Advanced methane emitters detection using satellite data and AI techniques

Abstract:

The accurate measurement of ground-level methane emissions from low-Earth orbit is a challenging but urgent task for efforts to mitigate climate change and address immediate industry sector needs.

Bluefield developed a proprietary software algorithm suite called BFX-2 to detect surface level methane emissions using satellite measurements (like European Space Agency’s Sentinel-5P “Tropomi”). It combines advanced image processing tools and modern atmospheric models and can automatically track abnormal methane levels over the whole surface of the Earth while limiting false alarm using artificial intelligence and machine learning techniques.

Using this tool, Bluefield already discovered dozens of significant methane releases (>100 tons/day) over the US territory as well as other parts of the world. Outreach has begun to help pinpointing the exact origin of emission points and understand the cause of the release (see Bloomberg recent article[1]).

Bluefield continues developing its own microsatellite constellation that will bring higher spatial resolution and better sensitivity, enabling the location and quantification every critical methane emitting site on Earth. The planned sensors’ sensitivity will detect over 95% of industrial emissions and pinpoint the location of the emitting source within 20 meters. This augmented satellite data will feed the BFX-2 analysis tool that allows directly seeing what was once invisible and channel resources to reduce emissions quickly and effectively.

[1] July 27, 2020 www.bloomberg.com/news/articles/2020-07-27/no-one-is-owning-up-to-releasing-cloud-of-methane-in-florida