PREFIRE Cloud Mask: Binary Clear/Cloud Indicators from Polar CubeSats

PREFIRE Satellite 1 and 2 Cloud Mask datasets contain binary clear/cloud indicators and probabilities derived from mid- and far-infrared spectrometer measurements. The data is produced by a machine learning technique using spectral radiance and auxiliary meteorology inputs to support the generation of other atmospheric property products. Science data retrieval began in June and July 2024, with global coverage concentrated in the polar regions.

Use Cases

• Identifying cloudy versus clear scenes in far-infrared spectral data to filter inputs for downstream products.
• Assimilating polar cloud cover probabilities into global circulation models to improve climate predictions.
• Analyzing spatial patterns of cloud frequency in polar regions to characterize the Earth's energy budget.
• Validating and improving machine learning-based cloud detection algorithms for satellite spectrometry.

Strengths

• Data is derived from a novel push broom spectrometer with 63 channels measuring a previously under-sampled part of the spectrum (5-53 µm).
• Provides specific spatial resolution details: scene footprints of 11.8 km x 34.8 km with eight distinct tracks per orbital swath.
• Cross-platform presence on NASA-affiliated portals indicates official mission data with structured metadata.

Limitations

• Column names and detailed schema are unknown across all sources, limiting precise understanding of data structure.
• Metadata for row count, file size, and specific license terms is consistently missing or listed as 'None'.
• The description for the total swath width is truncated in multiple sources, missing a complete figure.

Provenance

Source

National Aeronautics and Space Administration (NASA)

Collection Method

Machine learning technique applied to spectral radiance and auxiliary meteorology data collected by the PREFIRE Thermal Infrared Spectrometer aboard CubeSats.

Time Range

From 2024-06 29 (Satellite 2) and 2024-07-24 (Satellite 1) onward (ongoing).

Freshness

2026-04-09 23:04:41.662406

Geography

Global, with the greatest concentration of data in the polar regions.