基于体素特征的森林冠层叶绿素反演研究

Canopy Chlorophyll content in forested ecosystems plays important role in plant photosynthesis and can indicate vegetation stress which influenced by the environment. Hyperspectral remote sensing data has been widely used in inversion large area canopy chlorophyll content. However, due to lack of vegetation canopy vertical structure information, the mechanism was unclear when using canopy remote sensing data inverse canopy chlorophyll especially in non-uniformity vegetation areas. Therefore, the study on remote estimation of canopy chlorophyll resistant to complex vertical structure and variations in species is essential for evaluating forest ecology effect. In this study, three different forest community types such as coniferous, broad-leaved and mixed forest which contain different vertical structural characteristics was selected and the airborne hyperspectral and Light Detection and Ranging (LiDAR) data will using to inverse canopy chlorophyll. Firstly，the voxel was used to describe canopy three-dimensional structure, and then the 3D voxel characteristics were established by combining LiDAR and hyperspectral data. Second, the sensitive of canopy spectral responded with the canopy 3D structure was analysis and then the regularity of how canopy 3D structure parameters influenced on canopy chlorophyll inversion was researched based on the model simulated data and in-situ measurements. Third, the regularity between canopy chlorophyll and canopy 3-D structure in community level was applied to airborne hyperspectral and LiDAR data, and canopy chlorophyll content inversion framework considering the vegetation type and canopy 3D structure were established based on these canopy voxels.

森林冠层叶绿素含量在森林生态系统光合作用中起着重要作用，叶绿素也可作为植物受到胁迫的指示性指标，衡量植被对环境的响应。高光谱遥感数据在反演大尺度范围植被叶绿素方面有着广泛的应用。但是由于缺少森林冠层垂直结构信息，群落垂直与水平结构分布的不均匀性对遥感反演冠层叶绿素影响的机制尚不明确。建立一种适应不同群落类型及冠层结构特点的冠层叶绿素遥感反演方法能为森林生态功能评价提供更有效的基础数据。本项目拟利用机载激光雷达与高光谱数据，对针叶林、阔叶林、针阔混交林三种具有不同垂直结构特性的森林群落类型，研究群落冠层三维结构“体素”划分与描述方法；利用野外测量数据及几何-光学辐射传输模型模拟数据，分析冠层“体素”结构特征对冠层光谱的响应规律，揭示群落类型及冠层结构差异对冠层叶绿素反演的影响；利用冠层“体素”特征构建顾及森林冠层垂直结构特点及群落类型的叶绿素反演方法，实现非均质森林覆盖区域冠层叶绿素反演。

森林冠层叶绿素含量可作为植物受到胁迫的指示性指标，衡量植被对环境的响应，同时在森林生态系统的光合作用中起着重要作用。但是由于缺少森林冠层垂直结构及水平结构分布的不均匀性，遥感反演冠层叶绿素影响的机制尚不明确。针对以上问题，本项目建立一种适应估计冠层水平和垂直结构特点的冠层叶绿素遥感反演方法。以湖北省太子山林场马尾松林、马尾松+麻栎混交林、麻栎林三种具有不同垂直结构特性的106块样地为基础，利用激光雷达与高光谱数据研究了群落冠层三维结构“体素”划分与描述方法；利用Geosail模型模拟叶片和冠层尺度不同结构参数对冠层光谱的响应规律，揭示群落类型及冠层结构差异对冠层叶绿素反演的影响；利用构建顾及森林冠层垂直结构特点及群落类型的叶绿素反演方法，实现非均质森林覆盖区域冠层叶绿素反演。结果发现：1）利用激光雷达构建的体素结构可以反映出不同森林群落类型的差异，体素的大小不同对林分特征参数，特别是垂直分层的透过率有一定影响，但是这种差异不足以影响到对针叶林、阔叶林和针阔混交林的识别。2）在植被冠层中，叶绿素的分布有一定差异，冠层叶绿素含量最高的叶片总是顶层的两片叶片，由顶层叶片至底层叶片，叶绿素含量逐渐降低，从植物生理角度上说，顶层叶片更容易获取足够的光照进行光合作用。最顶端叶片的叶绿素含量与整个植株的叶绿素含量的相关性已经较高，决定系数达到0.88，顶端三片叶片的叶绿素含量与整个植株的叶绿素含量的决定系数已经高于0.97，说明顶端三片叶片的叶绿素含量已经能代表整个植株的叶绿素水平。3）在冠层叶绿素反演结果中来看，采用基于物理模型的反演算法反演的精度总体高于基于光谱指数的反演精度；在采用基于物理模型的反演算法中，要包含足够信息的绿色波段、红色波段、红边和近红外波段信息，且光谱波段宽度不高于30nm，冠层叶绿反演中误差为0.24g/m2。