Climate Datasets

A curated catalog of climate and weather datasets, focused on accessibility and clear metadata.

Dashboard

Browse the full catalogue at https://dpird-dma.github.io/climate_datasets/. You can filter by variable, time step, domain, access type, format, licence and grid spacing, sort the list, and share a filtered view as a link.

The dashboard lives in docs/ and is deployed to GitHub Pages on every push to main.

• Local preview: make preview then open http://localhost:8000
• Source of truth: data/datasets.json and data/vocab.json (synced to docs/data/)

Project Structure

• data/datasets.json: canonical dataset registry
• data/vocab.json: allowed values for the structured fields (access, formats, time steps, variables, domain)
• .agents/skills/: repo-scoped Codex workflows for safe dataset edits
• docs/: static dashboard (HTML/CSS/JS)
• scripts/: helpers for table generation and validation
• tests/: regression tests for dataset-editing helpers
• README.md: project overview + snapshot of the dataset table

Contribution Workflow (Short)

1. Add or update entries in data/datasets.json, preferably with the skill helpers below.
2. Run make all. It regenerates the README preview, syncs the dashboard data, checks reference labels, validates the registry (structure, field types, vocabulary and URL syntax) and runs the tests.
3. Commit the registry change with the regenerated files. CI fails if they are out of date.

See CONTRIBUTING.md for detailed guidance.

Skills

The repository ships two repo-scoped skills under .agents/skills/:

• add-dataset: adds a dataset from a required JSON snippet (or interactive prompts).
• update-dataset: updates a dataset with fuzzy name matching.

To use them directly from the repo:

• python3 .agents/skills/add-dataset/scripts/add_dataset.py --dry-run (paste JSON on stdin)
• python3 .agents/skills/update-dataset/scripts/update_dataset.py --name "SILO Point Data" --dry-run (paste JSON on stdin)

License

Catalogue content in this repository is licensed under CC BY 4.0. The datasets listed here keep their own licences, given per entry in the license field.

Dataset Catalog (Snapshot)

The table below is generated from data/datasets.json and previews the first three entries only. The full catalogue is on the dashboard.

Dataset	Resolution	Format	Variables	Method	Access Conditions	Temporal Coverage
DPIRD	station observations (~200 stations)	web, json, csv	Time series data - Evaporation, rainfall, solar radiation, air temperature, and others	Ad hoc handling of technical issues and missing values	API key registration	Minute to yearly intervals
SILO Point Data	station observations (~8000 stations)	web, json, csv, apsim	Continuous daily time series Evaporation, rainfall, solar radiation, air temperature, and others	Observational records or interpolated estimates for missing records	Accessible via SILO network	daily, from 1889 to current year
SILO Gridded Data	~5 km (0.05°)	NetCDF, GeoTiff	Evaporation, rainfall, solar radiation, air temperature, and others	Gridded daily climate surfaces derived either by splining or kriging the observational data	Free, open download from the AWS Public Data Program (no email or registration)	Daily and monthly, from 1889 to the current year (mean sea level pressure from 1957, Class A pan evaporation from 1970)

Dataset References