745
where c and d are positive constants, s is the angular distance from the sun to a point in the sky hemisphere
(0,<|>), and 0y > s the angle from zenith. Zq is the solar zenith angle. The parameters AO, Al, and A2 are functions
of the atmospheric conditions and control the isotropic, or mean, shape of the diffuse sky distribution, the
horizon or limh brightening, and the shape of the solar aureole, respectively. Figure 5 illustrates the output from
this model for a similar solar zenith and azimuth angle. The main differences between figures 4 and 5 are due to
the scaling of the maximum values within the particular graphics package used. Further work is required to
assess the wavelength dependency of the three parameters and to relate them to a physical measure of
atmospheric loading such as optical depth.
lOO'
Figure 5. Sky radiance anisotropy modelled using three parameters to control the functional form of the
distribution for a similar solar zenith and azimuth angle to figure 4.
3.2 Surface Bidirectional Reflectance Distribution Function (BRDF)
The BRDF of terrestrial surfaces is important in the derivation of surface albedo from remotely sensed data since
surface reflectance from nadir viewing sensors generally provide a poor measure of the albedo. The directional
reflectance of soils and vegetation canopies can be highly asymmetric in the forward and backscattering
directions and it is essential to measure and model these directional effects before interpreting remotely sensed
data from nadir and mult-look direction sensors. Figure 6 illustrates the asymmetry in directional reflectance
from an area of lateritie soils characteristic of the Tiger-Bush sub-site. This three dimensional visualisation is
one of 252 such images for this one target alone which contain data on the spectral variation of the directional
reflectances. A number of visualisation tools are still to be developed before the full information content of the
data can be extracted.
The BRDFs of all the characteristic cover-types will be used in aggregating the directional reflectances over
varying spatial scales from airborne and satellite remotely sensed data to develop a model for the surface
radiation budget at sub-site, super-site and GCM scales.