fiboa demos

Web demos and dashboards built on the Fields of the World harmonized field data: small, working examples of what open field boundaries make possible. Published at https://fiboa.org/demo/.

Demo	What it shows
nl_raster	Disappearing Dutch farmland: 18 editions of the Dutch parcel register (BRP) on one 2 m Cloud-Optimized GeoTIFF, read in the browser with OpenLayers.
relief	No field is flat: click a field and see its relief from the open 1 m height model of its country or state.
season	Every field has a season: the NDVI of a field in every clear Sentinel-2 image of a year, with the rain and warmth against a normal year.
flood	Fields under water: click a field and see, in every Sentinel-1 radar image of a year, how much of it stood under water, next to the daily rain.
rotation	Crop rotation: what grew on every field of the Netherlands, Flanders and Denmark, year after year, from all editions of their registers.
history	What was here before the fields: today's fields over old maps and aerial photos of the Netherlands, Flanders, France, Switzerland and North Rhine-Westphalia, with a swipe line to today.
regions	Farming regions: the Corn Belt, Napa Valley, Bordeaux and more, drawn from the crops of the fields, from an overview of A5 cells down to every field.

Develop

npm install
npm run dev      # http://localhost:5173/demo/
npm test         # data consistency and unit tests
npm run build    # static site in dist/

Every push to main is tested, built and deployed to GitHub Pages by .github/workflows/pages.yml.

Add a demo

1. Create a directory <name>/ with an index.html and its code under <name>/src/.
2. Add <name> to the demos list in vite.config.js.
3. Link it from the landing page index.html and from the table above.
4. Write <name>/src/prompt.js: the prompt that builds the demo and the conversation it came from, and call addPromptButton(PROMPT) (shared/prompt.js) in its main.js. It shows as "Built with a prompt" in the header.

Shared tokens and page frame live in shared/base.css. Data that a demo precomputes lives in <name>/data/, and the script that produces it in scripts/<name>/.

Data scripts

Precomputation runs in Python with pixi (pixi.toml), close to the data:

pixi run test                                   # pytest for the scripts
pixi run python scripts/nl_raster/stats.py --help

License

Code: Apache-2.0. Data derived from the catalog keeps the license of its source (CC0 for the Dutch BRP).