An Interferometric Coherence Optimization Method Using Particle Swarm Optimization in Radar Polarimetry for Improvement of Forest Height Estimation

Since natural forests store a large quantity of carbon, it is essential to estimate biomass in assessing carbon dioxide in the atmosphere and climate changes. The significant problem with interferometric radar techniques is the saturation level of different wavelengths and relevant polarizations due to the fact that structural parameters are not considered. Height is one of the structural parameters that can be estimated using interferometric techniques. Height estimation can be improved using coherence optimization. In this paper, DEM differencing, coherence amplitude, combined and polarization coherence tomography methods are applied to estimate the height using simulated data in L and P bands. DEM differencing method underestimates results up to 14m in P band and 11m in L band respectively. Coherence amplitude method in comparison with DEM differencing method reaches to better results and the differences reduce to 8m in P band and 2m in L band respectively. Results of combined and PCT method are approximately the same and the differences reduce to 2m but using coherence optimization in PCT method offers better results. Better results in comparison with previous methods were obtained using coherence optimization and particle swarm optimization proposed in this paper. The differences results were reduced to 0.5m.