Extract Numerical Weather Prediction (NWP) meteorology from the Met Office Unified Model (UM) — .pp files archived on JASMIN and CEDA — into analysis-ready yearly Zarr stores.
extract_um_met is a single command-line tool that extracts each world region,
joins them into your domain, and appends the result to a yearly store.
python -m extract_um_met run --domain SA --date 2016
# → {zarr_save_directory}/SOUTHAMERICA/SOUTHAMERICA_Met_2016.zarrThe raw archive is available via JASMIN (link your CEDA and JASMIN accounts first), or to download directly from CEDA for 2015–2024 only. Running this tool on JASMIN reaches the full record.
git clone https://github.com/elenafillo/extract_UM_met.git
cd extract_UM_met # config.yaml is read from here
module purge && module load jaspy # JASMIN Python env
export TMPDIR=/work/scratch-pw5/$USER/tmp && mkdir -p "$TMPDIR"
cp config.example.yaml config.yaml # then edit your copy
python -m extract_um_met make-native-grid # one-off
python -m extract_um_met run --domain SA --date 20160115 # single-day smoke testconfig.yaml is git-ignored (per-user paths); the shared structure — domains,
archive paths, grid specs — lives in config.example.yaml.
Documentation → elenafillo.github.io/extract_UM_met
| Page | What it covers |
|---|---|
| Setup | Environment, TMPDIR, writing config.yaml, native grids — start here |
| Extracting | The CLI: run / extract / make-native-grid, resume, SLURM, adding a domain |
| Concepts | Data types, Mk blocks, world regions, grid modes |
| Reference | Variables, model levels, output chunking, raw UM inventory |
| Troubleshooting | Symptoms, causes and fixes |
| Codebase | Module map and pipeline flow |
| Roadmap and contributing | Remaining work, open questions, how to open an issue, credits |
The pages are written in How_Tos/ and rendered by the Sphinx config
in docs/ — edit the sources in How_Tos/, never docs/.