S1 AI4LCC: Sentinel-1 Satellite Imagery with Land Cover Labels

Sentinel-1 Land Cover Dataset pairs satellite imagery from the Sentinel-1 mission with land-cover labels derived from the AI4LCC benchmark. The data is stored in Zarr format for efficient storage and scalable access. The dataset was created by author wtr001 and was last updated on 2026-06-19.

Use Cases

• Train land cover classification models based on paired Sentinel-1 imagery and AI4LCC-derived labels.
• Benchmark geospatial machine learning algorithms based on the AI4LCC benchmark dataset.
• Develop scalable data pipelines for large geospatial datasets based on the Zarr storage format.
• Analyze synthetic aperture radar (SAR) signatures for different land cover types based on Sentinel-1 data.

Strengths

• Data is stored in Zarr format, which the description states enables efficient storage and scalable access for large geospatial datasets.
• The dataset pairs Sentinel-1 imagery with land-cover labels, creating a ready-to-use resource for supervised machine learning tasks.

Limitations

• Column-level documentation is absent; field semantics must be inferred after download.
• Row count, file formats, and license information are unknown, which may limit suitability assessment.
• Freshness should be verified as the last update timestamp is 2026-06-19.

Provenance

Source

Labels derived from the AI4LCC benchmark dataset; imagery from the Sentinel-1 satellite mission.

Collection Method

Original labels and Sentinel-1 data were converted into Zarr format, and Sentinel-1 image tiles corresponding to label locations were added.

Freshness

Last updated 2026-06-19 04:51:43.