Structures visible on soil profiles contain transport-relevant information (pedological horizons, boundaries between them, biopores, skeleton, and other macroscopic features). The knowledge of the shape and connectivity of such structures is a first order approximation for predicting flow and transport under given boundary conditions. We measure the spectral reflectance of exposed soil surfaces (e.g. on soil profiles, soil samples, or the horizontal soil surface) to map the geometry of such structural regimes. We make use of the fact, that reflectance of light by dense particulate media depends on complex refractive index, particle size, and porosity. We propose a light scattering model based on the four-flux approach of Maheu et al. to describe radiative transfer of visible light (VIS) and infrared radiation (IR) through soil slabs. Principal input parameters
are wavelength, mean grain size, slab thickness, porosity and refractive index. In case of soil samples, where all parameters but one are known, the unknown can be estimated by comparing the measured and calculated reflectance and transmittance spectra. It follows for example, that the refractive index have to be known to solve the problem for the mean grain size. The refractive index for soils is rarely published. Therefore we demonstrate the variability of the refractive index for soils on the basis of three different soil materials. To test the model, we measured the reflectance and transmittance for soil samples as a function of particle size fraction and slab thickness. We found that the fitted complex refractive indices -- a constant of the material -- depend mainly on the type of soil material and less on particle size or slab thickness.