TanDEM-X极化干涉SAR森林冠层高度反演

Forest Canopy Height Retrieval by Pol In SAR with Tan DEM-X Data

森林树高是反映森林蓄积量、生物量、碳储量和立地质量的关键指标。极化干涉SAR技术是森林树高估测的有效手段之一,但数据获取及数据质量一直是极化干涉SAR树高反演研究的主要制约因素。以云南省勐腊县为实验区,基于德国DLR提供的双站模式Tan DEM-X Co SSC全极化干涉数据,采用三阶段算法、以及两种结合极化干涉优化的相位差分-相干幅度法反演了实验区的森林冠层高度。结果表明:三阶段算法的反演结果数值相对合理,与二类调查数据的空间分布格局相似;相位差分-相干幅度法的反演结果在林区出现部分负值,而在非林区却反演出冠层高度、且呈现条纹状,反演结果与二类调查数据的空间分布差异较大、相关性低。因此,三阶段算法结合双站模式的Tan DEM-X全极化干涉数据,能够在实际树高估测反演中发挥作用,为辅助森林资源调查工作提供了新手段。

By using the bistatie TanDEM-X polarimetric and interferometric CoSSC data from German DLR, we used Three-Stage algorithm and Phase Difference-Coherence Amplitude algorithm with two different strategies to retrieve the canopy height of the forest in Mangle of Yunnan Province. The canopy height retrieved by Three-Stage algorithm has reasonable values and similar spatial distribution to the forest inventory data. The two different strategies of Phase Difference-Coherence Ampli- tude algorithm had produced some negative height in forest area and positive canopy height in non-forest area with apparent fringe, and their estimations were badly correlated to the inventory data. Therefore, the bistatic TanDEM-X polarimetrie and interferometric CoSSC data combining with Three-Stage algorithm could play a role in the practical measurement of forest canopy height.