文献类型：期刊文献

中文题名：干涉、极化干涉SAR技术森林高度估测算法研究进展

英文题名：Development of Forest Height Estimation Using InSAR/PolInSAR Technology

作者：张王菲[1,2] 陈尔学[1] 李增元[1] 赵磊[1] 姬永杰[2]

第一作者：张王菲

机构：[1]中国林业科学研究院资源信息研究所;[2]西南林业大学林学院

年份：2017

卷号：32

期号：6

起止页码：983-997

中文期刊名：遥感技术与应用

外文期刊名：Remote Sensing Technology and Application

收录：CSTPCD;;北大核心:【北大核心2014】；CSCD:【CSCD2017_2018】；

基金：国家973计划项目"复杂地表遥感信息动态分析与建模"(2013CB733400)

语种：中文

中文关键词：干涉SAR;极化干涉SAR;森林高度;估测;反演

外文关键词：InSAR; PolInSAR; Forest height; Estimation; Inversion

分类号：TP79

摘要：在干涉、极化干涉SAR森林高度估测中,估测算法对结果精度起着决定性作用。通过对现有森林高度干涉、极化干涉SAR研究的系统性分析,总结了现有研究中森林高度估测算法的基本原理、模型假设及其应用局限性,并对这些算法在区域和全球尺度森林高度反演中的潜力进行了分析。总结发现,基于干涉SAR技术的DSM-DEM差分法在森林高度反演中精度较高,与极化干涉SAR算法相比,受到森林类型、结构的影响较小,在区域和全球尺度森林高度反演中具有很大潜力。但是其局限性在于是否能够获取大范围高精度的DEM;极化干涉SAR技术利用了森林的极化散射特点,不受DEM的限制,可以大范围地进行森林高度反演,但是在森林异质性大的区域,仍然需要进一步分析森林特征对不同波长相位及相干幅度的影响,根据森林的微波散射原理拓展微波散射模型,才能进一步提高估测结果和精度。此外,由于单基线干涉SAR、极化干涉SAR对森林垂直结构可见性差,因此,发展多维度、多基线SAR及其相应算法并朝这个方向拓展是未来采用干涉SAR、极化干涉SAR进行森林高度反演的主要方向。
Forest height estimation is one of the hottest research areas of InSAR/PollnSAR technology within its 30 years＇ development.Estimation algorithms play an important role in the forest height assess- ment by InSAR/PollnSAR technologies. This paper systematically reviewed the basic theories, model as- sumptions and then summarized the limitation and potential of these algorithms applied in forest height es- timation, especially performed in regional or global scale. It also deliberated the intrinsic characteristics of these algorithms like DEM difference method, three-stage inversion process, coherence amplitude method and so on.Analysis showed that the estimation results of DEM difference method had higher accuracy and less influence from forest types and structure.So it had great potential for global and regional forest height assessment, however,it was limited by the requirement of high accuracy DEM data in those area.The result accuracy of algorithms based on PolInSAR depended more on forest types, structures and also the robust of forest scattering models.It had no restriction of DEM and could perform in global and regional scale,but for the forest area with great heterogeneous, model and algorithm suitability and robust need to further stud- ying.Besides, for the poor penetration of single-baseline InSAR/PollnSAR, we should focus more on multi- dimension,multi-baseline technique for InSAR/PolInSAR application development in the future.