A recent joint experiment by Google and American Airlines has shown promising results in the fight against aviation emissions.
The experiment focused on aircraft contrails, the icy condensation clouds created by plane exhausts – also called vapor trails.
When humid air mixes with plane exhaust, it can create visible ice crystals that accelerate global warming by reflecting heat and light.
Google’s experiment used AI to direct planes around air vulnerable to contrail formation, resulting in net CO2 benefits.
What are contrails?
Contrails emerge when an aircraft engine’s kerosene exhaust binds with water particles in humid air, encouraging ice formation. These ice particles reflect light and heat, promoting global warming.
While seemingly benign, contrails have been associated with several negative impacts on the climate:
- Contribution to global warming: The impact of contrails on global warming is two-fold. Firstly, they trap outgoing longwave radiation from the Earth’s surface, creating a warming effect, much like other greenhouse gasses. Secondly, under specific atmospheric conditions, contrails can evolve into extensive cirrus cloud formations, which have a net warming effect on the Earth’s surface.
- Increased heat conservation: Although contrails and the clouds they can form reflect some of the sun’s energy back into space (a cooling effect), they also trap some of the Earth’s outgoing heat (a warming effect). On balance, the warming effect outweighs the cooling, leading to an overall contribution to global warming.
- The extent of the effect: While individual contrails are ephemeral, the cumulative impact of many flights can result in large areas of the sky covered by contrail-induced cirrus clouds.
How did Google’s experiment work?
Google adjusted flight paths using Google’s AI-based predictions, cross-referenced with Breakthrough Energy’s open-source contrail models.
The models directed planes around, over, or under areas likely to produce contrails. In some cases, the adjustments needed are merely a few meters.
They analyzed the results from 70 test flights over six months.
By utilizing AI models to help pilots select different flying altitudes, the researchers achieved a 54% reduction in contrail production.
Carl Elkin and Dinesh Sanekommu, both from Google, stated, “This is the first proof point that commercial flights can verifiably avoid contrails and thereby reduce their climate impact.”
The experiment also found that individual planes consumed 2% more fuel to avoid contrail formation, but across an entire airline’s operations, increases in fuel are a cumulative 0.3%, given that most contrails are generated by a limited number of flights.
Google estimates that the cost of evading contrails could range between $5 and $25 per tonne of CO2 equivalent, which results in a net benefit.
AI is powering innovation in climate science, although skepticism remains over the environmental impacts of the technology, as the data centers and hardware required to host and train models require enormous quantities of energy.