EO·Atlas
Spectral unmixing
Decomposes a mixed-pixel spectrum into the fractional abundances of its constituent endmember materials, resolving sub-pixel composition that a single hard class label cannot capture. Used to estimate vegetation, soil, and impervious-surface fractions.
Spectral unmixing answers the question: what materials are present within a pixel, and in what proportions? A single sensor pixel typically covers an area containing several surface types simultaneously. Rather than assigning one class label to the whole pixel, linear spectral mixture analysis (LSMA) decomposes the observed reflectance spectrum into a weighted sum of pure endmember spectra, returning a fractional abundance estimate for each endmember.[1]
The constrained least-squares formulation requires that fractional abundances are non-negative and sum to one. Endmembers commonly used in vegetation and land-cover applications are Green Vegetation (GV), Non-Photosynthetic Vegetation (NPV), Soil, and Shade, where the Shade endmember accounts for shadows and water.[2] The INPE PRODES Brazilian Amazon deforestation monitoring system has applied Landsat 30m linear SMA annually since 1988: GV fraction images are computed, and pixels showing GV fraction loss accompanied by Soil and NPV fraction gain are classified as clearing events.[3]
The foundational paper for linear SMA is Adams et al. 1993 (Remote Sensing of Environment), which established the endmember concept and constrained least-squares formulation.[1] The method requires enough spectral bands to separate the endmembers of interest, with hyperspectral data providing more robust separation than multispectral for applications with many endmember types. Unmixing error increases with endmember spectral variability and degrades in highly heterogeneous landscapes where the pure-pixel assumption does not hold.
- Advanced Hyperspectral Imager
via Gaofen-5 (GF-5)
GF-5 AHSI 330-band hyperspectral data demonstrated for spectral unmixing in soil, mineral, and vegetation applications; peer-reviewed literature confirmed.
- Terraprisma Satellite-Based Exploration Intelligence
- HyperScout H
HyperScout-H 25-band pixelated filter array enables per-pixel spectral unmixing for mixed surface composition retrieval
Identifies surface materials by matching an observed reflectance spectrum against a reference library of known signatures, typically after continuum removal to isolate diagnostic absorption features. Used for mineral and surface-composition mapping from imaging spectrometers.
Detects and quantifies localised trace-gas enhancements by applying a matched filter tuned to the target gas absorption signature across an imaging-spectrometer scene, isolating plumes against a variable surface background. Used for methane and carbon dioxide point-source mapping.
- [1]The spectral mixture problempeer reviewed2026-06-04(1993) Foundational linear SMA paper; endmember concept; constrained least squares formulation
- [2]Monitoring tropical forest degradation using spectral unmixing and Landsat time series analysispeer reviewed2026-06-04(2018) Landsat + LSMA for Brazilian Amazon forest degradation; GV/NPV/Soil/Shade endmembers; NDFI index built on fractions
- [3]Spectral Mixture Analysiscommunity2026-06-04(2023) Clear description of fraction image outputs, endmember selection, constrained model; PRODES application cited
Edited from public sources by SpectraWorks editorial. See the data dictionary for field definitions.
How to cite this page
Plain text
Spectral unmixing. EO-Atlas, SpectraWorks B.V. https://eo-atlas.org/topics/methodologies/spectral-unmixing (CC BY 4.0)
BibTeX
@misc{eoatlas:topics:methodologies:spectral-unmixing,
title = {Spectral unmixing},
author = {{EO-Atlas}},
publisher = {SpectraWorks B.V.},
note = {CC BY 4.0},
url = {https://eo-atlas.org/topics/methodologies/spectral-unmixing}
}