Search

Bidirectional reflectance distribution functions (BRDFs) of natural surfaces which are best represented by measured biconical reflectance factors (BRDFs) frequently do not fulfill the Helmholtz reciprocity requirement, although theory does not permit deviations. A new set of measured radiance data has been used to derive the BCRF by two different techniques to investigate the validity of the reciprocity requirement in measured data. Both techniques consider sky radiance. The interpolation technique replaces missing measurements by interpolated values. The resulting BRDFs show deviations from reciprocity but reproduce the measured radiances perfectly. This technique has been previously used. The new reciprocity technique replaces missing measurements by the requirement of reciprocity. The resulting BRDFs fulfill reciprocity completely but partly fail to reproduce the measurements. A discussion is given on the differences between the two sets of BRDFs, which are mostly small, and their possible reasons. The conclusion is that the truth is not known but may be in between the results obtained from the two techniques.

The directional reflected radiation of natural surfaces may change even if nothing save the distribution of the irradiation over the hemisphere varies. This is due to the angular anisotropy of the reflection properties of natural surfaces. The quantitative determination of this effect for four different vegetated surfaces is the aim of this investigation. In this paper, results for the first of the four surfaces, a savannah, are shown. The directional reflected radiation may change by ± 1% per degree change of the solar zenith angle and by ± 1 per 6% change of the spectral atmospheric turbidity factor at 0.52 μm.

The albedo of four vegetated surfaces was investigated to derive its variability with differing distributions of the irradiance. The results are based on measured values of the spectral biconical reflectance factor, which are combined with calculated spectral irradiances for low and high atmospheric turbidity. The solar zenith angle is varied from 0° to 80°. The derived spectral albedos are then integrated with respect to wavelength in order to achieve the albedo. It is found that the variability of the albedo with respect to the atmospheric turbidity is less than 0.01 in nearly all cases. The variability of the albedo with respect to the solar elevation angle, however, is larger than 0.02 in many cases. For solar elevation angles from 20° to 60°, the variability of the albedo of the four surfaces can be represented by a mean curve which fits the individual variabilities with an accuracy of 0.015.

The average variability of the reflected radiation field due to differing distributions of the irradiation in case of a savannah was given in a previous paper. In this paper, equivalent results are given for three more surfaces: bog, pasture land, and coniferous forest. Because the results are rather similar, mean values for vegetated surfaces can be derived. They indicate a change of the reflected radiance by ± 1% per degree change of the solar zenith angle and per 10% change of the optical depth of the atmosphere.