Underlying Topography Estimation over Forest Using Maximum a Posteriori Inversion with Spaceborne Polarimetric SAR Interferometry

Xiaoshuai Li,Xiaolei Lv,Zenghui Huang

DOI: https://doi.org/10.3390/rs16060948

IF: 5

2024-03-09

Remote Sensing

Abstract:This paper presents a method for extracting the digital elevation model (DEM) of forested areas from polarimetric interferometric synthetic aperture radar (PolInSAR) data. The method models the ground phase as a Von Mises distribution, with a mean of the topographic phase computed from an external DEM. By combining the prior distribution of the ground phase with the complex Wishart distribution of the observation covariance matrix, we derive the maximum a posterior (MAP) inversion method based on the RVoG model and analyze its Cramer–Rao Lower Bound (CRLB). Furthermore, considering the characteristics of the objective function, this paper introduces a Four-Step Optimization (FSO) method based on gradient optimization, which solves the inefficiency problem caused by exhaustive search in solving ground phase using the MAP method. The method is validated using spaceborne L-band repeat-pass SAOCOM data from a test forest area. The test results for FSO indicate that it is approximately 5.6 times faster than traditional methods without compromising accuracy. Simultaneously, the experimental results demonstrate that the method effectively solves the problem of elevation jumps in DEM inversion when modeling the ground phase with the Gaussian distribution. ICESAT-2 data are used to evaluate the accuracy of the inverted DEM, revealing that our method improves the root mean square error (RMSE) by about 23.6% compared to the traditional methods.

environmental sciences,imaging science & photographic technology,remote sensing,geosciences, multidisciplinary

What problem does this paper attempt to address?

The main problem that this paper attempts to solve is to accurately estimate the digital elevation model (DEM) of the terrain under forest - covered areas by using the inversion method based on maximum a posteriori probability (MAP) and combining with space - borne polarimetric SAR interferometry (PolInSAR) data. Specifically, the paper proposes a new method to overcome the problems existing in traditional methods, such as:

1. **Dual - candidate effect**: When using the random volume over ground (RVoG) model for parameter solution, there are two possible solutions for the ground phase, which leads to the uncertainty of parameter inversion. By introducing prior knowledge, especially the probability distribution model of the ground phase (such as Von Mises distribution), the paper effectively solves this problem, making the estimation of the ground phase more continuous and accurate.

2. **Computational efficiency**: The traditional MAP method needs to perform time - consuming exhaustive search to determine the ground phase, which greatly reduces the computational efficiency. For this reason, the paper proposes a four - step optimization (FSO) method, which uses gradient optimization techniques to significantly improve the computational efficiency while maintaining the accuracy.

3. **DEM accuracy improvement**: By using the improved MAP method, the paper shows in the experiment that its method can effectively reduce the height jump phenomenon in DEM inversion, and through the comparison and verification with ICESAT - 2 data, this method improves the root mean square error (RMSE) by about 23.6% compared with the traditional method.

In general, this paper aims to provide an efficient and accurate method for extracting DEM of forest - covered areas from space - borne PolInSAR data, so as to provide support for forest resource management, disaster prevention and other fields.