Polarimetric SAR data representations

SAR Polarimetry Target Analysis supports fully polarimetric data in various representations, as described on the following table.

Representation	Description
s4c	Nonsymmetrized scattering matrix (single look only)
S3c	Symmetrized scattering matrix
c4r6c	Nonsymmetrized covariance matrix (best for multilook)
C3r3c	Symmetrized covariance matrix
t4r6c	Nonsymmetrized coherence matrix (best for multilook)
T3r3c	Symmetrized coherence matrix
k16r	Nonsymmetrized Kennaugh matrix (best for multilook)
K9r	Symmetrized Kennaugh matrix

Typically, fully polarimetric data is distributed in s4c or S3c representation. Other representations are often the result of specific processing, such as data filtering. For example, a 10-channel covariance matrix (c4r6c) can be the result of applying a Boxcar filter (by running the PSBOXCAR algorithm) or an Enhanced Lee filter (by running PSPOLFIL).

The most common supported matrix types for single or dual polarimetric data are as described in the following table.

Type	Description
s1c	Incomplete scattering matrix (single-pol, single-look only)
s2c	Incomplete scattering matrix (dual-pol, single-look only)
c2r	Incomplete covariance matrix (dual-pol, detected, single-look or multi-look)
c2r1c	Incomplete covariance matrix (best for dual-pol multi-look)

The matrix type of a SAR data set can be found in its metadata. The matrix type is updated automatically whenever the data is processed.

You can perform a pseudo fully polarimetric analysis on compact-polarimetric data. For example, RCM or RISAT compact-pol data can be filtered (by running PSBOXCAR or PSPOLFIL). The data can then be transformed into a pseudo fully polarimetric covariance matrix by running PSRECONS.

For more detailed information about the various matrix types that can be used to represent polarimetric SAR data, see the PSCONV algorithm.