Category: Processing and quality
NoData
NoData is a special value that means a pixel has no valid measurement, for example due to clouds or missing coverage.
Also known as: missing data, invalid pixels, NA (raster)
Expanded definition
NoData indicates missing or invalid pixels. It is not the same as zero.
NoData handling matters for statistics and machine learning. If NoData is treated as a real value, averages and indices can be biased and models can learn artifacts.
Good products document what causes NoData (clouds, shadows, sensor gaps, edge effects) and provide masks or flags so users can handle it correctly.
Related terms
Cloud Mask
A cloud mask labels pixels likely affected by clouds so they can be excluded or handled differently.
Gap Filling
Gap filling estimates missing values in time series caused by clouds, shadows, or data dropouts.
Raster
A raster is a grid of pixels where each pixel stores a value, commonly used for satellite imagery and elevation models.