A new satellite AI can spot methane leaks from space twice as effectively as a human expert.
April 15, 2026
Original Paper
Global monitoring of methane point sources using deep learning on hyperspectral radiance measurements from EMIT
arXiv · 2604.10094
AI-generated illustration
The Takeaway
Methane is a 'super-pollutant' that is 80 times more potent at warming the planet than CO2, but spotting small leaks from gas lines or landfills is incredibly difficult. This paper introduces a vision-based AI (MAPL-EMIT) that scans satellite data and automatically identifies methane plumes with terrifying accuracy. It found twice as many plausible leaks as human analysts did looking at the same data. This turns our satellites into a global, automated 'leak police' that can pinpoint exactly who is polluting the atmosphere in real-time. For regular people, this means we finally have a way to hold companies accountable for their greenhouse gas emissions from 250 miles above the Earth.
From the abstract
Anthropogenic methane (CH4) point sources drive near-term climate forcing, safety hazards, and system inefficiencies. Space-based imaging spectroscopy is emerging as a tool for identifying emissions globally, but existing approaches largely rely on manual plume identification. Here we present the Methane Analysis and Plume Localization with EMIT (MAPL-EMIT) model, an end-to-end vision transformer framework that leverages the complete radiance spectrum from the Earth Surface Mineral Dust Source I