Applying polsar and polinsar to forest structure information extraction

The key research activities and achievements in the field of applying PolSAR and PolInSAR to forest structure information extraction in DRAGON 2 are summarized in this paper. The limitation of the ALOS PolInSAR dataset acquired in the Culai test site for forest height extraction because of its long temporal baseline (46 days), and how the PolInSAR coherence optimization methods can help improve the topography inversion accuracy under forest canopy were presented. We have analyzed and evaluated the capability of multiple polarization parameters extracted from different frequency PolSAR data for forest scar mapping in the Shibazhan test site, and developed the land cover classification method based on SVM (Support Vector Machine) using PolSAR data. With the L-band E-SAR PolInSAR data acquired in the test site in Germany, we developed forest above ground biomass (AGB) estimation approach based on polarization coherence tomography (PCT).