基于高光谱与激光雷达的林冠物种与功能多样性遥感监测方法研究

本研究立足于生态与遥感相融合，发展基于遥感的森林生物多样性监测方法。利用机载高光谱和激光雷达数据，从林冠物种-生化-光谱和物种-结构-点云的内在关联入手，在单木尺度上结合最优叶片生化组分和最优冠层结构参数，提出基于聚类算法的林冠物种多样性遥感估算模型；从单木功能性状角度出发，探讨反映多样性及生态系统功能的最优生理和形态性状，提出基于单木多维性状的林冠功能多样性遥感监测与区域制图方法；从方法适用性角度综合分析两种多样性监测方法所独有及共有的问题与影响因素，包括基于单木与像元估算的差异、聚类的光谱饱和以及监测尺度的依赖性问题，并在两个研究区开展对比分析，深入探讨森林冠层生物多样性区域尺度遥感监测方法的普适性，突破遥感技术在森林生物多样性估算上的方法创新，为生物多样性保护与评估提供新的思路。

This research focuses on forest species and functional diversity estimation at regional scale using remote sensing methods. Based on airborne hyperspectral and LiDAR data, the relationship between leaf biochemical and spectral properties and their inner linkage with species diversity will be firstly explored. Then a clustering model will be developed for mapping forest species diversity by inter-species variations in biochemical components and structural properties of individual trees. Meanwhile, a multidimensional trait space will be built up by ecologically relevant and functionally important traits of individual trees, and a moving window method will be applied to map the regional forest functional diversity. Finally, the applicability and limitation of the proposed diversity monitoring methods will be evaluated and compared among two subtropical forest study areas, mainly including the differences of individual tree-based and pixel-based estimations, the clustering saturation and the scaling issue. This research could provide a technical foundation for future regional forest biodiversity monitoring and will have important implications for forest conservation and management.