TREE CANOPY HEIGHT ESTIMATION USING MULTI BASELINE RVOG INVERSION TECHNIQUE

Abstract

Polarimetric Interferometric Synthetic Aperture Radar (PolInSAR) technique utilizes the characteristics of both SAR polarimetry and Interferometry. PolInSAR technique is proved to be very useful for vegetation parameters retrieval. Estimation of the tree canopy height parameter is very important for the estimation of the Above Ground Biomass (AGB). The baseline separation between different PolInSAR datasets has a very important role in the tree canopy height estimation due to the sensitivity of the baseline to the tree height and the forest structure. So for accurately estimating the tree canopy height of a forest with varying tree heights and species several pairs of PolInSAR datasets with different baselines separations are required. Multi-baseline Random Volume over Ground (RVoG) inversion technique is the most successful method for tree height inversion. UAVSAR, the Quad-Pol L-band airborne SAR of JPL/NASA acquired PolInSAR datasets over the Gabon forest as a part of the AfriSAR campaign. Nine PolInSAR SLC datasets of this campaign acquired over the Mondah Forest site of Gabon forest is used for this study. Tree canopy height map produced from this datasets shows that the tree height is varying at this site and has a maximum height of 50 m. The results obtained are validated using the field data collected by JPL/NASA during March 2016. The comparison of the results with the field data showed that both are in good agreement with an average deviation of 3.75 m.