Abstract
Underlying topography plays an important role in the national economic construction, military security, resource exploration and investigation. Since synthetic aperture radar tomography (TomoSAR) can achieve the three-dimensional imaging of forests, it has been widely used in underlying topography estimation. At present, there are two kinds of TomoSAR based on the applied datasets: single polarimetric TomoSAR (SP-TomoSAR) and fully polarimetric TomoSAR (FP-TomoSAR). However, SP-TomoSAR cannot obtain the underlying topography accurately due to the lack of enough observations. FP-TomoSAR can improve the estimation accuracy of underlying topography. However, it requires high-cost data acquisition for the large-scale application. Thus, this paper proposes the dual polarimetric TomoSAR (DP-TomoSAR) as another suitable candidate to estimate the underlying topography because of its wide swath and multiple polarimetric observations. Moreover, three frequently used spectral estimation algorithms, namely, Beamforming, Capon and MUSIC, are used in DP-TomoSAR. For validation, a series of simulated experiments was carried out, and the airborne P-band multiple polarimetric SAR data over the Lope, Gabon was also acquired to estimate the underlying topography. The results suggest that DP-TomoSAR in HH & HV combination is more suitable to estimate underlying topography over forest areas than other DP combinations. Moreover, the estimation accuracy of DP-TomoSAR is slightly lower than that of FP-TomoSAR but is higher than that of SP-TomoSAR.
Highlights
The digital terrain model (DTM) plays an important role in the national economic construction, military security, resource exploration and investigation and other fields [1,2,3]
This paper primarily explores the effectiveness of dual polarimetric (DP)-TomoSAR to estimate the underlying topography and selects three common spectral estimation algorithms (Beamforming, Capon and MUSIC) in TomoSAR to estimate the underlying topography based on dual polarization data
We found that when the signal-to-noise ratio (SNR) is lower, or the difference between surface scattering centres and canopy scattering centres is smaller, the inversion results of Beamforming, Capon and MUSIC algorithms based on DP-TomoSAR remarkably improved compared with SP-TomoSAR
Summary
The digital terrain model (DTM) plays an important role in the national economic construction, military security, resource exploration and investigation and other fields [1,2,3]. FP TomoSAR can acquire scatterers’ heights, and identify their scattering mechanisms [19,20] This technique is helpful for the interpretation of tomograms and can improve the accuracy of underlying topography inversion [21]. To obtain the forest backscattering power along the vertical direction, a number of spectral estimation methods have been proposed, which can be divided into three kinds: nonparametric spectral estimation, parametric spectral estimation, and sparse spectral estimation Among these methods, Beamforming [28,29], Capon [30] and MUSIC [31,32,33] are three frequently used estimators. The LiDAR data were applied to assess their performance under different polarimetric modes
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
