An efficient method for estimating tropical forest canopy height from airborne PolInSAR data

The study on PolInSAR forest canopy height(FCH) inversion is an essential branch within the SAR field, where the temporal decorrelation(TD) inherent in the interferometric complex coherence of repeat-pass Polarization Interferometric Synthetic Aperture Radar (PolInSAR) data significantly impacts the precision of FCH estimation. This study is grounded in airborne multi-baseline PolInSAR data, employing Random Volume over Ground (RVoG) + VTD and Random Motion over Ground(RMoG) as the FCH inversion models. The classical RVoG threestage method(TS) process was used to simplify the solution of RVoG+VTD and RMoG, and a novel, more efficient method was attempted to further streamline the solution of the RVoG+VTD model(RVoG+VTDs). The results demonstrate that the RVoG+VTDs model obtains the most satisfactory results, and the efficiency of the RVoG+VTDs model is 23 times more efficient than that of the RVoG+VTD; 19 times more efficient than that of the RMoG model, and 22 times more efficient than that of the RVoG model, respectively; and the inversion time consumed by the four models is 1.5 min, 35 min, 28 min, and 33 min, respectively. In the accuracy validation, the estimation accuracy of RVoG+VTDs still keeps the best accuracy, with corresponding R-2 values for the four methods (RVoG+VTDs, RVoG+VTD, RMoG, and RVoG) equal to 0.813, 0.774, 0.804, and 0.770, respectively. The RMSE values are 6.626 m, 7.277 m, 6.774 m, and 7.335 m.The study's conclusions illustrate the feasibility of solving the RVoG+VTD and RMoG model by fixing the extinction coefficients in the model, with RVoG+VTDs taking precedence over the model selection, which can provide a valuable reference for the application of spaceborne PolInSAR data.