植被冠层光学信号饱和特性对LAI反演低估的影响研究

Leaf area index (LAI) inversed from remote sensing dataset is affected by physical models, inversion method, and data quality. However, the canopy optical signal saturation is an important factor causing LAI underestimation, especially for the dense vegetation. The reason causing canopy optical signal saturation is the limited optical penetration depth. Therefore, based on the definition of penetration depth in electromagnetic theory, this study forms the theoretical definition and expression for the penetration depth of canopy optical signal. Moreover, based on the modified multi-layer radiative transfer model which with a parameter describing the vertical distribution of foliage area density, this study defines the expression of parameter which describing the contribution of each layer outgoing radiation to total canopy radiation. This study aims to analyze the characteristic of canopy optical signal saturation based on radiative transfer theory. In order to quantitatively analyze and validate the level of canopy optical signal saturation, this study designs and carries out the field experiments for dense crop canopy and forest types. At last, this study extracts the prior knowledge such as the vegetation height and vertical structural parameters making use of the LIDAR or microwave technology, and complements the vegetation information under the saturation position. Based on the above information, this study improves the LAI inversion accuracy during saturation for dense vegetation canopy.

遥感反演的叶面积指数精度受物理模型、反演方法和数据质量的影响，但是对于浓密植被冠层，植被冠层光学信号饱和是造成叶面积指数低估的重要原因。而导致冠层光学信号饱和的原因是光学有限的穿透深度，因此基于电磁波穿透深度概念，形成植被光学穿透深度的理论定义和表达。此外，基于考虑植被冠层垂直结构参数的多层模型，定义分层出射辐射对冠层总辐射贡献的参数表达，从辐射机理上探讨植被冠层光学信号饱和特性。设计并开展野外实验定量分析与验证浓密作物和森林类型植被冠层光学信号饱和程度。最后，借助激光雷达或微波技术提取植被先验知识如植被高度和垂直结构参数，补充饱和位置以下的植被信息，改善浓密植被冠层LAI反演精度。

基于遥感技术提取叶面积指数（LAI）参数时，对于浓密植被冠层，植被冠层光学信号饱和是造成LAI低估的重要原因。研究通过引入叶面积密度和叶片结构参数的垂直分布函数，扩展了多层多组分SAILH冠层反射率模型，利用河北怀来实验场玉米光谱实测数据，验证扩展的多层多组分SAILH模型与单层SAILH和ACRM冠层反射率模型的精度，结果表明扩展的多层多组分SAILH模型与地面实测光谱数据的均方根误差（RMSE）和偏差（BIAS）最小。根据冠层反射率分层贡献描述冠层反射率饱和问题，并用于分析由于参数垂直结构差异造成的LAI低估程度。研究中定义分层贡献率为不同高度处植被出射辐射占冠层总辐射比例，研究结果表明由于冠层内部消光作用导致LAI存在约6%-16%的低估。针对浓密植被冠层，采用融合归一化差值植被指数（NDVI）和近红外波段（NIR）反射率构建的新植被指数（NIRv）降低对LAI高值的饱和程度，此外联合使用多角度反射率数据，降低单一角度反射率饱和对LAI反演精度的影响，提高中高分辨率遥感数据LAI反演精度。