MAPL-EMIT uses AI to spot methane plumes from space

Google and NASA JPL combine hyperspectral imaging and deep learning to locate methane emissions and make their data accessible.

Presented on 1 September, MAPL-EMIT analyses the observations of the EMIT instrument installed on the International Space Station. The model identifies the contours of the methane plumes, estimates their enrichment and searches for their point of origin taking into account the surrounding landscape.

The team reports the detection of 84% of the plumes annotated by experts in its evaluation. The training mobilizes 3.6 million simulated plumes. The model, an inference library and data are published to facilitate the search. False positives remain possible; the confidence levels accompanying the results must therefore be taken into account.