Google Will Use AI to Avoid Contrails That Warm the Planet

AI to avoid contrails will be tested over one of the world’s busiest air routes. Google and a British consortium will use artificial intelligence, weather forecasts, and satellite imagery to identify areas over the North Atlantic where certain contrails may contribute to warming and selectively adjust the flight paths of some aircraft before they pass through those areas.

The project will run for 30 months in Shanwick airspace, which covers the eastern portion of the North Atlantic corridor. According to information released by Google, this region accounts for around 5% of global warming associated with contrails. The objective is not to eliminate contrails from every flight, but to determine whether small route deviations can avoid atmospheric conditions that are particularly favorable for the formation of persistent contrails with a climate impact.

The initiative therefore introduces a different approach within efforts to reduce aviation’s impact: acting on where an aircraft flies, while continuing to develop sustainable fuels, more efficient engines, and new propulsion technologies.

Why can some aircraft contrails warm the atmosphere?

The white lines that occasionally remain behind an aircraft are known as condensation trails, or contrails. Their formation is related to the atmospheric conditions encountered by aircraft at high altitudes.

During combustion, engines generate water vapor and particles. When hot exhaust gases come into contact with sufficiently cold and humid air, the vapor can condense and freeze around small particles, forming ice crystals. From the ground, the result can be seen as a white line trailing behind the aircraft.

However, not all contrails behave in the same way. Some disappear quickly and have a limited climate effect. Others form in atmospheric regions that are cold and humid enough for them to persist for longer periods, spread, and develop into structures resembling cirrus clouds.

This is where the problem that Operation Blue Skies aims to address emerges. Persistent contrails can alter the Earth’s energy balance. Although they can also reflect some solar radiation during the day, they are capable of trapping infrared radiation that would otherwise escape into space. The balance between these processes can produce a net warming effect, particularly under certain atmospheric conditions and at night.

Google states that warming contrails account for approximately one-third of aviation’s total climate impact. This estimate helps explain why avoiding even a fraction of the most harmful contrails could become a complementary tool for reducing the climate impact of air transportation.

AI to avoid contrails will identify where not to fly

The technological distinction of the project lies in attempting to predict, before an aircraft passes through an area, where conditions favorable for the formation of persistent contrails exist.

Google UK will develop artificial intelligence models capable of generating these forecasts. The system will combine machine learning capabilities with weather information and satellite imagery to identify regions of airspace where warming contrails are likely to form.

Once one of these regions has been detected, certain flights that would normally pass through it could receive a slight route adjustment.

The principle is relatively straightforward: if an atmospheric layer favorable to producing a persistent contrail occupies only a limited portion of the route, it may be more efficient to avoid that area than to allow the aircraft to generate a contrail capable of persisting for hours.

The complexity arises when applying this concept to real-world air traffic. Any change must take into account safety, aircraft separation, airspace availability, controller workload, additional fuel consumption, and weather conditions.

For this reason, Operation Blue Skies will not be solely an artificial intelligence experiment. It will also be an operational air traffic management trial.

10,000 flights a year will provide a laboratory over the North Atlantic

The program will take place in Shanwick airspace, an extensive region covering the eastern portion of the North Atlantic transatlantic routes.

Approximately 10,000 flights pass through this area during the trial hours each year, although only a small proportion of aircraft expected to enter regions sensitive to contrail formation will be rerouted.

This feature is essential to understanding the experiment: the goal is not to indiscriminately redesign the routes of thousands of aircraft.

The strategy is to identify situations in which a relatively small intervention could prevent a contrail with a potentially significant climate impact.

The program includes two operational trials lasting approximately four months each within its 30-month duration. This structure will make it possible to compare forecasts, actual operations, and subsequent observations to determine whether the route deviations effectively prevented the predicted contrails.

Satellites will verify whether the strategy actually works

Predicting a contrail is only the first half of the problem. The second is demonstrating that it would have occurred and verifying whether the route modification successfully prevented it.

To do this, Google will use satellite imagery and machine learning to track and verify the results.

The project also involves several organizations with distinct responsibilities. NATS, the UK air navigation service provider, will manage operations within the trial airspace, air traffic control coordination, safety assessments, and controller training.

Contrails.org will lead forecast assessment, trial design, and data analysis, while the Met Office will develop UK contrail forecasting capabilities.

Academic evaluation will also play an important role. Imperial College London and the University of Cambridge will analyze the results and provide independent verification of the climate impact.

This structure will make it possible to compare what AI predicts with what actually happens in the atmosphere.

Google contributes AI and £1.4 million in technology to the project

Google UK’s participation will be pro bono. The company will contribute approximately £1.4 million in kind through artificial intelligence research expertise, engineering, and high-performance computing infrastructure.

The program’s funding combines contributions from industry partners with resources from the UK Department for Transport through the ATI Programme. The public funds allocated to the project will go to academic participants, nonprofit organizations, and aviation industry partners.

The trial also builds on previous experience.

Google had previously worked with American Airlines, EUROCONTROL, and FlightKeys on research into contrail prediction and avoidance. Those efforts made it possible to use forecasts to make targeted adjustments during normal operations and explore whether routes could be modified without substantially disrupting flight operations.

Operation Blue Skies now increases the scale of the concept by extending it across an entire oceanic airspace under a state-backed program.

Avoiding contrails does not replace aviation decarbonization

The project raises an important distinction. Avoiding a condensation trail does not mean eliminating the CO₂ emissions generated by the aircraft. Sustainable aviation fuel represents another pathway currently being developed by the industry to reduce its carbon footprint.

An aircraft that slightly modifies its route continues to consume fuel and generate emissions. For this reason, contrail management should be understood as a complementary tool rather than a substitute for efforts aimed at reducing emissions directly associated with combustion.

The industry continues to need efficiency improvements, fleet renewal, lower-carbon aviation fuels and new technologies capable of accelerating aviation decarbonization and, when technically feasible, new propulsion technologies.

However, there is an important difference in terms of timing. Transforming fuels, engines, and airport infrastructure can require considerable investment and extended periods of time. Contrail management seeks to address part of aviation’s climate impact by using existing aviation infrastructure, complemented by improved forecasting capabilities.

The actual benefit will depend on whether the route deviations generate a climate benefit greater than any additional impact caused by changes in flight distance or fuel consumption.

The North Atlantic will test a new way of managing the skies

The upcoming trials will have to demonstrate whether AI to avoid contrails can work beyond controlled experiments and become a tool compatible with large-scale aviation operations.

The question will not only be whether algorithms can anticipate where a contrail will appear. It will also be necessary to determine whether controllers and aircraft can use that information safely, whether route modifications are sufficiently small, and whether satellites confirm a real reduction in the expected warming effect.

If the results are favorable, Operation Blue Skies could provide a model for incorporating climate information into future airspace management.

The conceptual shift is significant. For decades, route planning has been determined primarily by safety, weather, efficiency, time, and fuel. This experiment will add another variable: the climate effect of flying through a particular region of the atmosphere.

Thus, an apparently simple white line behind an aircraft becomes the focus of an experiment in which artificial intelligence, meteorology, satellites, and air traffic control will attempt to determine whether small changes in the sky can produce a measurable climate difference.

Source: Aerospace Testing International.