MoBot Ginkgo Tree — Classified Point Cloud
Terrestrial laser scan of the ginkgo tree at the Missouri Botanical Garden.
Data & Collections
The Fossett Laboratory keeps an open, machine-readable archive of geospatial data from its field campaigns and lab work. Everything is served in standardized, cloud-optimized formats through a STAC API, so a script can pull exactly what it needs as easily as a person can browse. Datasets include UAV photogrammetry models, lidar point clouds, multispectral imagery, and planetary terrain data.
Point cloud viewer
Several of the lab's lidar surveys can be opened in a browser, with nothing to install. The viewer streams the published COPC file directly from S3, requesting only the parts of the octree the current view needs rather than the whole file, so a survey of several hundred million points can be examined without downloading it. The same files are in the catalog above for download.
Other scenes
- MoBot Ginkgo Tree — Full Scan The unclassified scan of the same ginkgo, at 57 million points.
- Graham Chapel Lidar survey of Graham Chapel on the Washington University Danforth Campus.
- Proffit Mountain Lidar survey of Proffit Mountain, Missouri, site of the 2005 Taum Sauk reservoir failure.
- Confluence Farm Lidar survey of Confluence Farm, at the meeting of the Missouri and Mississippi rivers.
- Shaw Nature Reserve Lidar survey of Shaw Nature Reserve, Franklin County, Missouri.
- Cahokia Mounds Lidar survey of the Cahokia Mounds site in Illinois, covering Monks Mound and the surrounding plaza.
- Tyson Research Center — Airborne Lidar Airborne survey of Tyson Research Center, 332 million points with no color; shown by elevation.
Programmatic Access
For developers, researchers, and AI agents alike: discover, subset, and pull data through the same open STAC interfaces the major Earth-observation platforms use.
STAC Browser
Open in new tab →STAC API
Connect from Python, R, or any STAC client.
import pystac_client
catalog = pystac_client.Client.open(
'https://fossettlab.org/api/stac'
)S3 Bucket
Open datasets live in the public S3 bucket
bradleylab-public.
Bucket listing is disabled (standard S3 default); access individual files via
the asset URLs in each STAC item, or start from catalog.json.
Restricted datasets require a WashU login.
Data Citation
Each dataset row in the table above includes a Cite & stable URL expander with a permanent link to the STAC item plus suggested plain-text and BibTeX citations. Use those formats when referencing Fossett Laboratory datasets in publications, reports, or downstream STAC-aware tooling.
Licensing
Most datasets are released under Creative Commons Attribution 4.0 (CC BY 4.0). Check individual dataset metadata for specific terms.
Accessing Restricted Data
Some datasets are restricted to Washington University researchers. Contact Bill Winston with your WashU WUSTL Key and a brief description of your intended use.
STAC Quickstart
The Fossett Lab catalog is a SpatioTemporal Asset Catalog (STAC) — a standard for geospatial data discovery used across NASA, ESA, Microsoft, and the open-source geo community. STAC clients work the same way against our catalog as against any other STAC source.
What is STAC?
STAC describes geospatial datasets as a hierarchy of catalog → collections → items → assets. Every item carries a footprint, a datetime, and links to the actual data files (lidar point clouds, GeoTIFFs, etc.). Searches over bbox, time, or keyword work uniformly across providers.
Python (pystac-client)
# pip install pystac-client
import pystac_client
cat = pystac_client.Client.open(
"https://fossettlab.org/api/stac"
)
# List collections
for c in cat.get_collections():
print(c.id, "—", c.title)
# Search by bbox (e.g. Tyson Research Center) and time
search = cat.search(
bbox=[-90.59, 38.50, -90.53, 38.55],
datetime="2024-01-01/2026-12-31",
collections=["pointclouds"],
)
for item in search.items():
print(item.id, item.datetime,
item.bbox)
# Open the first asset, e.g. a COPC point cloud
for k, a in item.assets.items():
print(" ", k, a.href)R (rstac)
# install.packages("rstac")
library(rstac)
cat <- stac("https://fossettlab.org/api/stac")
# List collections
collections(cat) |> get_request()
# Search by bbox + time
results <- cat |>
stac_search(
bbox = c(-90.59, 38.50,
-90.53, 38.55),
datetime = "2024-01-01/2026-12-31",
collections = "pointclouds"
) |>
get_request()
# Inspect items + assets
items_length(results)
items_assets(results)Direct HTTP / curl
# Catalog root
curl https://fossettlab.org/api/stac
# All collections
curl https://fossettlab.org/api/stac/collections
# Search items in the pointclouds collection
curl "https://fossettlab.org/api/stac/search?collections=pointclouds&bbox=-90.59,38.50,-90.53,38.55"
# A specific item (returns full STAC item JSON with asset URLs)
curl https://fossettlab.org/api/stac/collections/pointclouds/items/2026-04-03_tyson_lidarCoordinate reference systems & formats
- Footprints & bbox: always WGS84 (EPSG:4326), as required by the STAC spec.
- Asset CRS: stored on disk in each dataset's native CRS (commonly UTM 15N / EPSG:32615 for Missouri sites). Each item's
properties.proj:epsgor asset metadata reports the actual CRS. - Point clouds: Cloud-Optimized Point Cloud (
.copc.laz) where available; legacy.las / .lazotherwise. Stream chunks with COPC-aware readers (PDAL, copclib). For what the bytes hold, see the lab's guide to reading a LAS/LAZ header, and what COPC adds. - Imagery: Cloud-Optimized GeoTIFF (
.tif) so range-requests work over HTTP — open directly with rasterio, GDAL, or QGIS. - 3D models: glTF / GLB and Gaussian splat formats where applicable.
Licensing & citation
Most datasets are released under Creative Commons Attribution 4.0 (CC BY 4.0). Each STAC item carries a license field; check it before redistribution.
When citing a dataset, use the format provided in the item's sci:publications or sci:citation metadata if present, or attribute the Fossett Laboratory and the dataset's collection date.