ISPRS Annals of the Photogrammetry, Remote Sensing and Spatial Information Sciences
Publications Copernicus
Download
Citation
Articles | Volume V-1-2020
https://doi.org/10.5194/isprs-annals-V-1-2020-375-2020
https://doi.org/10.5194/isprs-annals-V-1-2020-375-2020
03 Aug 2020
 | 03 Aug 2020

A COMPOSITE MODEL FOR REFLECTANCE AND POLARISATION OF LIGHT FROM GRANULATE MATERIALS

J. I. Peltoniemi, M. Gritsevich, J. Markkanen, T. Hakala, J. Suomalainen, N. Zubko, O. Wilkman, and K. Muinonen

Keywords: light, scattering, electromagnetic waves, polarisation, reflectance, BRDF, optical properties, Earth Observation, snow, soil

Abstract. Many natural land surfaces, such as sand or snow, consist of densely packed grains, often covered by dust, water droplets, contaminated with other materials such as possible oil leaks, hoar frost, and can also be internally cracked, porous, and heterogeneous. Most scattering models ignore these complications, but here a more detailed approach is taken to test all these effects. The current model is composed of three techniques: 1) Monte Carlo-based electromagnetic volume integral equation technique for non-spherical wavelength scale dust particles, 2) Monte Carlo ray tracing for stochastic-shaped grains much larger than the wavelength, with optional point scattering from dust cover, internal inclusions, and liquid surface layer, in a layer of an optical depths of few units, and 3) adding-doubling to combine smaller layers into an arbitrary, thick and vertically inhomogeneous medium. The model allows the medium to be built in a modular way, and after initialisation, rather complicated layered structures can be computed quickly and flexibly. The computed results are compared against experimental measurements of snow and sand. The model agrees with measurements usually within the measurement accuracy (∼ 0:05). The scattering is observed to depend significantly on grain size, shape, orientation, composition, fine structures, dust, and some other properties that need to be defined. Both, measurement and modelling, require much deeper attention to these properties.