森林冠层绿色FPAR的高光谱遥感反演研究

A vegetation canopy is composed primarily of photosynthetically active vegetation and non-photosynthetic vegetation (NPV;e.g.,branches and stems),but only the photosynthetically active radiation(PAR) absorbed by green leaf is used for photosynthesis. Therefore, there is a need to partition PAR absorbed by canopy into PAR absorbed by green leaves and by NPV.Accordingly,canopy fraction of absorbed photosynthetically active radiation (FPAR)should be partitioned into leaf FPAR (i.e.green FPAR)and NPV FPAR. Green FPAR is a key biophysical variable for estimating gross primary productivity (GPP) based on light use efficiency model, but it is difficult to directly measure green FPAR.Therefore, it is necessary to retrieve green FPAR using remote sensing technology . Presently, because of lack of the study on green FPAR, it is unclear about the correlation between green FPAR and canopy FPAR, influence factors of green FPAR and how to invert green FPAR using remote sensing data. Therefore, this project will carry out the research on green FPAR using remote sensing data. In this project, the forest is chosen as the study subject and a combined model of the geometric model and the turbid media model, GeoSail model, is used to calculate and partition canopy FPAR. The canopy structural and optical parameters of several tree species in Dinghushan ecosystem research station are measured as model input . Based on these parameters data and GeoSail model, a simulated dataset including canopy FPAR, green FPAR and canopy reflectance is constructed. On the basis of the dataset, the research topics conducted in this project include: 1) Analyze the quantitative relationship between canopy FPAR and green FPAR and discuss the inherent mechanism from the view of canopy structure; 2) Conduct the sensitivity analysis of input parameters, and disclose main influence factors of green FPAR; 3) Analyze correlation between green FPAR and canopy reflectance and construct inversion model of green FPAR based on remote sensing data. Further, the uncertaintity of model is analyzed and discussed based on flux data. This research aims to provide theory,model and method for remote sensing monitoring of green FPAR, and further provide effective data for productivity estimation and carbon cycle research.

冠层绿色叶片的光合有效辐射分量(绿色FPAR)能真实反映植被与外界进行物质和能量交换的能力。利用遥感反演实地测量困难的绿色FPAR，对提高生产力遥感估算精度具有重要意义。本项目针对国内外对绿色FPAR 遥感反演研究的不足，选择多个典型树种，在测量冠层组分光谱、冠层结构参数及背景反射率的基础上，采用模型模拟的方法构建不同情景下冠层FPAR、绿色FPAR 以及冠层光谱数据集，进而研究绿色FPAR 和冠层FPAR 之间的定量关系及内在关联机制；分析绿色FPAR 对参数的敏感性，揭示绿色FPAR 的主要影响因子；探索绿色FPAR 反演方法，构建稳健可靠的绿色FPAR 遥感反演模型；结合地面观测数据，检验模型精度，分析不确定性。本项目旨在为森林冠层绿色FPAR 的直接遥感监测提供理论、模型和方法支持，进而为生产力估算和碳循环研提供科学数据支持。

森林冠层包含光合组分和非光合组分，但只有光合组分吸收的光合有效辐射用于光合作用形成有机组织，冠层光合组织确定的光合有效辐射分量(绿色FPAR)能真实反映植被与外界进行物质和能量交换的能力，但是冠层组分分离的困难性导致冠层绿色FPAR研究进展缓慢，其变化特征和机制仍不清晰。本项目针对当前冠层绿色FPAR遥感反演研究不足的问题，以冠层辐射传输模型为手段，以落叶阔叶林和常绿针叶林为研究对象，从理论上探讨森林冠层绿色FPAR的遥感反演方法，取得了满意的结果。首先对传统的辐射传输模型进行改进，实现冠层绿色FPAR的分离，同时根据森林生态系统特征，设定模拟方案，模拟了森林冠层光谱特征和绿色FPAR，进而分析了冠层绿色FPAR随冠层结构的变化特征。结果表明，冠层绿色FPAR随着冠层植被指数和光合组分在冠层中的比例增加而增加；其次，采用相关分析法分析冠层绿色FPAR与高光谱数据之间的相关关系，表明冠层绿色FPAR与近红外波段反射率呈显著的线性正相关，而与可见光波段负相关，但是不存在敏感的独特窄波段；第三，从不同角度分析冠层绿色FPAR与NDVI、EVI以及SAVI的相关性并借助大气辐射传输模型分析绿色FPAR与植被指数关系的抗大气干扰性，发现EVI与冠层绿色FPAR相关性最高，并且两者之间的关系对大气影响的敏感度低；最后，基于森林生态观测站的涡度相关通量数据及气象数据检验冠层绿色FPAR与植被指数之间的关系。结果表明，无论是温带落叶林还是热带常绿林，EVI与森林冠层绿色FPAR都显著相关且具有最高的相关系数，证实了基于辐射传输模型的推演结果。本研究以辐射传输模型为依托，从理论角度推演冠层绿色FPAR的特征、敏感波段和指数，旨在为森林冠层绿色FPAR 的直接遥感监测提供理论和方法支持，本研究结果对于提高森林生态系统生产力的估算精度、推进生态系统碳循环的研究具有重要意义。