What changed
Some of the white trails behind aircraft spread into thin cloud layers that trap heat. They form only in particular cold, humid pockets, which raises a practical possibility: move a flight slightly up or down when conditions allow, rather than redesign the airplane.
Researchers ran a randomized crossover test with 44 commercial flights. Pilots on 22 treatment flights received recommended altitude changes based on contrail forecasts. Observers later detected 4 contrails in the treatment group and 11 in the control group—a 54.4% reduction per flight-kilometer.
What this could change for you
For a passenger, the flight would feel ordinary. The change happens in dispatch and the cockpit: a modest altitude adjustment in the right patch of sky could reduce a flight's warming effect.
The test also showed why AI can help with climate problems that are annoyingly specific. It combines weather, route, and altitude data quickly enough to point pilots toward a narrow action.
What it does not prove
This was a 44-flight feasibility study. It counted visually detectable contrails rather than directly measuring their total climate effect, and weather forecasts can miss the exact altitude where a trail forms.
Adjusted flights used about 2% more fuel, although only a small share of flights would need to move in a scaled program. Airlines must also preserve safety, traffic separation, comfort, and schedule reliability.
The bottom line
AI helped pilots avoid more than half of the detectable contrails in a small real-world trial, with a modest fuel tradeoff on adjusted flights. It is not a complete aviation-climate solution, but it is unusually actionable.
Primary research
Feasibility test of per-flight contrail avoidance in commercial aviation
Communications Engineering · 2024 · DOI 10.1038/s44172-024-00329-7
