What Is Polarimetric Synthetic Aperture Radar?

Polarimetric synthetic aperture radar (PolSAR) is a microwave remote sensing technology that transmits and receives radar signals across multiple polarization channels simultaneously, extracting detailed information about the physical and geometric properties of observed surfaces. By measuring how a target scatters electromagnetic energy at different polarization orientations, PolSAR produces far richer characterizations of terrain, vegetation, and built structures than conventional single-polarization radar systems can provide. The technology draws on radar engineering, electromagnetic wave theory, and signal processing, and has been an active area of research in the IEEE Geoscience and Remote Sensing Society since the launch of the airborne AIRSAR sensor in 1985.

A conventional synthetic aperture radar images a scene by exploiting the motion of a platform, typically a satellite or aircraft, to synthesize a large virtual antenna and achieve fine spatial resolution. PolSAR extends this by operating in fully polarimetric mode: the sensor transmits horizontally (H) and vertically (V) polarized pulses and records the backscattered energy in both H and V channels. The resulting 2x2 complex scattering matrix, or its statistical generalization as the covariance or coherency matrix, encodes all polarimetric information about the illuminated scene.

Scattering Decomposition

The central analytical framework in PolSAR is scattering decomposition, which separates the total observed backscatter into physically interpretable components. The Cloude-Pottier decomposition, based on eigenanalysis of the coherency matrix, yields three canonical scattering mechanisms: surface scattering from smooth or gently undulating terrain, volume scattering from randomly oriented structures such as forest canopies, and double-bounce scattering from dihedral-like arrangements such as vertical trunks against flat ground. The Freeman-Durden model, described in a foundational paper in IEEE Transactions on Geoscience and Remote Sensing, provides an alternative three-component decomposition with explicit physical basis functions. These decompositions are foundational tools because they translate abstract matrix quantities into geophysical interpretations without requiring any ground-truth training data.

Land Cover Classification

PolSAR data support classification of land cover at spatial resolutions that are unaffected by cloud cover or solar illumination, a critical advantage over optical sensors. Classification algorithms exploit the statistical distribution of the coherency matrix, commonly modeled as a Wishart distribution, to assign each resolution cell to a land cover class. Iterative schemes such as the Wishart classifier, refined by machine learning approaches, can distinguish cropland, forest, urban fabric, and wetland with high accuracy. Research published in IEEE Transactions on Geoscience and Remote Sensing has consistently demonstrated classification accuracies exceeding 90 percent for well-separated classes on airborne and spaceborne PolSAR data. The ability to map vegetation structure and biomass makes PolSAR particularly valuable for forest inventory and carbon stock estimation.

Target Detection and Change Detection

PolSAR's sensitivity to structural orientation makes it effective for detecting man-made objects embedded in natural backgrounds. Ships on open water, for instance, produce strong double-bounce or odd-bounce responses against a background dominated by surface scattering, allowing detection algorithms based on polarimetric contrast to operate with lower false-alarm rates than those relying on intensity alone. Change detection applications exploit differences in scattering mechanisms between two PolSAR acquisitions to identify areas that have undergone physical transformation, supporting damage assessment after earthquakes or floods. Multilook processing, which averages neighboring resolution cells to reduce speckle, and polarimetric optimization techniques both improve the sensitivity of change detection methods.

Applications

Polarimetric synthetic aperture radar has applications in a wide range of disciplines, including:

• Agricultural crop monitoring and yield estimation using crop-growth-phase-dependent scattering signatures
• Forest biomass estimation and deforestation mapping for carbon accounting
• Wetland and coastal habitat characterization where optical data are frequently obscured by cloud cover
• Urban infrastructure mapping and damage assessment following natural disasters
• Glaciology and sea ice classification for polar navigation and climate research