Skip to main content
U.S. flag

An official website of the United States government

Spatial Procedures for Automated Removal of Cloud and Shadow (SPARCS) Validation Data

This dataset was originally created by M. Joseph Hughes, Oregon State University, and was derived manually from Pre-Collection Landsat 8 scenes. Its purpose was to validate cloud and cloud shadow masking derived from the Spatial Procedures for Automated Removal of Cloud and Shadow (SPARCS) algorithm.

Return to Pre-Collection Cloud Cover Assessment Validation Datasets Overview


This collection of cloud validation data contains 80, 1000 x 1000-pixel, subsets of Pre-Collection Landsat 8 scenes in .tif format, each with both a manual cloud truth mask and a color composite preview image, in .png format.

Each product is provided as a thematic raster, and includes the classes of “cloud”, “cloud shadow”, “snow/ice”, ”water”, and ”flooded”. While these validation images were subjectively digitized by a single analyst, they provide useful information for quantifying the accuracy of clouds flagged by various cloud masking algorithms.

The mask and accompanying color composite preview image are provided in PNG image format, and includes all bands from the original Landsat Level-1 data prouduct (GeoTIFF) band-interleaved by pixel (BIP), Landsat Level-1 Quality Assessment (QA)band (GeoTIFF), and its associated Level-1 metadata (MTL.txt file).

Data citation:

U.S. Geological Survey, 2016. L8 SPARCS Cloud Validation Masks. U.S. Geological Survey data release. doi:10.5066/F7FB5146.

Hughes M.J. & Hayes, D.J. (2014). Automated detection of cloud and cloud shadow in single-date Landsat imagery using neural networks and spatial post-processing. Remote Sensing, 6(6), 4907–4926. doi:10.3390/rs6064907.

The interpretation of the values in each mask is listed below.

Value / Interpretation

0    Shadow

1    Shadow over Water

2    Water

3    Snow

4    Land

5    Cloud

6    Flooded

Download SPARCS Validation Data (1.4 GB)

Additional information about this product can be found in the metadata.