Choose and configure data sources¶
VoxCity builds each city model from four layers — building footprints/heights, land cover, tree canopy height, and terrain elevation. This guide shows how to select the source for each layer. For the full catalog of datasets and their coverage, see the Data sources reference.
Let VoxCity choose (auto mode)¶
If you omit the source arguments, VoxCity selects appropriate datasets based on your location:
from voxcity.generator import get_voxcity
voxcity = get_voxcity(rectangle_vertices, meshsize=5)
Specify every source (custom mode)¶
Pass each source explicitly to control exactly which datasets are used:
from voxcity.generator import get_voxcity
voxcity = get_voxcity(
rectangle_vertices,
meshsize=5,
building_source='OpenStreetMap',
land_cover_source='OpenStreetMap',
canopy_height_source='High Resolution 1m Global Canopy Height Maps',
dem_source='DeltaDTM',
)
Mix specified and auto sources (hybrid mode)¶
Specify only the layers you care about and let VoxCity auto-select the rest:
from voxcity.generator import get_voxcity
voxcity = get_voxcity(
rectangle_vertices,
meshsize=5,
building_source='Overture', # custom
# land_cover_source, canopy_height_source, dem_source auto-selected
)
Choosing a source¶
Pick sources based on your area’s coverage and the resolution you need:
Buildings —
OpenStreetMaphas the broadest reach; regional datasets such asEUBUCCO(Europe), Microsoft Building Footprints, orOverturecan give better coverage or heights in specific regions.Land cover — global options (
ESA WorldCover,ESRI,Dynamic World) work anywhere at 10 m; higher-resolution regional sources exist for the US and Japan.Canopy height — the 1 m Global Canopy Height maps give the finest detail; the ETH 10 m product is a global fallback.
Terrain —
DeltaDTMandFABDEMare global; 1 m national LiDAR DEMs are available for the US, England, France, and Australia.
Consult the Data sources reference for exact coverage, resolution, and provenance before committing to a source.