openAMUNDSEN-DA

An ensemble-based snow data assimilation framework for the open-source snow-hydrological model openAMUNDSEN.

Software scope

openAMUNDSEN-DA is an open-source data assimilation framework built around openAMUNDSEN. It preprocesses configured snow observations, prepares deterministic assimilation sequences, executes an ensemble particle-filter workflow and validates the resulting tables, grids, plots, maps and report.

The documentation is deliberately technical. The openAMUNDSEN-DA framework and its Rofental application are described in the EGUsphere preprint by Wagner et al. (2026). This repository documents the software interface and operational workflow.

openAMUNDSEN-DA technical workflow from mounted setup inputs through observation summaries and prepared steps to validated results

The host setup is mounted at /data. openAMUNDSEN-DA summarizes configured snow observations, prepares inspectable event inputs and writes a validated result set.

Start here

Installation explains the Python package and wheel-based Docker image.
Input data defines required model, forcing and observation inputs without prescribing product-generation algorithms.
Configuration documents the strict setup, project and step ownership boundary.
Running the model covers the supported single-domain and subdomain command sequences.
Output data defines the manifest, compact NetCDF, diagnostics and cleanup contract.
Example data sets describes the shipped Rofental example.
How to Use is the reviewed, continuous Rofental walkthrough.

Citation

If you use openAMUNDSEN-DA in your work, please cite the preprint and the archived software version you used. Find release records through the openAMUNDSEN-DA concept DOI.

Wagner, F., Rottler, E., and Strasser, U.: openAMUNDSEN-DA v0.9: an ensemble based snow data assimilation framework for the open source snow hydrological model openAMUNDSEN, EGUsphere [preprint], https://doi.org/10.5194/egusphere-2026-4575, 2026.

License and availability

The stable v0.9.4 Python package is available from PyPI. The tested multi-architecture container is available from GHCR, and release archives and evidence are available from GitHub Releases. Archived releases are available under the openAMUNDSEN-DA concept DOI. The version-specific v0.9.4 DOI will be added after Zenodo archives the release. The MIT License permits commercial use subject to its terms.