4

Contrail Modeling Systems

The goal of a contrail modeling system is to simulate the formation, persistence, and evolution of contrails and contrail-generated cirrus as well as their radiative impacts. Simulating contrails typically requires a set of multiple models for different aspects of the problem, hence a “modeling system.” For an end-to-end system, one starts with a model for the engine exhaust that accounts for the combustor conditions and technology, the fuel type, and the operating conditions for each flight segment, as described in Chapter 2. Chapter 3 on measurements reflects the emphasis on needing observations to feed the models. In addition, there are only a handful of current contrail models and this chapter does not attempt to describe these models in detail as they are previously published in the literature. Chapters 5 and 6 provide more specifics about how the models will be used and what simulations are needed. This chapter describes the modeling systems that are initialized with the characterization of the exhaust at the exit plane of the engines and then simulate the initial formation of contrails behind the aircraft and their evolution and dissipation. Engine emissions are placed into an exhaust plume model that includes the airflow around the airframe and accounts for the pressure and temperature fluctuations plus vertical motion of the aircraft’s wake vortices as they collect the engine exhaust and descend 100 m or more, all the while mixing the vortices with the ambient atmosphere. After the dynamical energy in the vortex dissipates, the final mixing of the exhaust plume and any contrails into the atmosphere (e.g., into general pollution layers or cirrus clouds) is controlled by the large-scale atmospheric wind shear and static stability. Aircraft engine exhaust consists of gases and particles that can continue to react with each other or the ambient atmosphere once they leave the engine exit plane. Gases can be converted to other species or condense to existing or new particles. Most importantly here, the particles can condense ambient water vapor to form initial contrail ice particles (<10 micron diameter), which can grow into ice crystals (20–100 micron diameter). Like natural cirrus clouds, these ice crystals will eventually sublimate through gravitational settling into drier regions or large-scale atmospheric subsidence and warming. Contrail modeling systems need to include the dynamical mixing of the aircraft wake with the ambient atmosphere plus the chemistry and microphysics of the particles.

WAKE VORTEX MODELS

A number of research groups have developed models of the wakes and contrails from jet aircraft.¹ In addition, there have been detailed observations of the wake vortex and contrails that can test these models (Gayet et al.

___________________

¹ See, for example, Fritz et al. (2020), Kärcher et al. (2015), Lewellen and Lewellen (2001), Paoli et al. (2013), Paugam et al. (2010), Picot et al. (2015), Sussmann and Gierens (1999), and Unterstrasser et al. (2008).

2012; Schumann et al. 2013). Many studies have shown that the intensity of contrails (i.e., their radiative impact on climate) depends on not simply the Schmidt–Appleman criterion (ice supersaturation ratio, designated as relative humidity over ice, RHi >100 percent; Appleman [1953]), but also other factors controlling the wake vortex such as the jet-to-wing-span ratio, aircraft weight and number of engines, ambient wind shear, and stability (Llewellen and Llewellen 2001; Paoli et al. 2013; Saulgeot et al. 2023).

Vortex dynamics follows several stages, as does the microphysics controlling the contrail ice particles. Initial visible contrails form from the expanding exhaust of each engine within 1 s. These contain all the exhaust products, especially the soot, engine oil, and trace metal particles that are seeds for the contrail ice particles, and collectively, these descend from the aircraft’s downwash. After about 20 s, the pair of rotating wing-tip vortices begin to sweep up the exhaust and contrail ice, redistributing it in rotating vortices ranging from flight level to several hundred meters below (see Lewellen and Lewellen 2001). After about 400 s, the vortices decay and the remaining non-uniform distribution of ice particles enters the mature phase where the plume is dispersed by large-scale atmospheric processes (wind shear, turbulence, sublimation). Typically, the contrail ice particles gravitationally settle, falling through the atmosphere, and sublimating about 700 m below flight level (Schumann et al. 2015). The vortex phase is the most complex for particles and ice physics: the plume cools, and formation of sulfate aerosols, freezing on solid nuclei, condensation, heterogeneous nucleation, and coagulation of particles occurs. In the presence of a sufficient number of soot aerosols, heterogeneous freezing of ice occurs, but homogeneous freezing of sulfate particles might take place if soot concentrations are low enough. Organic compounds are also present in the exhaust and, like sulfate, can condense on the non-volatile soot particles or mix with sulfate particles. While these small particles can be treated as passive tracers within the bulk air flow in plume models, contrail ice has its own dynamical transport.

GLOBAL CLIMATE MODELS

The mature phase of the exhaust plume is described as the diffusion regime where the large-scale atmospheric structure defined by wind shear and static stability control the final evolution of the exhaust plume and any contrails (e.g., Fritz et al. 2020; Paoli et al. 2013), and in this regime, mixing of compounds in the plume with ambient air compounds becomes important. In addition, the plume is advected (i.e., transferred) along with the ambient air. Ice crystal growth in this regime takes place primarily from interaction of the exhaust particles and contrail ice with the ambient water vapor.

The climate impacts of contrails and contrail-induced cirrus all come from this mature phase in which the contrail vortex model is linked with a global-scale weather forecasting or climate model. These global models calculate the integrated radiative forcing and dehydration impacts of a contrail and place it in the context of the background atmosphere (water vapor, clouds, surface albedo).²

Global climate models (GCMs) are used directly for predicting contrail impacts (Bock and Burkhardt 2016; Burkhardt et al. 2010; Chen and Gettelman 2016; Gettelman et al. 2021), but these do not resolve clouds and individual contrails, instead generating them statistically in each of their large, 100-km grid cells and letting them evolve statistically according to the meteorological conditions and the cloud physics of the climate model. The GCM derives its initial plume parameters from observations and plume models. The advantage of a GCM is a closed mass budget wherein the evolution of water in a cloud will pull water from the ambient atmosphere and ensure proper interaction with existing and subsequent cloud formation. Climate models also have closed energy budgets and can be used to predict the radiative effects of contrails (given their parameterized cloud microphysics).

CONTRAIL PLUME MODELS

Other examples of contrail modeling involve calculation of individual flights within an atmosphere defined by current weather prediction systems. Examples are the contrail cirrus prediction model (CoCiP; Schumann 2012) as well as the plume model of Fritz et al. (2020) which focuses on the chemical and microphysical evolution of the

___________________

² See Chen and Gettelman (2016), Gettelman et al. (2021), and Schumann et al. (2015).

plume. Singh et al. (2024, see Tables 2 and 3) reviewed many of these models. These plume models are optimized to calculate contrails globally given identified aircraft flight paths and aircraft emissions plus the meteorological conditions from weather forecasting systems such as the European Centre for Medium Range Weather Forecasting. These models can include all global aircraft flights but must parameterize the dynamical processes described in the large eddy simulation models (e.g., Lewellen and Lewellen 2001; Paoli et al. 2017).

CoCiP is a plume model that takes exhaust particle number and forms contrails under the right conditions. The advection and evolution of the contrails is followed with a Lagrangian Gaussian plume model. Mixing and bulk cloud processes are treated quasi-analytically or with a numerical scheme. Contrails disappear when the bulk ice content sublimates (as a result of changes in the background atmosphere) or precipitates (when the ice particles are large enough). Water from the background ambient atmosphere condenses and is the primary mass of contrails and contrail cirrus; yet, when using a forecast model, the water vapor in the forecast model is not changed. This process is different when CoCiP is coupled with a climate model wherein the atmospheric humidity profile changes when ice crystals from the contrails fall to lower altitudes and release the water vapor (Schumann et al. 2015). Forecast/reanalysis models can provide the enabling background meteorology for these plume models (i.e., ice-supersaturation profiles, wind shear, stratification) but have not been designed to re-assimilate the redistribution of water vapor. While climate models can be, and have been, merged with plume models, the link with the forecast models is needed to predict specific time periods and to optimize flight paths for reduced contrail climate impacts. There is currently no seamless prediction system.

ICE-SUPERSATURATION FORECAST MODELS

One aspect of forecast models that needs improvement is the prediction of ice-supersaturation regions, including the amount of supersaturation, as these determine contrails formation, their persistence, and their optical depth (i.e., climate impact). Current forecast models are not optimized to accurately predict these regions and the amount of ice supersaturation.

SYSTEM FOR CONTRAIL PREDICTION

NASA has several modeling/observation systems that could be used to predict ice supersaturation and the formation of contrails using either National Oceanic and Atmospheric Administration–based models (e.g., RAPP through NASA Langley Research Center) or NASA’s own assimilation system (GEOS, a weather forecast model optimized to assimilate NASA Earth system observations). Such systems could be used to predict contrails statistically or with an embedded plume model such as CoCiP. The GEOS system currently operates at higher resolution than most climate models (0.25° latitude, ~25 km).

Finding: Individual models of the atmosphere (weather forecasts) and contrails need to be linked into a seamless system so differing models and meteorological data can be tested for contrail prediction.

Recommendation: As part of a national strategy, NASA should support development and assessment of models for all scales of contrail prediction. These models range from wake vortex to global climate to contrail plume to ice-supersaturation forecast.

SUMMARY

There is currently no complete model system for predicting contrails. To optimize and test a contrail prediction system, such a modeling system needs to undergo verification as with any weather forecast system. Any contrail modeling system should be able to support stakeholder needs. The elements needed for development of an operational system to forecast and verify contrails are discussed in the Chapter 5.

REFERENCES AND FURTHER READING

Wolf, K., N. Bellouin, and O. Boucher. 2023. “Long-Term Upper-Troposphere Climatology of Potential Contrail Occurrence Over the Paris Area Derived from Radiosonde Observations.” Atmospheric Chemistry and Physics 23:287–309.

Wolf, K., N. Bellouin, and O. Boucher. 2024. “Distribution and Morphology of Non-Persistent Contrail and Persistent Contrail Formation Areas in ERA5.” Atmospheric Chemistry and Physics 24(8):5009–5024.

Zhou, C., and J.E. Penner. 2014. “Aircraft Soot Indirect Effect on Large-Scale Cirrus Clouds: Is the Indirect Forcing by Aircraft Soot Positive or Negative?” Journal of Geophysical Research: Atmospheres 119(19):11303–11320.