星载LiDAR产品辅助的InSAR森林高度反演

InSAR forest height retrieval based on spaceborne LiDAR products assistance

针对星载重轨InSAR森林高度反演受时间去相干制约与模型解算辅助数据难以获取的问题。考虑ALOS-2 PALSAR-2干涉数据特点,采用一种顾及时间去相干影响的半经验散射模型,利用最新发布的星载激光雷达ICESat-2 ATL08高程产品中的植被高度数据作为辅助数据,并结合主成分分析思想(PCA)对ALOS-2 PALSAR-2相干幅度信息与树高的关联模型进行参数解算。实验结果表明,在ICESat-2树高数据辅助条件下,通过散射模型可以较好抑制时间去相干的影响,进而反演出可靠的模型参数及森林高度(RMSE约为3 m)。本研究验证了联合星载重轨干涉SAR与星载LiDAR数据实现大范围、大尺度森林高度反演的可行性。

Aiming at the problem that the spaceborne heavy-orbit InSAR forest height inversion was restricted by the temporal decorrelation and the acquisition of auxiliary data for model solving.Considering the characteristics of ALOS-2 PALSAR-2 interferometric data,this paper used a semi-empirical scattering model that took into account the effect of temporal decorrelation.By using the canopy height data from the newly released space-borne LiDAR ICESat-2 ATL08 elevation product as auxiliary data,the association model between ALOS-2 PALSAR-2 coherence amplitude information and tree height could be solved with the principle of principal component analysis(PCA).The experimental results showed that under the auxiliary condition of ICESat-2 canopy height product data,the effect of temporal decorrelation could be well suppressed by scattering model,and then the reliable model parameters and forest height inversion result could be obtained(RMSE was about 3 m).This study validates the feasibility of realizing large-scale forest height inversion with joint space-borne repeat-pass interferometric SAR and space-borne LiDAR datas.