GEDI (Global Ecosystem Dynamics Investigation)

The first space-borne lidar designed to measure forest structure and biomass at global scale. Mounted on the ISS; samples Earth’s temperate and tropical forests via 8 parallel tracks of 25 m footprints. Generates the canonical NASA above-ground biomass product family.

What it sees

• Full-waveform lidar at 1064 nm — penetrates canopy, returns vertical profile of vegetation density
• 25 m footprint diameter; ~60 m along-track spacing; 8 footprint tracks across ~4.2 km swath
• 3 separate lasers, 2 split into 4 beams each = 8 simultaneous tracks
• L1B waveform → L2A canopy heights → L4A per-footprint biomass density (Mg/ha)

Why it matters

• Above-ground forest biomass at hectare-resolution sampling globally — closes a gap that satellite radar (Sentinel-1, NISAR) and optical (Landsat) cannot fill alone.
• L4B 1-km gridded biomass is the operational NASA product for carbon accounting + REDD+ + national biomass inventories.
• Pairs with NISAR (s03 →) for the new global forest-carbon stack: GEDI samples in 3D; NISAR mosaics in 2D.

Where to get the data

• earthaccess Python: short_name="GEDI_L4A_AGB_Density_V2_1_2056" (footprint biomass), GEDI_L4B_AGB_Density_V2_2056 (gridded)
• ORNL DAAC for L4: daac.ornl.gov
• LP DAAC for L1B–L2B: lpdaac.usgs.gov

What it enables

• Per-footprint above-ground biomass mapping for any forested AOI ≤51.6° latitude
• Canopy height + relative height percentile (RH98) products for forest disturbance
• Carbon-stock estimation paired with HLS surface-reflectance (q01 →)
• Reference data for foundation models (Prithvi-EO, AlphaEarth Foundations use GEDI as ground truth)
• Hydrography (water-surface elevation as a side product — feeds SWOT validation)

Gotchas

• No polar coverage. ISS orbit caps at ±51.6° latitude. Boreal forests above this latitude are excluded.
• Footprint geometry is sparse + non-grid. Don’t try to xarray-stack as a raster; treat as a sparse point cloud. Use geopandas + spatial joins.
• Primary mission ended March 2023. Extended mission resumed August 2024 — but there’s a gap in 2023-08 to 2024-08. Filter by date carefully.
• L4A has known biases in dense tropical canopies; the saturation point is ~250 Mg/ha for most allometric models. Above that, biomass estimates compress toward an asymptote.
• HDF5 reading is non-trivial — multiple beams, deep group structure. Use the h5py recipe in gedi-l4a-tutorial or open via earthaccess.open + custom reader.

Related missions

• NISAR — L-band SAR, complementary 2D coverage of GEDI’s 3D sampling
• BIOMASS (ESA): P-band SAR for forest biomass, launched April 2024
• ICESat-2 (s10 →): cryosphere-focused lidar; secondary use for canopy
• MAAP: NASA+ESA cloud platform for combined GEDI+BIOMASS+NISAR analysis

Related datasets