LBA-ECO LC-22: Land Cover from MODIS, Mato Grosso, Brazil

2000-2001 and 2003-2004 land cover classifications for Mato Grosso, Brazil, derived from MODIS vegetation indices. A decision tree classifier, trained on field observations and Landsat TM data, identifies seven land-cover classes from 250-meter resolution NDVI and EVI time series. The data is provided as two GeoTIFF files.

Use Cases

• Tracking deforestation and agricultural expansion based on annual land cover classifications.
• Modeling vegetation phenology and ecosystem dynamics using the 16-day MODIS NDVI/EVI time series.
• Validating regional-scale land cover maps with field-observation-trained classifier outputs.
• Analyzing spatial patterns of land cover change at 250-meter resolution over a multi-year period.

Strengths

• Classifier trained on field observations and Landsat TM data, suggesting ground-truth validation.
• Provides multi-temporal coverage for two distinct periods (2000-2001, 2003-2004).
• Uses two established vegetation indices (NDVI and EVI) from the MODIS sensor.
• Cross-platform presence on NASA and government portals indicates established use.

Limitations

• Metadata conflict: 'last updated' dates differ significantly between sources (2004 vs. 2026).
• Sparse column and row metadata; specific data dimensions and file sizes are unknown.
• Temporal coverage is limited to two short, non-consecutive periods in the early 2000s.

Provenance

Source

National Aeronautics and Space Administration

Collection Method

Derived from MODIS Collection 4 vegetation indices using a decision tree classifier trained on field and Landsat data.

Time Range

2000-2004

Freshness

2026-04-10 01:45:07.909843

Geography

Mato Grosso, Brazil