Google and NASA JPL Release AI Model Mapping Methane Plumes Worldwide
Google Research and NASA JPL have released a deep learning model that maps methane plumes globally from space, locating plumes at 24 of the world's 25 worst-emitting landfills. Europe's Methane Regulation, which ordered a satellite monitoring tool, excludes sectors like waste management.
The model, a vision transformer trained on 3.6 million simulated plumes, reads raw spectra from NASA's EMIT instrument on the International Space Station at 60 meters per pixel.
While Europe has a similar methane hotspot explorer built by Dutch researchers, it covers the planet daily at a lower resolution (5.5 by 3.5 kilometers per pixel) compared to the new model's focus on specific landfill cells at 60 meters resolution.
The new model finds about 50% more plumes than human analysts and 23,000 additional plumes, including at 24 of the 25 largest-emitting landfills.
The plume database, available on Earth Engine under a Creative Commons license, the model on Kaggle, and the code on GitHub, is a free resource. However, Europe's own rules stop short in addressing methane emissions from waste landfills, despite projections of significant future emissions.
Airmo and Copernicus CO2M are European initiatives launching satellites in 2027 to enhance methane monitoring capabilities. For now, Google's model offers a powerful tool for global methane mapping.