2016 年 36 巻 3 号 p. 225-235
Recently, to assess climate change, multi-angular optical remote sensing, especially satellite remote sensing, has attracted attention for precise monitoring of vegetation activity. Most ground objects have anisotropic properties such that the radiance of an object varies under different illumination and observation angles. Such optical properties are defined by Bi-directional Reflectance Distribution Functions (BRDF). Using BRDFs, one can estimate the three-dimensional structure of a target such as tree height or crown shape, which is difficult to estimate based solely on one-directional optical remote sensing.
This review classified recent BRDF studies of vegetation into three categories: (1) observational studies, (2) simulation studies based on physical models, and (3) simulation studies based on semi-empirical models. Next, we discussed the present status of each category. We described some difficulties of integrating each category to develop BRDF studies: (1) observational data insufficiency and (2) impossibility of measuring many parameters used in BRDF models. To overcome these problems, we proposed a new BRDF observation system using an unmanned aerial vehicle (UAV). Additionally, we introduced a Bidirectional Reflectance Simulator (BiRS), which can estimate BRDFs based on observable parameters alone.