植被二向性反射统一模型研究

Faced with severe challenges from global changes, only establishing reliable vegetation BRDF unified models, can we find feasible retrieval methods and provide reliable global products of vegetation parameters for different vegetation types. In this study, we abandon the traditional modeling ideas of radiative transfer models for continuous vegetation and geometry optical model for discrete vegetation. Therefore, different types of vegetation are unified, considering discrete vegetation as clumped leaf clusters at different scales . Based on the common character of bidirectional reflectance of different canopies, we start from the theories and analysis and establish a BRDF unified model with definite physical meaning and analytical form. At the same time, a multiple scattering model is set up, which can accurately describe the multiple scattering between soil and canopies and is easy for retrieval. Our purpose is making complex bidirectional reflectance simple and clear. Then taking LAI and clumping index retrieval as an example，retrieval strategies and methods of different vegetation parameters will be studied. The scaling effects of remote sensed vegetation parameters can also be re-analyzed to adapt the needs of the global vegetation biological parameters monitoring.

面对全球变化的严峻挑战，就千差万别的植被类型而言，只有建立了可靠的植被BRDF统一模型，才有可能找到可行的反演方法，提供全球意义上的可靠的植被参数产品。本研究将抛弃连续植被-辐射传输模型，离散植被-几何光学模型的传统BRDF建模观念，将离散植被特性视为叶簇在不同尺度下群聚方式，由此将各类植被统一起来。基于不同植被二向性反射特性的共性，从理论分析着手，建立一个有明确物理意义，具有解析形式的BRDF统一模型。同时建立能准确描述土壤-冠层间多次散射的，且易于反演的多次散射模型。力图使复杂的二向性反射的表达变得简单明了，并以LAI和聚集指数反演为例研究不同植被生物参数的反演策略与方法，以适应全球植被生物参数监测之需要。

面对全球变化的严峻挑战，就千差万别的植被类型而言，只有建立了可靠的植被BRDF统一模型，才有可能找到可行的反演方法，提供全球意义上的可靠的植被参数产品。在对现有模型分析的基础上，本研究取长补短，给出了适用于不同入射光条件和植被类型的一次散射与多次散射反射率解析表达。首先，基于几何光学原理，把立足点由树冠移至植被元素，一次散射项的非各向同性特征可以归因为太阳-目标-传感器三者间的几何关系、多尺度群聚叠加和太阳直射与天空散射对目标辐照度之比三个方面。视不同植被类型为多尺度群聚叠加的结果，以泊松分布为基础，本文重新给定了四分量面积比例的求取方法，使之适用于各种植被类型，进而获得一次散射项的近似解析表达式。假定G=0.5,本文借助“再碰撞概率”理论，考虑了冠层内部、冠层与土壤间的多次散射过程，本文也获得了不同天空散射光条件下多次散射项的近似解析式，并揭示了多次散射项的形成原理。利用玉米冠层BRDF数据及加拿大黑杉林BRDF数据对模型的可靠性进行了初步验证。以LAI和聚集指数反演为例研究不同植被生物参数的反演策略与方法，以适应全球植被生物参数监测之需要。