A multi-model archive of long-range reforecasts from AI weather models, initialized 2000–2025 and rolled out to 50 days — beyond the models' traditional forecast lengths — at daily cadence on a global 0.25° grid. The archive supports benchmarking of AI weather models, training of downstream ML models, and multi-model blending. It also powers AI-Almanac, UChicago-LAUDE's interactive, human-centered forecast platform. Real-time forecasts initialized from operational ECMWF IFS analyses are planned alongside the reforecasts.
This repository holds the source for the archive's documentation site,
published with GitHub Pages from docs/ at
https://envfluids.github.io/reforecast/. It does not contain the
forecast data itself.
Full data documentation lives in docs/: an
overview of the models, archive structure,
zarr format, and sizing methodology, plus
documentation of every model's variables, units, temporal aggregation, and
measured storage cost.
One zarr store per model per initialization
(forecasts_<model>/init_YYYYMMDDT00.zarr), float32, dimensions
(time, [number,] prediction_timedelta, lat, lon).
Adam Marchakitus — marchakitus@uchicago.edu — University of Chicago