FrameIt

Focused Regional Analysis of Modeled Events with Integrated Tracking

A modular Python tool for subdomain extraction, object tracking and diagnostics from atmospheric and oceanic model outputs.

Features

• Subdomain extraction — extract fixed or mobile subdomains from large model outputs (AROME, MesoNH)
• Object tracking — follow meteorological objects (tropical cyclones) with multiple tracking methods:
  • Prescribed track
  • Pressure-wind algorithm
  • Fixed box
  • AI-based tracking (utrack)
• Normalized outputs — standardized NetCDF outputs to facilitate inter-model and inter-simulation comparisons, and enable AI dataset creation
• Data reduction — reduce data volume for easier manipulation and sharing with partners
• Polar projection — stereographic projection centered on the tracked object
• Modular architecture — plug in custom trackers and diagnostics
• Model support — currently compatible with AROME and MesoNH model outputs,ocean and wave model extension planned

Installation

FrameIt requires conda to manage dependencies, as some packages (e.g. xesmf) are not available via pip.

1. Clone the repository

git clone https://github.com/meteofrance/frameit
cd frameit

2. Create and activate the conda environment

conda env create -f environment.yaml
conda activate frameit_env

3. Install FrameIt

pip install .

Requirements

• Python 3.10+
• conda (via Miniconda or Anaconda)
• Compatible with AROME and MesoNH model outputs

Quick start

1. Copy and edit the example configuration file:

cp example/conf_example.yml my_config.yaml

2. Edit my_config.yaml to match your simulation setup (input files, tracking method, subdomain size, output paths).

3. Run FrameIt:

frameit run my_config.yaml

That's it! FrameIt will extract the subdomain, track the object and write normalized NetCDF outputs to the directory specified in frameit_output_dir.

Documentation

The full documentation is available at: [https://meteofrance.github.io/frameit/]

It includes:

• A detailed description of all configuration parameters
• A description of the available tracking methods
• A description of the output files

Citation

If you use FrameIt in your research, please cite it as follows:

@software{soufflet_frameit_2026,
  author    = {Soufflet, Clément and Hoarau, Kevin and Colomb, Adrien and Laxenaire, Rémi},
  title     = {FrameIt: Focused Regional Analysis of Modeled Events with Integrated Tracking},
  year      = {2026},
  publisher = {Zenodo},
  doi       = {10.5281/zenodo.22128192},
  url       = {https://doi.org/10.5281/zenodo.22128192}
}

This DOI always resolves to the latest release. To cite the exact version you used, add the corresponding version field and replace the DOI with the version DOI listed on the Zenodo record.

License

FrameIt is released under the Apache 2.0 License.

Copyright 2026 Soufflet Clément - Météo-France

Authors

FrameIt was developed at Météo-France.

• Clément Soufflet — Météo-France
• Kevin Hoarau — CNRS
• Adrien Colomb — Météo-France
• Rémi Laxenaire — Météo-France