United States Land Cover Satellite Imagery 2016-2024

A dataset integrates Sentinel-2 satellite imagery with Dynamic World land cover labels for the entire United States from 2016 to 2024. It includes six spectral bands (Red, Green, Blue, NIR, SWIR1, SWIR2) and pixel-level land cover type annotations.

Use Cases

• Train a model for pixel-level land cover classification using the six spectral bands (Red, Green, Blue, NIR, SWIR1, SWIR2) as input features.
• Analyze land cover change over time across the United States by comparing labels from 2016 to 2024.
• Develop a change detection algorithm for deforestation or urbanization using the time-series of spectral bands and land cover labels.
• Benchmark remote sensing models on a dataset covering the entire United States with Sentinel-2 imagery and Dynamic World annotations.

Strengths

• Covers the entire United States, providing national-scale spatial coverage.
• Includes a nine-year time range from 2016 to 2024 for temporal analysis.
• Integrates data from two established sources: Sentinel-2 satellite imagery and the Dynamic World dataset.
• Provides six spectral bands per observation, offering multi-spectral input for models.

Limitations

• The dataset size is listed as 0.0 B, indicating a very limited scope or a metadata placeholder rather than the actual data.
• Specific row count, column structure, and file formats are unknown, hindering assessment of scale and usability.

Provenance

Source

Integrates Sentinel-2 satellite imagery and Dynamic World land cover annotations.

Collection Method

Pixel-level integration of multi-band satellite imagery with land cover labels.

Time Range

2016 to 2024.

Freshness

Last metadata update was 2026 -03-15.

Geography

Entire United States.