顾及生物物理参数的极化干涉SAR几何和物理观测量联合反演植被高度及DTM

When using traditional surveying methods to measure topography, it is difficult to remove vegetation height. Polarimetric SAR interferometry (POLInSAR) technique provides opportunity to solve this problem. However, there are some problems for POLInSAR. For example, the effect of vegetation biophysical parameters on scattering mechanism is not clear; the scattering model is imperfect; it is difficult to establish function relationship between physical parameters and observations as only simplified multi-step method or intelligent optimization method is used; accuracy assessment is unimplementable, multi-platform and multi-baseline integration is unrealizable. The project will investigate the influence of vegetation parameters (vegetation height, DBH (diameter at breast height), canopy-fill-factor, CW (crown width),forest crown closure, etc.) on POLSAR scattering mechanisms, based on which establish the sophisticated scattering models. Then, we will introduce modern survey adjustment theory to develop survey adjustment theory that unites POLInSAR geometric and physical observations, including establishing survey adjustment descriptions, survey adjustment criteria, methods for parameters estimation and accuracy assessment. After that, we will develop a synergetic inversion method for extracting vegetation height and DTM by using multi-platform and multi-baseline POLInSAR data. The project is important for solving the problems of removing the vegetation height in topographic mapping and DEM production. It is also significant for enriching the theory of survey adjustment and POLInSAR.

常规测绘手段获取地形时无法准确去除植被高，极化干涉SAR为解决这一问题提供了契机。而目前极化干涉SAR存在植被生物物理参数对极化散射的影响机理不明确、散射模型不完备，解算采用简化的分步算法或智能优化算法，导致难以建立观测量与物理量之间的函数关系、无法评定精度、难以实现多基线和多平台协同反演等问题。本项目拟研究树高、胸径、冠层填充因子、冠幅、郁闭度等植被生物物理参数对极化SAR散射机理的影响，建立精准的极化散射模型。在此基础上，引入现代测量平差的思路，发展极化干涉SAR几何与物理观测量联合平差理论，包括极化干涉SAR优化的测量平差表达、平差准则、参数解算和精度评定方法；建立基于联合平差的多平台、多基线极化干涉方法协同反演植被高和DTM。项目对解决地形图测绘和DTM制作中去除植被高的难题、丰富测量平差和极化干涉理论具有重要意义。

针对现有极化干涉SAR技术无法精确去除植被高并获取高精度林下地形的问题，本项目首先分析植被冠层多样性、排布规律、地形坡度、土壤湿度对极化SAR散射机理的影响。在此基础上，建立了引入植被生物物理参数的散射模型，包括考虑植被垂直结构的各向异性，枝干垂直向密度分布，地形坡度和时间去相干效应等，使得利用极化干涉SAR技术反演树高和林下地形适用于不同的场景。接着，建立了极化散射模型与PolInSAR几何和物理观测量之间的映射关系，以及PolInSAR几何与物理观测量的联合平差理论，包括函数模型、随机模型、平差模型、平差准则以及精度评定方法，克服了传统PolInSAR难以建立观测量与物理量之间的函数关系、无法进行精度评定的缺点；最后，建立多基线、多波段PolInSAR框架下植被生物物理参数与观测量之间的关系，把PolInSAR几何与物理观测量的联合平差扩展到多基线、多波段配置下，准确估计植被高度及DTM，克服了PolInSAR技术难以实现多基线和多平台联合解算、解算精度低等缺点。本项目解决了地形图测绘和DTM制作中去除植被高的难题、丰富了测量平差和极化干涉理论。