A new AI-powered tool could put a major dent in atmospheric pollution by automatically detecting methane "super emitters" from space.
Researchers at the University of Oxford partnered with the team at Trillium Technologies to develop the technology. It uses machine learning to analyze satellite data and pinpoint the biggest methane offenders with over 81% accuracy, according to findings published in Nature Scientific Reports, as summarized by the university.
While carbon dioxide often steals the pollution spotlight, methane can actually be up to 80 times more powerful at trapping heat, the university reported. The good news — methane only sticks around for around 12 years, compared to centuries for carbon, according to the International Energy Agency.
That means cutting back on methane could put a serious speed bump ahead of rising global temperatures. Some quick math shows it could prevent nearly 0.3 degrees Celcius (more than 0.5 degrees Fahrenheit) of warming over just the next 20 years, the university reported based on information from the UN Environment Programme.
This new AI sidekick makes hunting down methane leaks easy. Before, it took a ton of time and elbow grease to comb through satellite images. The methane was tricky to spot since it's invisible to human eyes and most satellite cameras, per the university.
But with hyperspectral satellites, Oxford's brainiacs trained their AI to filter out the noise and zero in on methane's fingerprint. The AI sifted through over 167,000 satellite images captured by NASA to learn the ropes.
The researchers aren't keeping this intel to themselves. They've made the AI's code and training data freely available online so other scientists can join the methane-busting party.
Next, they're looking at installing the AI right on the satellites. That way, it can send an alert back to Earth as soon as it catches a whiff of a methane leak.
Lead researcher Vít Růžička is enthusiastic about this new development, saying, "This would allow for a swarm of satellites to collaborate autonomously: an initial weak detection could serve as a tip-off signal for the other satellites in the constellation to focus their imagers on the location of interest."
Oxford professor Andrew Markham, the study's supervisor, is just as optimistic. "This approach could easily be extended to other important pollutants, and building on earlier work, our ambition is to run these approaches on-board the satellites themselves, making instant detection a reality," Markham said in the university's report.
Elsewhere, researchers are working on technology to help detect methane in animal barns, as other companies are also working on using satellite images to detect methane leaks. All of these efforts aim to help improve air quality and the health of the public.
With tools like this in our toolkit, cutting back on methane could become much simpler. So, keep your eyes on the skies — this methane-mapping AI might just become humanity's next hero.
Join our free newsletter for weekly updates on the coolest innovations improving our lives and saving our planet.
Cool Picks
