182
Chapter 19
of Boston is not very much in evidence in this image, Cape Code is obscured
by clouds.
Comparing panels C and D with the albedo shown in panel A confirms
the conclusion that, from visual inspection, the albedo derived from perpixel AVHRR and GOES inversions are reasonable in that the results
obtained show a distinct relationship to landscape features. If the BRDF
inversions were unreliable, the albedo image would be expected to be less
clear, especially given the angular extrapolation shown. Similar images of
albedo at other solar zenith angles show that albedo increases when the Sun
approaches the horizon, and that contrast between different surfaces
diminishes. Not unexpectedly, the white-sky albedo is very similar to the
black-sky albedo at 45° solar zenith angle. An image of the ratio between
nadir reflectance and albedo also shows that the transformation of
reflectance into albedo is dependent on the surface features viewed, due to
the associated differences in the BRDF. Using the modified Rahman-PintyVerstraete BRDF model arrives at somewhat higher values for albedo but
otherwise similar results.
Of particular interest is panel B of Figure 1 which shows not the
magnitude of albedo but its functional shape. The image shows the ratio of
black-sky albedo at 45 and that at 0° solar zenith angle, demonstrating
regional differences in the solar zenith angle-dependence of the albedo.
These are also related to landscape properties, due to differences in the
underlying BRDF. Dark areas indicate a weak dependence on solar zenith
angle, bright areas an increase in the albedo with solar zenith. Generally, an
increase is expected from volume scattering as from dense plant canopies
while the shadowing effects of sparser canopies should lead to a smaller
increase or even a slight decrease. This expectation is fully met by the results
shown. The heavily wooded areas display a strong increase (values of the
ratio of up to 1.8) while the urban and urbanized areas with their gaps in the
vegetation cover and buildings show the effects of shadowing (ratios
between 0.8 and 1.2). An interesting case is the Cape Cod region, which is
the only area in the image that is vegetated but sparsely, due to sandy soils. It
promptly shows decreased volume scattering.
Note that while the spatial distribution of the albedo ratio is supported by
the NDVI and snow gap images, it is not a simply redundant retrieval of
information. Details of the image show a wealth of spatial information that
differs from that shown in panels C and D, reflecting the unique spatial
distribution of surface reflectance properties of the visible band used. It is
hoped that future work will allow evaluation of this surface-related
information, making available an additional data dimension for the inference
of land surface biophysical properties.
Chapter 19
of Boston is not very much in evidence in this image, Cape Code is obscured
by clouds.
Comparing panels C and D with the albedo shown in panel A confirms
the conclusion that, from visual inspection, the albedo derived from perpixel AVHRR and GOES inversions are reasonable in that the results
obtained show a distinct relationship to landscape features. If the BRDF
inversions were unreliable, the albedo image would be expected to be less
clear, especially given the angular extrapolation shown. Similar images of
albedo at other solar zenith angles show that albedo increases when the Sun
approaches the horizon, and that contrast between different surfaces
diminishes. Not unexpectedly, the white-sky albedo is very similar to the
black-sky albedo at 45° solar zenith angle. An image of the ratio between
nadir reflectance and albedo also shows that the transformation of
reflectance into albedo is dependent on the surface features viewed, due to
the associated differences in the BRDF. Using the modified Rahman-PintyVerstraete BRDF model arrives at somewhat higher values for albedo but
otherwise similar results.
Of particular interest is panel B of Figure 1 which shows not the
magnitude of albedo but its functional shape. The image shows the ratio of
black-sky albedo at 45 and that at 0° solar zenith angle, demonstrating
regional differences in the solar zenith angle-dependence of the albedo.
These are also related to landscape properties, due to differences in the
underlying BRDF. Dark areas indicate a weak dependence on solar zenith
angle, bright areas an increase in the albedo with solar zenith. Generally, an
increase is expected from volume scattering as from dense plant canopies
while the shadowing effects of sparser canopies should lead to a smaller
increase or even a slight decrease. This expectation is fully met by the results
shown. The heavily wooded areas display a strong increase (values of the
ratio of up to 1.8) while the urban and urbanized areas with their gaps in the
vegetation cover and buildings show the effects of shadowing (ratios
between 0.8 and 1.2). An interesting case is the Cape Cod region, which is
the only area in the image that is vegetated but sparsely, due to sandy soils. It
promptly shows decreased volume scattering.
Note that while the spatial distribution of the albedo ratio is supported by
the NDVI and snow gap images, it is not a simply redundant retrieval of
information. Details of the image show a wealth of spatial information that
differs from that shown in panels C and D, reflecting the unique spatial
distribution of surface reflectance properties of the visible band used. It is
hoped that future work will allow evaluation of this surface-related
information, making available an additional data dimension for the inference
of land surface biophysical properties.