Vertical resolution enhancement by applying polarimetric spectral analysis techniques to multibaseline InSAR data

Polarimetric spectral analysis techniques have been recently introduced for processing multibaseline interferometric synthetic aperture radar data that are affected by the speckle phenomenon. In this paper, their phase estimation accuracy for retrieving the three-dimensional structure of scatterers from measurements corrupted by multiplicative noise is examined. The performance of the polarimetric beamforming, Capon, and MUSIC algorithms is investigated by Monte Carlo simulations and compared with the polarimetric Cramér-Rao bound. It is demonstrated by the CRLB that radar probing systems possessing polarization diversity offer the potential of considerably improving the information extracted from the observation space enlarged by scattering polarimetry. Moreover, it is shown that the polarimetric algorithms lead to noticeably enhanced estimation precision, in particular in terms of refined resolution, under the condition of polarization diversity.