A collection of datasets from various Dutch institutions to demonstrate a Spatial Data Infrastructure built on Portolan.
Three collections from Kadaster, the Dutch government agency responsible for land registry, cadastral mapping, and geographic data infrastructure. Kadaster maintains several national key registrations (basisregistraties) and publishes them via PDOK.
11.4 million buildings (panden) from the BAG (Basisregistratie Adressen en Gebouwen), the authoritative Dutch building and address registry. Current records only, with construction year, usage function, floor area, and dwelling unit count. This is the recommended starting point for building analysis.
bag-light.parquet (933 MB)identificatie (BAG ID), bouwjaar (construction year), gebruiksdoel (usage), status342 municipalities (gemeenten), 12 provinces (provincies), and 1 national territory
(landgebied) — the 2026 edition. Three separate GeoParquet files. The municipality codes
(identificatie) are CBS codes used as join keys across all Dutch government data.
gemeentegebied.parquet (8 MB), provinciegebied.parquet (1.6 MB), landgebied.parquet (0.4 MB)identificatie (CBS code), naam (name), code (prefixed code like GM0363)24.2 million building footprints harmonized to the EU INSPIRE Buildings schema. Includes
version history (current + superseded records). Filter endLifespanVersion IS NULL for
current buildings only (~12.6M).
buildings.parquet (2.1 GB)localId (BAG ID), anyPoint (construction date as ISO string), endLifespanVersion (null = current)Both datasets come from the same BAG source, but serve different purposes:
Use BAG Light for most analysis — it has more attributes, cleaner data, and is smaller. Use INSPIRE Buildings only when you need version history or EU-standard coordinates.
The identificatie field in bestuurlijke_gebieden contains CBS municipality codes (e.g.,
0363 for Amsterdam). These 4-digit codes are the universal key across all Dutch government
data. They appear in:
identificatie)Use bestuurlijke_gebieden as the reference geometry for any municipality-level analysis.
Each collection has its own AGENTS.md with detailed field descriptions, value distributions,
and query patterns:
bag_light/AGENTS.mdbestuurlijke_gebieden/AGENTS.mdinspire_buildings/AGENTS.mdHand-maintained agent guide (ported from this collection's llms.txt in August 2026). Update it alongside collection.json.