层析合成孔径雷达（TSAR）是传统二维SAR图像原理的三维扩展，是分析结构较为复杂的体结构场景的重要技术手段。森林垂直结构在森林生态功能及碳循环研究中具有重要作用，已经成为林业遥感研究的重要发展方向。为弥补现有碳储量研究大多集中于热带雨林和森林垂直结构研究多针对平坦地形及未考虑时间去相干的不足，本项目针对复杂地形区域，开展时间去相干与干涉复相干相关关系的定量研究，并建立时间去相干补偿的北方/半北方森林垂直结构反演模型。采用仿真与真实机载数据研究时间去相干模型，应用机载重轨多基线InSAR/Pol-InSAR数据，分别针对较为平坦和地形起伏较大的研究区建立时间去相干补偿的森林垂直结构反演模型。从反演的森林垂直结构中提取易于进行验证的树高等生物参数，分析地形对树高等生物参数反演精度的影响，以期建立复杂地表时间去相干补偿的北方/半北方森林垂直结构反演模型和提高树高等生物参数的反演精度。

Tomographic Synthetic Aperture Radar (TSAR) is the extension of SAR images from traditional two-dimension to three-dimension, and it is an important technology for analyzing more complex structure scenario, such as forest and city. Forest is an important carbon repository in terrestrial ecosystem, and forest vertical structure plays a key role in forest ecological functions and carbon cycle research, and it has become an important trend of forest parameter inversion in forestry remote sensing. As carbon stocking research focuses on tropical forest and the existing forest vertical structure study aims on flat terrain and without considering temporal de-correlation, we carry out this project to compensate for the lack of these. We focus on complex terrain areas to carry out the quantitative research between temporal de-coherence and interferometric coherence, and establish forest vertical structure inversion model in boreal/hemi-boreal forest with temporal de-correlation compensation. Simulation and real airborne data are applied to study temporal de-coherence model, and airborne repeat-path multi-baseline InSAR/Pol-InSAR data are used to establish forest vertical structure inversion models with temporal de-coherence, respectively in relatively flat test site (Ramingstorp) and complex terrain test site(Krycklan). Forest height and other biological parameters extracted from forest vertical structure established in our project is verified using in-situ and Lidar data, and the effects of terrain for the accuracy of forest height and other biological parameter is analyzed. Establishing forest vertical structure inversion model with temporal de-correlation compensation in complex terrain for boreal/hemi-boreal forest and improving the inversion precision of forest height and other biological parameters are expected.