Method Of TanDEM-X InSAR Sub-canopy Topography Inversion By Combining Phase And Coherence Information

High-resolution,high-precision,and large-scale sub-canopy topography plays an important role in resource surveys,military defense and other aspects.In the forest area,the optical remote sensing measurement is a digital surface model,and the forest height needs to be deducted to obtain the sub-canopy topography,but it is difficult to obtain high-resolution,high-precision tree height data in practice.Interferometric However,Synthetic Aperture Radar(In SAR)technology combines microwave imaging and interferometry to obtain terrain elevation.Due to the penetrability of microwaves,the echo signal contains information on the vertical structure of forests,which provides the possibility for subcanopy topography mapping.In particular,the new-generation dual-station In SAR system represented by Tan DEM-X provides high-quality interferometric data for sub-canopy topography mapping,and has important application potential in sub-canopy topography mapping.However,due to the weak penetration of the X-band,the corresponding In SAR DEM contains obvious vegetation high signals,and the Tan DEMX system mainly acquires single-polarization data,and the lack of observation information makes it impossible to directly separate the vegetation scattering and the surface scattering.Scatter information.Aiming at the problems existing in the acquisition of sub-canopy topography from Tan DEM-X In SAR data,the following research will be carried out in this paper.The details are as follows:(1)A single-baseline Tan DEM-X In SAR sub-canopy topography inversion method is proposed,the SINC model and spaceborne Li DAR ground point data are introduced,and a joint altimetry model of In SAR phase and coherence amplitude is constructed to achieve high precision and high resolution under non-penetrating conditions subcanopy topography inversion.The new method first estimates the penetration depth using a simplified vegetation scattering model(SINC model);then,introduces sparse surface elevation points and estimates the phase center height using the empirical relationship between penetration depth and phase center height;finally,the estimated The phase center height was removed from the In SAR DEM to obtain the sub-canopy topography.The single-baseline data covering the experimental areas of Krycklan in Sweden and Teruel in Spain were selected for the experiment.The results showed that the accuracy of the two experimental areas increased by 48% and 30% respectively.(2)A dual-baseline Tan DEM-X In SAR sub-canopy topography inversion method was developed.Based on the RVo G model,an In SAR scattering phase center height inversion model under penetration conditions was established,which effectively suppressed the effect of forest scattering on In SAR sub-canopy topography measurement.The idea of the new method is to introduce multi-baseline data to solve the problem that it is difficult to accurately estimate the height of the phase center when there are differences in forest tree species,density and other attributes.Therefore,this paper uses the dual-baseline data combined with the RVo G model to establish a phase center height estimation model,and deducts it from the In SAR DEM to obtain the sub-canopy topography.In this paper,the double-baseline data of the experimental areas of Krycklan in Sweden and Teruel in Spain are used to invert the sub-canopy topography.The accuracy of the two experimental areas is 2.72 m and2.73 m,respectively.(3)A large-scale sub-canopy topography inversion method with joint single/dual baselines is constructed,which provides a method reference for large-scale sub-canopy topography inversion using Xband dual-station satellites.The new method is based on the distribution characteristics of large-scale data and combines single and double baselines to invert sub-canopy topography: 1)For single-scene image areas,based on the single-baseline inversion method,the SINC model with empirical parameters(C-SINC model)is used to represent the phase center height;2)The double-baseline inversion method is used for the overlapping area of the two images,and the empirical parameters of the CSINC model are estimated by the inversion results,and the phase center height is represented by the phase information of the RVo G model.Finally,combined with the single/double baseline phase center height estimation model,the large-scale sub-canopy topography covered by multiple images was obtained.The Teruel experimental area in Spain is selected to verify the new method,and the inversion accuracy is 5.98 m.