Glossary

anemoi: The ECMWF-led machine learning weather prediction software stack used by EAGLE for training and inference. See the anemoi-core repository.
AORC: Analysis of Record for Calibration. A NOAA gridded meteorological forcing dataset used for land surface and hydrologic applications.
checkpoint: A saved snapshot of a trained machine learning model that can be used to resume training or perform inference.
CONUS: Continental United States.
EAGLE: Experimental AI Global and Limited-area Ensemble forecast system.
ECMWF: European Centre for Medium-Range Weather Forecasts.
ERA5: The fifth-generation atmospheric reanalysis produced by the European Centre for Medium-Range Weather Forecasts (ECMWF).
GEFS: Global Ensemble Forecast System. NOAA’s global ensemble weather prediction system.
GFS: Global Forecast System. See the NOAA GFS page.
HRRR: High-Resolution Rapid Refresh. See the NOAA HRRR page.
inference: The workflow stage that uses a trained model checkpoint to generate forecasts.
LAM: Limited-area model.
MET: Model Evaluation Tools. A community-developed verification package used to evaluate numerical weather prediction and machine learning weather forecasts.
Miniforge conda: The Miniforge-based Conda installation used by EAGLE runtime environment setup. See the Miniforge repository.
MPI: Message Passing Interface. A standard used for parallel computing across multiple processes.
netCDF: Network Common Data Form. A file format commonly used for storing and sharing scientific array-based data.
NRT: Near real-time. Refers to workflows that generate forecasts using the latest available input data shortly after it becomes available.
PrepBUFR: A BUFR-formatted observational dataset used for numerical weather prediction and forecast verification.
prewxvx: The component used by EAGLE to prepare forecast output for wxvx verification steps.
ufs2arco: The component used by EAGLE to prepare Zarr-formatted forecast and analysis data. See the ufs2arco repository.
Ursa: A NOAA RDHPCS platform currently supported by EAGLE. See the NOAA RDHPCS Ursa page.
uwtools: The workflow utility framework used by EAGLE to compose configuration, validate schemas, and execute drivers. See the uwtools repository.
visualization: The workflow stage that plots verification results derived from processed forecast statistics.
wxvx: The verification component used by EAGLE to run MET-based statistical verification tasks.
YAML: YAML Ain’t Markup Language. A human-readable data serialization format commonly used for configuration files.
Zarr: A chunked, cloud-friendly array storage format used in EAGLE data preparation workflows.