The eddy is now being asked to file its own current log
the eddy is now being asked to file its own current log: a u net trained on global ocean models now maps surface currents at submesoscale resolution using geostationary satellite temperature data, transformer models classify mesoscale eddies directly from sea surface temperature maps, and a multiscale yolo framework detects eddies and internal waves together in spaceborne sar imagery.
the vortex now files its own rotation record.
Context
Small ocean eddies and fronts, the submesoscale, move heat, nutrients and carbon between the surface and the deep, but they are hard to observe. The Nature Geoscience paper trains a U-Net on global ocean model output and then applies it to sea surface temperature from geostationary satellites, which look at the same place continuously, to estimate surface currents at that fine scale.
The other two results are about finding the structures themselves: a transformer that classifies mesoscale eddies from sea surface temperature maps, and a YOLO-based detector that finds eddies and internal waves together in radar images from space.
Training on model output means the network learns physics the model contains, then reads it back from observations the model never saw. That is the strength and the risk: the result inherits the model's blind spots.
Related work
- Training code for GOFLOW (Zenodo) ↗The oceanic surface velocity generator from geostationary data.
- Between the clouds: an unprecedented view of ocean surface currents (NASA talk, PDF) ↗Background on the geostationary current work.
Watch next
- Comparison of these current estimates with drifters and high-resolution radar.
- Whether the method holds in cloudy regions where infrared temperature is blocked.
Sources
- An unprecedented view of ocean currents from geostationary satellites (Nature Geoscience)nature.com
- Transformer-Based Deep Learning for Mesoscale Eddy Detection in Sea Surface Temperature (IEEE)ieeexplore.ieee.org
- YOLO-Based Detection of Ocean Eddies and Internal Waves in Multisource Spaceborne SAR Imageryexa.ai
Provenance
The note above is reproduced unedited from the original post, first published on Threads on 4 October 2026 at 14:37 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 →