data api · Global

A vertical vegetation structure model of the Earth

To better support carbon and biodiversity monitoring and to support the SDGs, EUDR, and 30x30, we have developed a global vertical vegetation structure model (VSM) to describe the full 3D-structure of vegetation at high spatial resolution.

Visit source ↗

Claim this listing

Represent A vertical vegetation structure model of the Earth? Send a claim for review so we can discuss listing updates. Claims do not automatically grant ownership or verification.

Explore paid sponsorship →

What's in it

Data offered docs-verified: Relative height profile RH0–RH100 (101 relative height metrics, stored in decimetres) · Predictive uncertainty quantiles Q2 and Q0 (not yet published)
Coverage docs-verified: Global land (GEDI training data spans about 51.6° N to 51.6° S; higher latitudes rely on extrapolation) · 2020 (available), 2024 (in preparation)
Resolution docs-verified: 10 m (main tiles, 10 980 × 10 980 px per Sentinel-2 MGRS tile); 1 km (global mosaics, averaged from 10 m tiles to 0.01°)
Formats docs-verified: Cloud Optimized GeoTIFF (Int16; lossless LERC+ZSTD for tiles, ZSTD for mosaics)
Access docs-verified: Free and open: public, anonymous HTTPS/S3 access via https://data.source.coop (no sign-in); licence CC BY 4.0 (attribution required)

Useful for

Use cases llm-inferred: Pull the vertical vegetation profile at site coordinates for ecological desk studies and habitat context · Canopy top height and understory density as inputs for land-cover roughness assumptions in flood or drainage modelling · Screen woodland structure and canopy layering ahead of planning-stage ecological surveys · Compare canopy structure across years for change screening, with imagery-driven differences noted
Audience llm-inferred: ecologists · GIS analysts · hydrologists · flood consultants · planners · environmental consultants

Limitations

Constraints docs-verified: Model predictions, not lidar measurements; accuracy varies with biome, terrain slope and canopy density · Dense tall forests: predicted heights biased toward the mean · GEDI samples only about 51.6° N to 51.6° S; higher latitudes depend on model extrapolation · RH metrics are not identical to airborne-lidar canopy height models · Year-to-year differences mix real change with input imagery differences · Global mosaics (~1 km) smooth fine-scale structure; use 10 m tiles for quantitative work · Only 2020 is available and only the median (Q1); Q2 and Q0 are coming soon; 2024 in preparation · Full product volume is about 360 TB per year; tiles must be selected via the Sentinel-2 MGRS grid

Licence: see product README
Authentication: Public; anonymous HTTPS/S3 access via data.source.coop
GitHub stars: Unknown
Last commit: Unknown
Last successful page check: 14/10/2026
Page health: page reachable
Package: Not recorded
Verification: Not independently verified

Provenance & evidence

Verified via Source Cooperative API /api/v1/products/geoai-ucph (public product, updated 2026-10-12).

Metadata observed 14/10/2026 · research sweep: manual-2026-10-15 (source.coop)