Polarimetric classification of trees

Since scattering matrix acquisition capability will be equipped with space-borne radar systems such as ALOS PALSAR to be launched in 2003, radar polarimetry is now an indispensable tool for monitoring the Earth´s cover. The advantage of using the vector nature of electromagnetic waves is the information on the amplitude, phase and their relations pertaining to target. In this paper, we retrieve a useful index for classifying broad leaf trees and conifer (needle like leaf) trees based on the scattering characteristics of tree canopies. The index used is the polarimetric correlation coefficient derived by the elements of the Sinclair scattering matrix. It is shown that correlation coefficient defined in the circular polarization basis serves to distinguish conifer tree and broadleaf tree