LBA-ECO LC-03: SAR-Based Land Cover and Biomass for Brazilian Amazon

LBA-ECO LC-03 provides three related geospatial products for four study areas in the Brazilian Amazon: Manaus, Tapajos, Rio Branco, and Rondonia. The dataset includes orthorectified JERS-1 and RadarSat SAR images, derived land cover classifications distinguishing four classes, and biomass estimates in tons per hectare. Data providers explicitly note the products have not been validated or quality-assured for general use.

Use Cases

• Analyzing land cover change based on SAR image classifications.
• Estimating regional biomass stocks using derived biomass GeoTIFFs.
• Comparing forest structure (short vs. tall vegetation) across different Amazonian sites.
• Studying radar backscatter signatures from L-band (JERS-1) and C-band (RadarSat) sensors.

Strengths

• Provides three integrated products (raw imagery, classification, biomass) for each of four distinct study areas.
• Includes multi-season L-band SAR data (Fall and Spring JERS-1) and C-band RadarSat data for each site.
• Offers high-resolution orthorectified imagery (12.5m and 25m) for the Manaus and Tapajos sites.

Limitations

• Data providers state the data have not been validated or quality-assured, advising extreme caution and recommending against use of derived products.
• The dataset has missing or incomplete data and metadata, diminishing usability.
• Spatial resolution varies by site (25m for Manaus/Tapajos, 90m for Rondonia/Rio Branco), limiting cross-site comparability.

Provenance

Source

National Aeronautics and Space Administration

Collection Method

Derived from orthorectified JERS-1 and RadarSat Synthetic Aperture Radar (SAR) imagery.

Time Range

Includes imagery from 1978 for the Rondonia and Rio Branco sites.

Freshness

Metadata last updated 2026-03-13 on datagov, but the underlying data appears to be from 1978 for some sites.

Geography

Four study areas across the Brazilian Amazon: Manaus (Amazonas), Tapajos National Forest (Para), Rio Branco (Acre), and Rondonia (Rondonia).