基于几何光学模型反演林冠郁闭度的普适性方法研究

针对定量遥感技术在森林监测上的需求，构建基于几何光学理论的林冠郁闭度逐像元反演模型，提出基于多源遥感数据的区域性空间尺度转换方法，根据林冠类型、分布特征、数据尺度效应对反演模型的输入参数进行优化与敏感性分析，解决利用物理学模型在大区域开展长时间序列森林动态监测的普适性问题，实现科学、合理的区域森林生态效益评价。

Monitoring forest structure by quantitatively deriving canopy structural variables, such as crown closure, would enable further understanding and assessments of changes in forest ecosystem services and provide insight into processes that can be of influence. This research focuses on the inversion of physical based Li-Strahler geometric-optical model for estimating the forest crown closure at regional scale. The pixel-based input parameters will be evaluated individually, which influences the accuracy of the model inversion. In addition, a regional scaling-based method will be extracted that can upscale the information from higher spatial resolution data to lower spatial resolution data, and better derive the inputs of the lower spatial resolution image for the model inversion. The inverted Li-Strahler model combined with the scaling method makes it possible to estimate forest crown closure in large areas. By using multi-temporal data, a change detection in forest and the evaluation on forest ecological benefit can also be carried out.