一种改进的PolInSAR PCT方法反演植被垂直结构

A modified PolInSAR PCT method to invert vegetation vertical structure

针对传统PCT方法中"相干相位-幅度联合反演算法"的缺点,采用RVOG模型,利用改进的非线性迭代算法反演植被高、地表相位。改进的非线性迭代算法不仅充分利用不同极化方式对应的复相干系数,同时兼顾复相干系数的先验统计误差,提高参数解算的可靠性,进而提高PCT结果的反演精度。最后,采用两景德国E-SAR数据进行实验,实验结果表明:文中提出的方法能较好地反映植被的垂直结构信息;植被冠层对应的平均相对反射率函数近似服从高斯分布;反演的相对反射率值与植被的种类、密度存在一定关联。

There are shortcomings for ＂phase-amplitude joint inversion algorithm＂ of PCT method. A modified nonlinear iteration method is applied to inverting the vegetation height, ground phase and temporal decorrelation, based on RVOG model. The modified method can not only take full advantage of different complex coherence values of polarizations, but also can consider the prior statistics errors of different complex coherence values. This makes parameter estimations more reliable and products more accuracy result. The modified method can prove two necessary parameter, vegetation height and ground phase. Finally, the new approach is validated on E-SAR data of Germany. It demonstrates that this method can reflect vegetation vertical structure well. The mean relative reflection ratio function follows the Gaussian distribution. And the mean relative reflection ratio values have relationship with vegetation species and density.