遥感成像机理蒙特卡罗模拟的若干计算技术

Key Computing in Monte Carlo Method for Remote Sensing Mechanism Study

给出植被遥感像机理研究中蒙特卡罗方法几个关键性计算技术的详尽推导,它们包括各种叶角分布下的叶片取向,叶冠层在任意方向的遮光系数,漫反射光取向和叶片局部坐标与冠层全局坐标的转换。这些计算涉及到一些辐射传输,概率论,解析几何和数值积分方面的知识,相信对相关工作会有借鉴意义。最后用强后向散射验证了文中的计算方法。

Several key computing methods in Monte Carlo simulation for investigation of remote sensing imagery mechanism are introduced. The leaf angle distribution function describing inclination and orientation of leaves in a canopy is written in a probability density function. Six essential leaf angle distribution functions and the integrating algorithm are given. The light intercepted probability at any direction in a canopy is expressed by the G-function. Integrating nodes and weights of 24-nodes Gauss double precision integration algorithm are listed. The way to obtain the photon fly direction for Lambertian reflection and emission is deduced. Photons move and hit leaves in canopy global coordinate system though the photons are launched or reflected in leaf local coordinate system. Judging rules whether a photon hits the above surface or the under surface of a leaf or neither both are discussed. Two matrixes, which transforms the direction in leaf local coordinate system into the global coordinate system, are showed.