Google Research and NASA JPL have released a deep learning model that maps methane plumes globally at 60 metres, finding about 50% more than human analysts and 23,000 additional plumes, including at 24 of the 25 largest-emitting landfills. Europe’s Methane Regulation ordered a satellite monitoring tool and a super-emitter alert system, but it covers only oil, gas and coal.
Google Research and NASA’s Jet Propulsion Laboratory have released a deep learning model that maps methane plumes worldwide from space, Google Research said. It flags around 50% more plumes than human analysts, and 23,000 more than the existing product.
The model is a vision transformer trained on 3.6 million simulated plumes. It reads raw spectra from EMIT, a NASA instrument on the International Space Station, at 60 metres per pixel.
Europe already runs a version of this. The Copernicus atmosphere service switched on a methane hotspot explorer in February last year, built by Dutch researchers on Sentinel-5P data.
The two are not doing the same job. Sentinel-5P covers the planet daily at roughly 5.5 by 3.5 kilometres per pixel, a region rather than a site.
Sixty metres resolves a landfill cell. The paper reports plumes at 24 of the world’s 25 largest-emitting landfills, against 17 for the NASA product it replaces. UCLA built last year’s list of the worst sites from Carbon Mapper analysis of EMIT and Planet Labs data.
Which is where Europe’s own rules stop short.
The Methane Regulation applies to oil, gas and coal. It obliges the Commission to build a global monitoring tool on satellite data, and Article 31 alerts member states to super-emitting events.
Waste is outside it. Prognos and the ifeu Institute estimated in January that municipal waste landfilled in Europe between 2022 and 2050 will release roughly 1.5 billion tonnes of carbon dioxide equivalent as methane.
Europe’s high-resolution answer is still being built. AIRMO raised EUR 5M in March to launch its first satellite in early 2027, and the Copernicus CO2M mission is due in November of that year.
What is available now costs nothing. The plume database sits on Earth Engine under a Creative Commons licence, the model is on Kaggle and the code on GitHub.
The limit is where the instrument looks. EMIT was built to map mineral dust in arid regions, and it flies on a space station orbit inclined at 51.6 degrees. Europe was never the target.
So the sharpest methane map in existence is a by-product of an American dust mission, given away.
European Earth observation has the opposite trouble, as TNW has reported: plenty of data and too few buyers. A regulator wanting facility-level methane attribution tomorrow would download Google’s file, then find the emitters it resolves best are the ones the regulation does not reach.
Get the TNW newsletter
Get the most important tech news in your inbox each week.
Facts Only
* Google Research and NASA JPL released a deep learning model for mapping global methane plumes from space.
* The model flags about 50% more plumes than human analysts.
* The model identifies 23,000 additional plumes compared to existing products.
* The model is a vision transformer trained on 3.6 million simulated plumes.
* The model reads raw spectra from the EMIT instrument on the ISS at 60 meters per pixel.
* Europe’s Methane Regulation covers oil, gas, and coal.
* The European regulation mandates a satellite monitoring tool and a super-emitter alert system.
* Sentinel-5P data is used by the Copernicus atmosphere service for methane hotspot exploration.
* The new model resolves plumes at 24 of the world’s 25 largest-emitting landfills.
* Municipal waste in Europe between 2022 and 2050 is estimated to release roughly 1.5 billion tonnes of carbon dioxide equivalent as methane.
* The EMIT instrument was built to map mineral dust in arid regions.
Executive Summary
Google Research and NASA’s Jet Propulsion Laboratory released a deep learning model that maps methane plumes globally from space. This model flags approximately 50% more methane plumes than human analysts and identifies 23,000 additional plumes compared to existing products. The model is a vision transformer trained on 3.6 million simulated plumes and reads raw spectra from the EMIT instrument on the International Space Station at 60 meters per pixel.
Europe has implemented some monitoring measures, with the Copernicus atmosphere service launching a methane hotspot explorer based on Sentinel-5P data. However, this existing system covers oil, gas, and coal emissions but does not encompass waste methane. The high-resolution mapping capability of the new model resolves landfill cells at 60 meters and identifies plumes at 24 of the world’s 25 largest-emitting landfills, surpassing the coverage of a NASA product it replaces for these sites.
The current regulatory framework in Europe focuses on oil, gas, and coal, leaving municipal waste methane emissions outside the direct scope of existing rules. Efforts are underway to develop higher-resolution solutions, with projects like AIRMO and the Copernicus CO2M mission aiming to provide this localized data. The underlying satellite instrument, EMIT, was originally designed for mapping mineral dust in arid regions, not specifically European monitoring targets.
Full Take
The narrative juxtaposes a powerful, open-source, high-resolution scientific discovery against the slow pace of regulatory response and existing data limitations. A significant pattern emerges in the tension between technological capability and governance structure: advanced analytical tools are being developed faster than the regulatory frameworks necessary to act upon that data. The release of an exceptionally sharp methane map derived from a U.S. dust mission, yet immediately becoming a public resource available on platforms like Earth Engine, highlights an asymmetry in observational advantage—the most detailed mapping exists outside the direct jurisdiction of the regulatory body concerned with emissions control.
The context reveals a structural challenge where regulation (Methane Regulation) addresses specific energy sources (oil, gas, coal) rather than diffuse waste methane, creating a gap precisely where the high-resolution data excels: landfill emissions. This suggests that even when satellite technology exists to resolve facility-level emissions, political and jurisdictional boundaries remain the primary impediment to systemic change. The distribution of this powerful information—where it is available freely versus where regulatory action is mandated—shapes agency.
The implication for human agency lies in recognizing that data availability does not equate to policy implementation. If a regulator wishes to immediately enforce facility-level methane attribution, they must navigate the existing framework, which deliberately excludes waste, rather than simply downloading the most advanced scientific product. This dynamic forces an interrogation of whether regulatory bodies are optimizing for political feasibility or true environmental impact resolution.
Bridge Questions: If high-resolution data is freely available, what institutional mechanisms exist to force adaptation when regulators operate on slower, less granular mandates? How can frameworks be designed to prioritize emission sources based on modeled risk rather than existing legal definitions? What accountability structures can bridge the gap between cutting-edge observation and actionable policy?
Sentinel — Human
The text effectively synthesizes technical findings from multiple sources to build a critical argument about the disconnect between advanced global monitoring technology and specific regional regulatory frameworks.