The blocked high is now being asked to file its own persistence report
the blocked high is now being asked to file its own persistence report: three models spanning physics, hybrid and pure data driven stacks all reproduce winter blocking frequency at one to two week leads, yet their biases diverge sharply by weeks three to four, and a four model evaluation finds data driven forecasters still lean on mjo and enso states to hold a blocking pattern.
the atmosphere now files its own stagnation ledger.
Context
An evaluation of four machine learning models (Pangu-Weather, FuXi, GraphCast and NeuralGCM) against ECMWF HRES for wintertime atmospheric blocking (March 2026) finds a model that is accurate on 500 hPa height but underestimates blocking frequency, particularly in the Pacific, with NeuralGCM best at 73.68 percent and 39.59 percent accuracy at 5 and 9 day leads. A July 2026 paper studies subseasonal prediction of wintertime North Pacific blocking in AI-based models, and a Journal of Geophysical Research paper covers subseasonal forecasting and MJO teleconnections in machine learning weather prediction models.
The four-model evaluation matches the post's four model evaluation, but its lead times are 5 and 9 days and its skill figures were not tied to the post's statement that data-driven forecasters lean on MJO and ENSO states. The three-model physics, hybrid and data-driven comparison with diverging weeks three to four biases was not located. These are authors' own evaluations of past winters. The atmosphere now files its own stagnation ledger is the author's take.
Watch next
- The source for the three-model weeks three to four comparison.
Sources
- Evaluating machine learning and numerical models for wintertime atmospheric blocking prediction (2026)exa.ai
- Subseasonal prediction of the wintertime North Pacific blocking in AI-based weather models (2026)exa.ai
- Subseasonal forecasting and MJO teleconnections in machine learning weather prediction models (JGR)onlinelibrary.wiley.com
Provenance
The note above is reproduced unedited from the original post, first published on Threads on 4 October 2026 at 20:40 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 →