Releases: rayford295/vgi-spatial-bias
Release list
Colorado Springs study region data (2019)
Second study region: 2x2 km Colorado Springs tile. Assets: merged LiDAR point cloud (cs_lidar_2km.laz, EPSG:6350, ground/non-ground classified, ~5 pts/m2), NAIP 4-band imagery clipped to the tile (EPSG:6350, 0.6 m), the original full-resolution NAIP scene (EPSG:26913), and the complete source bundle (raw LiDAR tiles + statewide Colorado OSM 2019 building/major-road extracts, county-joined). Used by src/prepare_data.py (colorado step) and the region pipeline. USGS/USDA public domain; OSM data (c) OpenStreetMap contributors, ODbL.
Campus RS reference data (2019)
Remote-sensing reference data for the 2x2 km UIUC campus tile: NAIP 4-band (RGBN, ~0.7 m) aerial imagery, EPSG:6350 — input to src/naip_segmentation.py. USGS/USDA public domain. LiDAR tiles are on I-GUIDE storage (see notebook).
OSM 2019 Illinois statewide extracts (buildings + roads)
Statewide Illinois OpenStreetMap 2019 snapshot, shapefile format (WGS84), temporally matched to the 2019 USGS 3DEP LiDAR acquisition.
- Buildings: 1,197,659 polygons (
gis_osm_building_IL_WGS84) — fields: osm_id, lastchange, fclass, type, height, levels, county join (COUNTY_NAM, CO_FIPS) - Roads: 765,328 lines (
gis_osm_roads_2019_IL) — fields: osm_id, lastchange, fclass, ref, oneway, maxspeed, bridge, tunnel, surface, county join - Extent (WGS84): -91.51, 36.98 → -87.51, 42.51 (full state)
The campus-extent subsets used by the VGI comparison pipeline are committed in the repo root as osm_buildings_2019.geojson / osm_roads_2019.geojson. These release assets are the full statewide source for scaling the analysis across the urban→rural gradient.