Polarimetric radar backscattering from reciprocal random targets

Radar backscattering from reciprocal random targets is studied employing a covariance matrix approach. Polarization signatures for backscattered power and correlation observables are derived. Optimal polarization for the power return in two orthogonally polarized radar channels can be determined. Polarizations which extremize mean copolar power can be proved to reduce the correlation of the backscattered wave components exactly to zero. In the case of cross-polar optimal polarizations, a particular correlation difference rather than individual interchannel correlations tends to zero. The utilization of these necessary conditions for the power extrema, furthermore, allows the introduction of an efficient numerical algorithm to compute optimal polarization states for a given target covariance matrix. The presented polarimetric concept is demonstrated to generalize the well-established theory of characteristic polarizations. Analysis of chaff radar data illustrates the efficiency of the outlined approach