Google's AI has found a way to successfully cut climate warming from airplane contrails over the Pacific Ocean in a new trial. After using AI-based predictions and open-source contrail models, Google's systems helped pilots avoid pockets of humid air that can generate the most persistent contrails, cutting their warming effect by up to 40%.
Contrails are the white lines of condensation that appear in the wake of airplanes as they pass through colder air in the upper atmosphere. Mostly, these dissipate quickly, but in some cases, they don't, trapping heat and contributing toward climate change. Estimates suggest that contrails may be responsible for up to a third of all aviation-related warming.
Google has now completed several trials showing that its AI can substantially mitigate this, with little additional expense or effort. The idea uses AI-based analysis of satellite imagery to model where humid air is likely to produce more persistent contrails, and then has airline pilots adjust their altitude slightly to avoid them.
Google's latest test with Cathay Pacific used the models on 80 flights, including several over the contrail-heavy Hong Kong-to-Singapore route. It now claims this helped cut the warming impact of those contrails by roughly 40%.
This comes after Google conducted a 70-flight trial in the US earlier this year with American Airlines and found that pilots reduced contrail generation by 54%. It's not clear how significant that effect would be on contrail warming, but Google reported that it would require only an additional 2% fuel use.
Increased fuel use may reduce the climate benefit and will need to be accounted for in overall net-positive estimates, but Google suggested that not all flights would need to make these adjustments. It claims that by focusing on the most impactful flights, the increase in fuel usage could be as little as 0.3% across an airline's operations.
There's already additional interest in this initiative in Europe. In mid-August, Google launched Operation Blue Skies, in cooperation with several major UK entities, to target the Shanwick Oceanic corridor in the North Atlantic. Over the next 30 months, the 10,000+ flights that pass through it during trial hours will be required to adjust their altitudes to avoid generating contrails.
The UK's Met Office, the NATS air navigation service provider, Imperial College London, and the University of Cambridge will help analyze the data, along with Google and other organizations, to determine the program's effectiveness. If it does have a measurable impact, this could serve as a blueprint for future trials and the expansion of the scheme to further reduce aviation-related climate warming.
In the near future, Google is planning a larger-scale trial with Cathay Pacific, involving more flights, in hopes of accelerating contrail research and delivering fast, effective impacts on climate change mitigation. Still, AI use has its own climate concerns. Google's own emissions jumped massively over the past few years due to its AI data centers.