Landsat Annual Land Cover Maps for Colombian Amazon

Annual land cover maps for the Colombian Amazon from 2001 through 2016. The maps were generated by applying a Random Forest classifier to time segments from the Continuous Change Detection and Classification algorithm on Landsat pixel surface reflectance data. This dataset was produced by the National Aeronautics and Space Administration and covers eight land cover classes including forest, pastures, and secondary forest.

Use Cases

• Analyzing deforestation rates by tracking annual changes in the 'forest' and 'secondary forest' classes.
• Modeling urban expansion and agricultural conversion using the 'urban' and 'pastures' land cover labels.
• Assessing surface water changes over the 16-year period via the 'water' class.
• Studying land surface reflectance properties using the 'highly reflective surfaces' classification.
• Evaluating the effectiveness of land cover classification algorithms on time-series satellite data.

Strengths

• Provides a 16-year continuous annual time series from 2001 to 2016 for longitudinal analysis.
• Land cover classification was performed using a Random Forest model with manually collected training data.
• Maps are derived from Landsat data using the peer-reviewed Continuous Change Detection and Classification algorithm.
• Dataset is hosted on multiple authoritative platforms (Data.gov, NASA Earthdata), indicating vetting and importance.

Limitations

• The manually collected training data used for classification is not included with the dataset.
• There is a conflict in the 'last updated' metadata: Data.gov lists 2026-03 13, while NASA Earthdata lists 2016-12-31.
• Specific column names, row counts, and file sizes are not provided in any source description.

Provenance

Source

National Aeronautics and Space Administration

Collection Method

Classification of Landsat time segments using a Random Forest classifier and manual training data.

Time Range

2001-2016

Freshness

2026-03-13 04:05:17.015382

Geography

Colombian Amazon