Welcome to Anemoi’s documentation!

The Anemoi framework provides a complete toolkit to develop data-driven weather models – from data preparation through to inference. The framework is composed of several packages which target the different components necessary to construct data-driven weather models. To aid development and deployment, each package collects metadata that can be used by the subsequent packages. The framework builds upon on established Python tools including PyTorch, Lighting, Hydra, Zarr, Xarray and earthkit.

Figure showing the structure of the Anemoi packages

Possible uses of Anemoi are for developing:

  • Global deterministic forecast models

  • Regional deterministic forecast models (limited-area models or stretched grids)

  • Coupled atmospheric and ocean models

  • Downscaling models

  • Ensemble/probabilistic models

License

Anemoi is available under the open source Apache License.

How to Cite Anemoi

If you use Anemoi in your work, we recommend you cite the following paper as the recommended reference: Lang, Simon, et al. “AIFS – ECMWF’s Data-Driven Forecasting System.” arXiv, 2024.

BibTeX:

@article{lang2024aifsecmwfsdatadriven,
title={AIFS -- ECMWF's data-driven forecasting system},
author={Simon Lang and Mihai Alexe and Matthew Chantry and Jesper Dramsch and Florian Pinault and Baudouin Raoult and Mariana C. A. Clare and Christian Lessig and Michael Maier-Gerber and Linus Magnusson and Zied Ben Bouallègue and Ana Prieto Nemesio and Peter D. Dueben and Andrew Brown and Florian Pappenberger and Florence Rabier},
year={2024},
eprint={2406.01465},
archivePrefix={arXiv},
primaryClass={physics.ao-ph},
url={https://arxiv.org/abs/2406.01465}
}

Getting Started

Contributing

Anemoi Package Documentation