Persistent Scatterer Selection Incorporating Polarimetric SAR Interferograms Based on Maximum Likelihood Theory

Persistent scatterer (PS) interferometry analysis has been performed using single-polarimetric synthetic aperture radar (SAR) interferograms, but newer SAR satellites enable the use of polarimetric data for such analyses. In this paper, to analyze polarimetric SAR interferograms, we proposed the maximum likelihood method at the PS selection stage. With the proposed method, interferograms in a certain polarimetric component exhibiting high coherence contribute to the estimation of surface displacements, whereas noisy interferograms of another polarimetric component are less used. The proposed method was examined by analyzing full- and dual-polarimetric interferograms using data obtained from ALOS/PALSAR images in full-polarimetric acquisition mode. The results showed an increase in the number of coherent pixels used to map surface displacements, and that number increased particularly in the analysis of full-polarimetric data. For example, in the case of full-polarimetric interferograms in linear HV basis, the number increased by a factor of 1.8-3.8 compared with those of single-polarimetric interferograms and by a factor of 1.14 compared with a simple integration of the results of single-polarimetric data. Our method increases the accuracy of the detection of the detailed spatial displacement pattern of a displacement phenomenon.