Forest height estimation by means of compact PolInSAR data

Abstract

Forest monitoring and management have become a necessity to restrain deforestation and human encroachment. In this respect, the polarimetric synthetic aperture radar interferometry (PolInSAR) techniques have demonstrated their capability for this purpose. Recently, a new mode of SAR systems called compact polarimetry (CP) has been proposed, which can overcome some limitations of fully polarimetric (FP) SAR system. This study aims to illustrate the capability of compact PolInSAR (C-PolInSAR) data for forest height estimation. For this purpose, a novel method based on random volume over ground (RVoG) model and PolInSAR decomposition technique is proposed. The main motivation of the proposed method is to directly and simultaneously estimate the ground only and volume only coherences from the C-PolInSAR data without the need for any prior information. The proposed method was verified with the L-band single-baseline PolInSAR data acquired over the Remningstorp test site, southern Sweden. In addition, light detection and ranging (LiDAR) data were used to vertify the results. The experimental results demonstrated that the C-PolInSAR data i.e., pi/4-PolInSAR and circular transmit-linear receive PolInSAR (CTLR-PolInSAR) data could provide a significant accuracy for forest height estimation compared to the fully PolInSAR (F-PolInSAR) data by 4.2 m and 4.49 m improvements, respectively.