An improved adaptive decomposition method for forest parameters estimation using polarimetric SAR interferometry image

Forest parameters estimation using polarimetric synthetic aperture radar interferometry (PolInSAR) images is one of the greatest interests in remote sensing applications. Applying the model-based decomposition concept to PolInSAR data opened a new way for forest parameters estimation. However, the method tends to underestimate the forest height due to reflection symmetry assumption and required the numerical solution of nonlinear equation system. In order to overcome these limitations, an improved adaptive decomposition technique with PolInSAR data is proposed. In this approach, an accurate topographical phase and asymmetry volume scattering model are applied to the model-based decomposition technique for polarimetric SAR interferometry image. The accurate topographical phase is first estimated and then used as the initial input parameter to our numerical method. This approach is not only avoiding large error generated by the constant topographical phase in fluctuating forest areas but also improve the accuracy of forest height estimation and the magnitude associated with each mechanism. The performance of proposed method is demonstrated with simulated data from PolSARproSim software and SIR-C/X-SAR spaceborne PolInSAR images over the Tien-Shan areas, China. Experimental results indicate that forest parameters could be effectively extracted by proposed method.