Global Land Cover SWIR Median Composite for 2021

2021 global land cover composite generated from Sentinel-2 satellite Short-Wave Infrared (SWIR) band medians. The consortium, led by VITO in Belgium, produced worldwide 10-meter resolution imagery per 1x1 degree tile after cloud removal. Data is provided in Cloud Optimized GeoTIFF format.

Use Cases

• Classify land cover types using the SWIR B11 and B12 band reflectance values from the median composite.
• Detect changes in vegetation or water bodies by comparing the 2021 SWIR median values with other temporal layers.
• Train semantic segmentation models on the 10-meter resolution SWIR imagery to map features like forests or urban areas.
• Analyze spectral signatures in the SWIR spectrum for specific materials or land conditions across different geographic tiles.

Strengths

• Global coverage at 10-meter spatial resolution.
• Data processed to remove clouds and cloud shadows for clearer analysis.
• Organized into standardized 1x1 degree tiles for manageable processing.

Limitations

• Excludes ocean areas and inner ice sheets, creating data gaps for those regions.
• Represents a single year (2021), limiting temporal trend analysis without other years.
• Specific sample size, row count, and validation metrics are not provided.

Provenance

Source

European Space Agency (ESA) WorldCover project, produced by a consortium led by VITO.

Collection Method

Generated from Sentinel-2 satellite imagery by calculating the median value from the SWIR (B11, B12) band time series for 2021.

Time Range

2021

Freshness

null

Geography

Worldwide land coverage, excluding oceans and inner ice sheets.