Exhaust plumes stream out the back of aeroplanes, drawing straight white lines across the blue vault of the sky. Known as contrails, these artificial clouds are made of ice crystals that coalesce around particles of soot from engine emissions. Under certain conditions of temperature and humidity, these formations don’t dissipate into thin air but broaden out and hang around for hours.

But at the edges of these clouds is a silver lining. First, the effect is uncommon; although all flights emit soot, only a few percent cause planet-heating contrails. Second, meteorologists understand much about cloud formation, the atmospheric parameters which are required for damaging contrails to form (cool and moist). Apparently, those zones with the atmospheric conditions most prone to creating contrails can be pancake-shaped: if not necessarily round, they spread over a wide area but are not very deep.

Taken together, this atmospheric data and analysis suggests that where a contrail risk zone occurs, aeroplanes can avoid creating contrails simply by avoiding a certain volume of air.

‘Those conditions evolve over time, and based on weather forecasts, climate models, aircraft models, we try to anticipate if a planned trajectory will generate contrails, and put them into an alternative trajectory to avoid them,’ says Julien Lopez, Head of Green Operations for French aviation engineering company Thales, which is leading 10 partners in the French DECOR* project, and leading the three-year, €9mn, 20-member pan-European CONCERTO project. Both projects, scheduled to end in mid-2026, aim to test out contrail avoidance systems, among other solutions to reduce the climate impact of aviation.

Early results suggest that there could be an engineered solution to the problem posed by contrails, one that is easier to implement than switching to decarbonised fuel. So says a report published in September by US think tank the International Council on Clean Transportation. It contends that contrail avoidance is an important mitigation in short-lived climate pollutants, controls of which ‘can complement GHG [greenhouse gas] mitigation by delivering substantial near-term reductions via easier-to-implement technology solutions’.

How does it work? 
Avoiding persistent contrail zones in practice is not simply a matter of radioing a set of instructions to a pilot, who takes hold of the stick and swerves aside. Pilots must stick to flight plans pre-agreed with local authorities, which include aspects of the journey including trajectory, speed, altitude and timings, unless safety reasons (hazardous weather) or operational reasons (a delay to the start of the flight) cause a change in the plan. Although Thales offers autopilot software, contrail avoidance is being incorporated in its fuel efficiency software, called FlytOptim.

Generally speaking, the higher planes fly, the more fuel-efficient the flight. Earth’s atmosphere is similar to a swimming pool whose bottom is the surface of the earth. It’s much easier to fly the higher the flight, since there are fewer air molecules to push out of the way. On the other hand, the aeroengine must take in enough oxygen molecules to combust the kerosene fuel to create thrust, which places a ceiling on the upper edge of flight. This depends on engine technology. Most commercial jets cruise at altitudes of 35,000–40,000 ft (10.6–12.2 km).

Contrail-producing zones may occur at approximately this altitude. Faced by such an obstacle, a decision must be taken about how to avoid them: flying above or below? This is what the trial Thales flight planning software does: to avoid a contrail danger zone, it reprogrammes the flight. Most, but not all, of the time that leads to a drop in altitude. Either way, there is no other change in direction or trajectory.

In addition to performing lots of simulations – 160,000 flights – the DECOR* and CONCERTO research have also taken to the skies. Within the former project, Thales worked with project partner Amelia, a regional short-haul airline. Initially the system was tested on small commercial jets, the ERJ145, but has since expanded to a larger short-haul jet, the Airbus A320, on flights as long as Paris-Algiers. Long-haul test flights are not part of CONCERTO research, but Lopez is confident that when the system is expanded to include them, the results will be no different. ‘There is no issue at all in extrapolating the findings,’ for larger airframes, he says. ‘The final contrail impact does vary depending on engine. But if they are in a contrail-prone area, they will generate them; only the magnitude of final contrail impact will vary, but we can predict it thanks to the models.’

The researchers then verified that the adjusted flight path prevented the formation of a contrail, by comparing the path to meteorologists’ reports of the actual – as opposed to the predicted – weather, produced after the fact, to ensure the plane did not enter the high-risk zone. For additional verification, a ground-based camera scanned for persistent contrails in a small 60 km-wide section of sky on a particular high-risk route. The project also relied on climate scientists’ own validation work of their predictive models.  

The first goal of the project was to ensure that it was feasible technically to develop solutions that will be capable of providing in real time rerouting for a hundred or thousands of flights every day, says Lopez. ‘There was a problem about how much data is required for weather forecasts, flight planning and computing capabilities. But if you want to industrialise this, it is something you must tackle, and we have demonstrated that we can do that.’

He continues: ‘The second goal of the research was to demonstrate the possibility to bring new features into the dense operations of airlines without wreaking havoc. That is difficult. Dispatchers have to dispatch several flights in a constrained time; addressing fuel use and timing of operations; but technically, we have demonstrated that works too.’

As there is only so much air over any particular part of the Earth, and aeroplane stacking heights are fixed and large to provide a safe margin of error, contrail avoidance could become more difficult or even impossible in congested airspace, such as over London in the early morning as the long-haul flights come in. For this reason, the CONCERTO R&D programme aims to take a holistic approach to managing a large number of flights in a particular area.

Towards the end of the year, this system will be tested in air traffic control in ‘shadow mode’: considering in real-time the actual traffic over a huge airspace (all Sweden and Denmark), computing all the rerouting for contrail avoidance, and validating them, but not actually flying them.

Despite the positive early results of the trial, Lopez says that to convince industry once and for all, the next step would ideally involve massively scaling up the trial to cover not just one carrier, but all commercial flights over a large area of sky, such as across Continental Europe for a whole year, to prove the system can function in such complex circumstances. Another refinement would be to install contrail sensors into aeroplanes.

At what cost? 
As fuel costs are the biggest single ongoing operational expense of an airline, flight plans are designed to optimise fuel efficiency. Any change will have an effect on the total fuel bill, and therefore on the bottom line. No matter how clever the system is, a big increase in fuel costs will just not appeal to air carriers.

It is estimated that 3% extra would be the upper limit of that cost, says Lopez, based on trial results. That number was imposed by the airline participating in the trial as the maximum fuel premium that it would be willing to pay. And results from the trial found that the premium was more like 1% in practice: because of the relatively few number of flights deviated, as well as the relatively small deviation required.

But Amelia’s flexibility as a regional carrier offers it the option to compensate for that fuel premium, says Lopez. If diverted, it can reduce its flight speed slightly, arriving just a few minutes late, but having used only the quantity of fuel planned for the journey.

He adds: ‘Scientists continue to develop their models, they are never optimal. We wanted to know, are they good enough? The last five years have seen a lot of progress, and lots of people well-known in this field say that if we do focus on the big-hitter flights we will never make a mistake, because the contrail environmental impact is huge versus the fuel, and thus the CO₂ environmental impact. We might have a fuel burn increase of 3%, but that is set against more than a 25% reduction in the climate impact. That makes this a no-regret policy; it’s always a good decision. And as the science improves, we can tackle more areas to avoid them all.’

• Further reading: ‘Technology – not feedstock – is key bottleneck to net zero aviation, trade body contends’. The aviation industry’s push towards net zero carbon emissions by 2050 is not principally limited by sustainable aviation fuel (SAF) feedstock availability, but by the pace of technology rollout and infrastructure deployment, according to the International Air Transport Association (IATA).
• Find more about Eni’s first plant to produce SAF at the Gela biorefinery in Sicily. The plant has a capacity of 400,000 t/y, representing almost a third of European SAF demand in 2025 based on Wood Mackenzie forecasts, following implementation of the European Union’s ReFuelEU Aviation regulation.