PREDICTION AND MEASUREMENT OF SOIL BIDIRECTIONAL REFLECTANCE

A model for soil bidirectional reflectance distribution functions (BRDF's) in visible and reflective infrared wavelengths is introduced and compared to data acquired in the field. The model is based on the representation of soil surfaces by a collection of opaque spheres sitting on a Lambertian horizontal surface. Soil reflectance is expressed as a function of the horizontal area obscured by a sphere, the amount of surface in shadow, the relative brightness of a sphere, and the fraction of diffuse illumination. These terms are expressed as functions of the view direction, solar direction, and a sphere area index. The model is not sensitive to increases in the sphere area index beyond a value of 0.4. For comparison, soil reflectance factor data were acquired on a tilled field from many view directions and for a range of solar directions. The observed reflectance factor distributions were consistent with those predicted by the function; maximum reflectance occurred in the antisolar direction and reflectance decreased with increasing phase angle. Increasing the surface roughness by different tillage methods did not substantially alter the directional anisotropy of the soil reflectance factors. The model was fit to the data by a nonlinear least squares procedure. The relative rms difference between the predicted and observed reflectance factors was typically less than 15% when the solar zenith angle was less than 50-degrees. The model provided a realistic representation of tilled field reflectance distributions.