Operation Blue Skies: Using AI to reduce aircraft contrails over the North Atlantic

Operation Blue Skies is a 30-month trial that will use AI models and satellite data to reduce aircraft contrails over the eastern North Atlantic. For builders, the project shows how machine learning can inform real-time operational decisions in aviation, but the results are years away and depend on safety, forecasting accuracy, and policy adoption.

What Operation Blue Skies is testing

The project brings together Google, the Met Office, the UK Department for Transport, air traffic control provider NATS, the University of Cambridge, Imperial College London, and Contrails.org. Funding comes from the UK government's ATI Programme, with Google contributing AI models, machine learning, and satellite imaging pro bono at an estimated value of £1.4 million TechRadar.

The trial focuses on the Shanwick Oceanic zone, where about 5% of global contrail warming occurs. Over two four-month periods in the winters of 2026-2027 and 2027-2028, researchers will observe roughly 10,000 flights. Some aircraft will receive slight route deviations to avoid pockets of cold air where contrails form, while others serve as a control group.

How AI fits into contrail reduction

Google's role is to provide AI models that learn from the gathered data and build a picture of contrail formation over the North Atlantic. The Met Office is developing new contrail forecasts, and Contrails.org is designing trial parameters and forecast assessments. Academic partners independently verify the data and outcomes.

The core idea is that AI can predict where contrails will form based on atmospheric conditions, then recommend flight path adjustments that minimize contrail creation while keeping the CO2 penalty low. Google's technical lead, Dr. Paul Hodgson, told the BBC that contrail warming is responsible for roughly a third of aviation climate warming, while the CO2 penalty from diversions is less than 1% TechRadar.

What this means for AI builders

For teams building AI products in environmental monitoring, satellite imagery analysis, or operational decision-making, Operation Blue Skies is a concrete example of how machine learning can be applied to a high-impact climate problem. The project requires accurate forecasting, real-time routing optimization, and integration with safety-critical air traffic systems.

However, this is a research trial, not a commercial product. The timeline is long: formal trials don't start until winter 2026-2027. Builders should watch for published results and open datasets, but immediate product opportunities are limited. The project also highlights the importance of independent verification: academic partners will validate data and outcomes, which is a pattern worth replicating in other AI-for-climate initiatives.

Caveats and unknowns

Several factors could limit the project's impact. First, it covers only the eastern North Atlantic corridor, which represents a small fraction of global contrail warming. Second, the success depends on forecast accuracy and the ability to safely reroute flights under NATS oversight. Third, the findings may not generalize to other airspaces with different traffic patterns and weather conditions.

Google's contribution is pro bono, so the commercial viability of AI-driven contrail reduction is not yet proven. If the trial succeeds, the findings could become standard operational procedure, but that requires regulatory adoption and airline buy-in. For now, Operation Blue Skies is a well-structured experiment that builders should follow, not a shift in the current aviation landscape.