Licensing

For businesses and professionals building on urban tree data.

First-pass page

This page describes the public data as it stands today. If you need something it does not cover, get in touch and we will work out specifics.

What you can do

The public datasets are free to use.

Be aware of share-alike. Part of the data is OpenStreetMap-derived and falls under ODbL, as a derived database. If you need data without this obligation, you can filter, see below.

Request more cities! Adding a city is straightforward where the city publishes an inventory, and anyone can request one for the public data. If the city does not publish data, we may still have options.

Excluding OSM rows

Every row carries a data_source and merged_sources. Filter data_source to remove prefix OSM_, and where an OSM node was merged into a municipal record the row says so in merged_sources. Excluding both is a filter you can run against the public files:

INSTALL httpfs; LOAD httpfs;

COPY (
  SELECT * FROM read_parquet([
    'https://storage.googleapis.com/trilogy_public_models/duckdb/trees/ussfo_tree_info_v2.parquet',
    'https://storage.googleapis.com/trilogy_public_models/duckdb/trees/usnyc_tree_info_v2.parquet'
    -- ... one URL per city; the datasets page lists them all
  ], union_by_name = true)
  WHERE NOT starts_with(data_source, 'OSM_')
    AND NOT coalesce(contains(merged_sources, 'OSM_'), false)
) TO 'trees_no_osm.parquet' (FORMAT parquet);

That is every city at once. You give up the OSM rows outright, and a further slice of municipal records that had an OSM node merged into them; the datasets page carries the current totals and the municipal/OSM split, which move with every refresh. It has to be the per-city files: the all-cities rollup does not carry merged_sources, so the second half of the predicate has nothing to read.

If you keep the OSM rows, you owe © OpenStreetMap contributors wherever they are shown, and ODbL's share-alike applies to databases you derive. We treat our own aggregate as a derived database rather than a Collective Database, because our trees are a mix of OSM and non-OSM within each city collective-database guideline.

Per-source licences for every city are on the datasets page.

What is actually on offer

	Public data
Cost	Free
Coverage	Every city we hold
Format	Parquet over HTTPS
OSM rows	Included (ODbL), excludable with one filter
Refresh	Our schedule
Support	GitHub issues

Get in touch

Email arborary.world@gmail.com if you're interested in talking trees!