Not all the radiation that reaches the downlooking scanner
originates at the surface element being viewed. There is a quantity
called path radiance, which is extraneous radiation that is scattered
into the field of view by the atmosphere.
Figure 5 shows that path radiance depends strongly on visual range
and wavelength. The wavelength dependence is much different than that
for the irradiance shown in Figure 4. The reason for this difference
is that path radiance depends on the reflectance spectrum of the objects
surrounding the material being viewed. In this instance, the spectrum
is that of green vegetation.
Path radiance also depends on the scan angle from nadir and the
location of the sun. Figure 6 shows the scan-angle dependence of path
radiance for three different sun positions on a very hazy day. The
scan plane includes the sun, and one readily can see the characteristic
peak in backscattered radiance at the antisolar angle. When the scan
plane does not include the sun, the path radiance is a much more symmetric
function of scan angle. Figure 7 shows how path radiance would vary
throughout one day if the aircraft were flying East or West at an
altitude of 1 km on a medium hazy day. The scan is in the North-South
plane. Note the increase and decrease of the radiance at any given scan
angle as the time changes throughout the day. Also note the increase
in path radiance at noon when the sun moves into the scan plane. This
type of information is useful for planning flight lines and times.
A scanner measures the total radiance from the view direction; thus,
the atmospheric effects of path radiance and transmittance are both
present in its signals. Figure 8 simulates the total radiance that
would be received by the scanner from a diffusely reflecting surface
with the reflectance spectrum of green vegetation. A comparison of
Figures 7 and 8 shows that the path radiance is a substantial component
of the total received radiance. Both figures show the large and systematic
changes in signal level that are associated with time of day, that is,
with sun position.
Now let us examine the scan angle dependence more closely. Figure
9 presents a comparison between experimental measurements of sky radiance
and calculated values for three different surface albedos. Note the
excellent agreement between the shapes of the experimental and the
calculated curves of sky radiance. Total radiance calculated with the
model was compared to total radiance measured by a multispectral scanner
for several fields of soybeans. The data are presented in Figure 10.
One immediately notes the difference in angular response. We believe
that much of this difference is due to the surface reflectance character-
istics. The simulation model assumed a perfectly diffuse, or Lambertian,
surface, whereas results presented by Suits [13] and results of other
investigators show that agricultural crops, like soybeans, do have
definite bidirectional reflectance characteristics. The model is being
modified to include such reflectarices.
16