphere-2025-278) by Caleb Akhtar Martínez, Sebastian D. Eastham, Jerome P. Jarrett
General commentsIn this paper, the authors conduct urgently needed tests of contrail simulation codes, CoCiP and APCEMM. These codes, in particular the first one, are applied by a number of groups to offer contrail avoidance services. The presented results point to serious questions about the reliability of these services.
The research strategy is sound and reasonable: first a comparison for a "standard" case is conducted, which already displays the most striking disagreement between the two models. The disagreement is explained to originate from two model differences: A contrail is treated in CoCiP as a Gaussion plume (not further resolved spatially) with bulk microphysics (whether actually a monodisperse size distribution is applied must be checked, see below), while APCEMM treats them with 2D-spatial (cross sectional) resolution and with spectral microphysics. The basic comparison is then followed by sensitivity studies, where 6 system parameters are varied one-by-one. The two models agree on the sign of the sensitivities for 5 of the parameters, but not on the sensitivity to the most important parameter, namely the relative humidity in the supersaturated layer. The lifetime-integrated optical depth, used as a proxy for climate impact, decreases with RHi in CoCiP, but increases in APCEMM. All these differences are explained by the authors with the construction of and the assumptions in the two models.The tests have been conducted in a special background situation, which is good for a first test. The authors state the necessity to check further parameters and that much work remains before one can trust the contrail avoidance service. I agree with this. I do not agree with the final conclusion that it would be currently better to avoid all contrails instead only the ones with strongest warming. With the current ATC system this would not be possible, at least not in dense airspaces.
This paper deserves publication in ACP, but e few questions need still to be resolved and the paper can also be improved in various aspects, as detailed below.
Major issues:
1) I am not a CoCiP user, so I am not familiar with its representation of a crystal size spectrum. My question here is whether it is actually monodisperse which means that every ice crystal has the same size (variance zero), or whether it rather may have a variance which is implicit. Note that many bulk cloud physics models (1-moment or 2-moment) implicitly assume a size distribution, where the variance is a function of the mean size. Please check, how CoCiP does this, i.e. whether it is monodisperse or perhaps monomodal.
Later, you say that CoCiP uses the ice mass and number and that the single size of all crystals is computed from these two values. There may be reasons for this, but in cirrus models, if they have 2-moment schemes (e.g. ice mass and number), one of the reasons for this is that this allows more freedom in treating size distributions implicitly.
Figure 5 left shows that dN/dt = -N/τ, that is the crystal loss rate is not constant. Wouldn't a monodisperse distribution of crystals, falling with identical speed, lead to a constant loss rate? Furthermore, if all crystals do what the crystal in the contrail centre does, why then is there ongoing crystal loss instead of instantaneous vanishing of all crystals?In line 200 you write "average ice particle". If the distribution is monodisperse, all ice crystals are average.
I think, your interpretation is valid for both cases, but your statements should be correct.
2) Section 2.2, Eqs 1 and 2: Why is the spatial integral only over the width of a contrail and not along its length? This quantity as used here seems to have some similarity to the "total extinction" of Unterstrasser and Gierens and a similar quantity introduced by Lewellen. These authors use them as proxies for climate impact, perhaps an instantaneous one. But in the present case, I fear this could not serve the intended comparison. The total radiative effect of a contrail should be the lifetime integral of the vertical optical thickness at every point of the contrail (width X length). For the current purpose, length should somehow increase with lifetime, and it seems that this effect is overlooked. Wouldn't the differences between CoCiP and APCEMM be even larger if the integrals would cover the complete contrail area over the complete lifetime?
Minor issues:
The title can be improved. It is not clear what "lifetime optical depth" may be. I think, the problem is not the optical depth, but the lifetime-integrated radiation effects or the change of the radiation energy flow integrated over the contrail lifetime.
When the same expression appears in the abstract, it should be written as "lifetime-integrated optical depth". As an explanation is given (proxy), the expression is acceptable, but in the title it should be changed. At the end of the abstract you explicitly write "contrail climate impact", why not so in the title?Abstract, line 19: "a strategy avoiding all contrail formation is still expected to yield a reduction in climate impact". I believe such a strategy does not work for practical reasons (ATC problems) and does not exist therefore. I also do not believe that such a strategy has been proposed, as written in Line 25.
Line 29: why hypothetical? Why not as well in actually occurring situations?
Line 44: cross sectional area of ~ 100 km², what kind of cross section do you mean? Say a contrail is 1 km deep, than it must be 100 km broad in your example.
Line 47: "The limited comparisons that have already been performed for large eddy simulations indicate disagreement in this regard", can you be more specific? As far as I remember, the cited papers weren't model comparison papers. What do you mean?
Line 50: "inconsistencies"? Probably you simply mean model differences or disagreements or contradictory results. To my view, two different models, independently developed, can neither be consistent nor inconsistent.
Lines 68-70: The two sentences "Teoh shows..." and "For this reason" are not logically connected, to my opinion. To only consider long-lasting contrails is justified without Teoh's results. Whether the latter turn out tenable can be doubted in view of your results, in particular the probably wrong sensitivity to layer supersaturation lets me doubt to which degree Teoh's results are believable.
Line 109: "Equivalent"? Is there a subtle meaning that I do not understand or do you just mean "Equal"?
Line 216: correct "observable in from satellites".
Figure 6: The figures are not entirely understandable. Partly, because they are just scatter plots and it is not clear which CoCiP point is paired to which APCEMM cross. Then, while CoCiP should indeed be represented by a single group of points for "fallstreak only", there should be 2 groups of crosses for APCEMM: "fallstreak only" and "all phases". I suggest, to have the integrated optical thickness on the y-axis, while the x-axis should be numbered 1-14, that is the number of the sensitivity experiment. Then each number would have one blue point for CoCip, and, say a red point and a red cross for APCEMM "fallstreak only" and "all phases". A similar outline would also work for the rhs panel.
Figure 6, caption: correct "unobserbavle".
Line 300: "Varying the layer RHi causes the lifetime optical depth to decrease in CoCiP and increase in APCEMM". This sentence is a bit unclear, since the word "varying" includes both decreasing and increasing. Please correct.
Line 345/6: The Schmidt-Appleman criterion says nothing on contrail persistence, so I suggest to add ice supersaturation as a condition. Avoidance of all, that is, including very short contrails, is not useful and probably worsening the climate (unnecessary fuel consumption and emissions).
Line 385: "our results suggest that contrail avoidance strategies which focus on avoidance of all contrails will have the greatest chance of producing a real climate benefit". I would not subscribe to this conclusion. It renders contrail avoidance practically impossible, in particular for ATC reasons. The ATC sectors where no contrails form would become overcrowded if they are neighbours to sectors where contrails can form. I think, the appropriate conclusion of your test is that more work is needed to "calibrate" the models to realistic behaviour and to test whether the promised results are satisfying. Your Section 5 points to this direction and I fully agree to the statements of Sect. 5.
Citation: https://doi.org/10.5194/egusphere-2025-278-RC1 -
CC1: 'Comment on egusphere-2025-278', Sina Hofer, 25 Mar 2025
A comment to egusphere-2025-278
Contrail models lacking post-fallstreak behavior could underpredict lifetime optical depth
by Martínez, Eastham, and Jarrett
Sina Hofer and Klaus Gierens
The model intercomparison conducted by the authors is very welcome as it supports our opinion that the fruits of "contrail avoidance" are not hanging as low as was assumed for a long time. To our opinion, much work needs still to be done.
In putting their results and model assumptions into perspective, the authors feel the need to argue that ISSRs can last many hours. In order to support this, the authors write:
"This is further supported by a recent preprint (Hofer and Gierens, 2024) which analyzed a larger ECMWF dataset and found that contrail lifetime is most commonly limited by sedimentation, as opposed to advection of the contrail out of the ISSRs."
Here, we find an error and a misinterpretation. The error is: we have used data from the ICON model of the German Weather Service, not ECMWF data.
The misinterpretation is: Although the movement of ISSRs is often aligned to the wind, this does not imply that contrail lifetimes are most commonly limited by sedimentation. To the contrary, in another recent preprint (Hofer and Gierens 2025) we show that contrail termination by sedimentation and contrail termination by synoptic processes (contrails leaving the ISSR with the wind and large-scale subsidence turning super- into subsaturation) have similar time-scales of a few hours. It is difficult to say in advance which time-scale is shorter.A similar statement is found in the conclusion: "the predominant mechanism for contrail evaporation is through sedimentation, as opposed to advection (Hofer and Gierens, 2024; Irvine et al., 2024)." We do not know Irvine et al. wrote (by the way, this was 2014, not 2024), but it is not our statement.
More arguments for long-lasting ISSRs can be found in case studies by Bakan et al. and Spichtinger et al., if needed.
Statistical arguments about the unobservable fraction of contrail lifetime can be found in a paper be Gierens and Vazquez-Navarro (2018). You might want to check whether your results agree with those from the statistical arguments.
References
Bakan, S., Betancor, M., Gayler, V., and l, H. G.: Contrail frequency over Europe from NOAA–satellite images, Ann. Geophys., 12, 962–968,
1994.Gierens, K., Vazquez-Navarro, M., 2018: Statistical analysis of contrail lifetimes from a satellite perspective. Meteorol. Z., 27, 183-193 DOI 10.1127/metz/2018/0888.
Hofer and Gierens, 2024: Kinematic properties of regions that can involve persistent contrails, egusphere-2024-3520 (Note: title changed during revision.)
Hofer and Gierens, 2025: Synoptic and microphysical lifetime constraints for contrails, egusphere-2025-326.
P. Spichtinger, K. Gierens, and H. Wernli, 2005: A case study on the formation and evolution of ice supersaturation
in the vicinity of a warm conveyor belt’s outflow region, Atmos. Chem. Phys., 5, 973–987, 2005
www.atmos-chem-phys.org/acp/5/973/Citation: https://doi.org/10.5194/egusphere-2025-278-CC1 - RC2: 'Comment on egusphere-2025-278', Anonymous Referee #3, 30 May 2025
Viewed
| HTML | XML | Total | BibTeX | EndNote | |
|---|---|---|---|---|---|
| 326 | 94 | 19 | 439 | 14 | 21 |
- HTML: 326
- PDF: 94
- XML: 19
- Total: 439
- BibTeX: 14
- EndNote: 21
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1