LBA-ECO LC-21: Selective Logging Detection in the Brazilian Amazon (1999-2001)

Fifteen GeoTiff files map selective logging activity across five Brazilian states (Para, Mato Grosso, Rondonia, Roraima, Acre) from 1999 to 2001. Each 30x30 meter pixel is classified as either logging damage (1) or no logging (0) using the Carnegie Landsat Analysis System (CLAS). This dataset provides a spatially detailed, automated detection of forest canopy damage for a critical period in the Amazon's timber production regions.

Use Cases

• Quantifying the spatial extent of selective logging damage over a three-year period based on annual binary classification maps.
• Analyzing patterns of forest disturbance and canopy damage at a 30-meter resolution using the CLAS methodology.
• Calibrating or validating newer remote sensing models for detecting subtle logging activity versus clear-cut deforestation.
• Assessing the effectiveness of automated image analysis combining atmospheric modeling and pattern recognition for forest monitoring.

Strengths

• Provides detailed, automated detection at a high spatial resolution of 30 x 30 meters.
• Covers a specific, three-year time range (1999-2001) across five major timber-producing states.
• Data is hosted on multiple authoritative platforms (NASA, Data.gov), indicating importance and vetting.

Limitations

• Metadata completeness is medium; column names and exact file sizes are not provided.
• There is a conflict in the reported temporal coverage: the title suggests 1999-2002, while all descriptions state 1999-2001.
• The most recent update date (2026-04 10) on Data.gov is likely an administrative error or placeholder, as the data itself is from the early 2000s.

Provenance

Source

National Aeronautics and Space Administration (NASA)

Collection Method

Automated analysis of Landsat ETM+ images using the Carnegie Landsat Analysis System (CLAS).

Time Range

1999-2001

Freshness

2002-12-31 23:59:59.999000

Geography

Brazilian Amazon states: Para, Mato Grosso, Rondonia, Roraima, Acre