The smokestack is now being asked to file its own emission slip
the smokestack is now being asked to file its own emission slip: a deep point object detector locates and quantifies greenhouse gas sources from satellites in a single pass, a targeting mode of an orbital spectrometer resolved co2 hotspots at nine power plants with deep learning, and a physics guided network estimates column carbon dioxide from a geostationary satellite using sixteen spectral bands.
the atmosphere now files its own source ledger.
Context
GHGPSE-Net, in Geoscientific Model Development 2026, is a point-object-detection deep network that does detection, localization and quantification of greenhouse gas point sources together, without a segmentation step, on a single concentration map; it was trained on synthetic data from an atmospheric transport model. A GMD paper of 18 June 2025 trains a convolutional network exclusively on simulations from eight power plants in Germany, then applies it to 39 OCO-3 Snapshot Area Map images covering nine power plants in the USA, Europe and China, comparing the results with average annual reported emissions. An arXiv paper (2605.23991) develops DeepXCO2 from the GOES-East Advanced Baseline Imager, which has 16 spectral channels at about 2 km spatial resolution and 10 minute revisit.
These are three separate studies, each with its own limits. The point-source detector was trained and evaluated on synthetic data. The nine power plant study trained on simulations of eight German plants and had a relative error near 20 percent on synthetic data, with real-image results only slightly worse than that; the abstract calls the results promising, and it is a test on 39 images. The GOES paper estimates column carbon dioxide from an instrument built for other applications, and its accuracy figures were not read. The atmosphere now files its own source ledger is the author's framing, and reported emission verification still depends on independent inventories.
Related work
- Earlier note in this series: the monsoon ↗Same pattern: a natural system forecast by learned models.
Watch next
- The GOES paper's validation results and any peer-reviewed version.
Sources
- GHGPSE-Net (Geoscientific Model Development, 2026)gmd.copernicus.org
- Quantification of CO2 hotspot emissions from OCO-3 SAM CO2 satellite images using deep learning (Geoscientific Model Development, 18 Jun 2025)gmd.copernicus.org
- Quantification of atmospheric carbon dioxide from GOES East (arXiv 2605.23991)arxiv.org
Provenance
The note above is reproduced unedited from the original post, first published on Threads on 4 October 2026 at 18:52 IST. Sources are the papers and datasets the note draws on.
View the original post ↗Embed this note
More notes
The air is now being asked to keep its own ledger
the air is now being asked to keep its own ledger: ecmwf’s aifs compo becomes the first ai model to forecast atmospheric composition globally every three hours, cleanair simulates 365 days of pm2.5 over china in ten seconds, and a unified framework maps six pollutants at one kilometer across the whole country. the air now files its own composition report.
read the note →The current is now being asked to draw its own map
the current is now being asked to draw its own map: china’s langya 2.0 predicts six ocean phenomena including internal waves and mesoscale eddies, a deep net called wenhai resolves eddies globally with air sea flux formulas built in, and scripps infers surface currents from the way temperature patterns deform in satellite images. the ocean now files its own circulation report.
read the note →The soil is now being asked to report its own carbon
the soil is now being asked to report its own carbon: a nix color sensor paired with generative data augmentation predicts soil organic carbon without a lab, random forest drives 74 percent of soil health mapping studies, and sentinel 2 tracks five year carbon change across france and italy from 922 samples. the dirt now files its own carbon account.
read the note →