The forecast is being learned, not solved
the forecast is being learned, not solved: deepmind’s weathernext 3 runs hourly forecasts at five-kilometer resolution into search and maps, germany’s dwd has run the aicon gnn operationally since march, and a radar-only system now nowcasts precipitation 410 minutes ahead. the equation solver is becoming the baseline, not the frontier.
the weather is becoming a model you can argue with.
Context
Google DeepMind describes WeatherNext 3 as the first global weather model that generates forecasts every hour of the day, delivered into Search, Maps and Gemini. Google for Developers lists 15-day probabilistic forecasts initialized hourly with 64 ensemble members at 5 km resolution.
It ingests live geostationary satellite observations as a direct input.
Hourly forecasts at 5 km resolution in Google products match the documentation. Germany's DWD running the AICON graph network operationally since March, and a radar-only system nowcasting precipitation 410 minutes ahead, were not seen in the sources read, so unsupported here, not refuted.
Related work
- Google DeepMind: WeatherNext 3 ↗Product page.
- Google for Developers: WeatherNext models ↗Model guide.
- Google for Developers: WeatherNext ↗Overview.
Watch next
- Find DWD's AICON documentation.
Sources
- Google DeepMind, WeatherNext 3deepmind.google
- Google for Developers, modelsdevelopers.google.com
- Google for Developers, overviewdevelopers.google.com
Provenance
The note above is reproduced unedited from the original post, first published on Threads on 2 October 2026 at 17:51 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 →