Geography Reference
In-Depth Information
some degree of atmospheric correction. Even for airborne
remote sensing, which involves a shorter stream-sensor
path length than satellite data, the atmosphere exerts a sig-
nificant influence on the water-leaving radiance that must
be accounted for. Also note that the atmosphere affects
the downwelling irradiance incident upon the river.
The critical implication of the preceding discussion of
the image chain is that the radiance measured above a
river channel comes from a number of different sources
and has been influenced by a number of different radiative
transfer processes. These diverse energy pathways can be
summarised by expressing the total, at-sensor upwelling
spectral radiance L T ( λ ) as the sum of four components:
and the volume reflectance of the water column. If a
study's objectives are to map bathymetry or substrate
composition, L B ( λ ) is the radiance component of interest
and the other terms in Equation (3.14) must somehow
be accounted for. Next, L C ( λ ) represents radiance from
the water column itself and encompasses those pho-
tons that entered the water but were scattered into the
upward hemisphere before reaching the bed. L C (
)is
the radiance component most useful for measuring con-
centrations of suspended sediment or other suspended
or dissolved materials. The combination of the first two
terms in Equation (3.14) constitutes the water-leaving
radiance L W (
λ
), which includes all of
the photons that have interacted with the river channel in
a potentially informative manner. The surface-reflected
radiance L S (
λ
)
=
L B (
λ
)
+
L C (
λ
L T (
λ
)
=
L B (
λ
)
+
L C (
λ
)
+
L S (
λ
)
+
L P (
λ
)
(3.14)
) is comprised of photons that never entered
the water and are thus unaffected by depth, bottom type,
or water column optical properties. Similarly, path radi-
ance L P ( λ ) scattered into the sensor's field of view by
the atmosphere contributes to L T (
λ
These components are illustrated in Figure 3.6. The
bottom-reflected radiance L B (
) consists of photons that
have interacted with the streambed. L B (
λ
) thus depends
not only on depth, due to the exponential attenuation
of light with distance traveled through the water col-
umn, but also on bottom type, which determines the
bottom contrast between the reflectance of the streambed
λ
) but is essentially
independent of the river attributes of interest. This parti-
tioning of the radiance signal into component parts, some
of which provide useful river information and some of
which do not, is an important concept to keep in mind in
any application of remote sensing to rivers.
λ
3.3 Optical characteristics
of river channels
L p ( λ )
With this overview of the image chain comprising remote
sensing of rivers, we now shift our attention to the spectral
characteristics of some important features of the resulting
images. Each of the radiative transfer processes and sur-
face interactions described above depends, to some extent,
on wavelength. Radiance measurements in a number of
spectral bands thus provide information that can be used
to identify and remove extraneous contributions to the
at-sensor radiance signal and gain additional leverage
for inferring channel attributes from remotely sensed
data. Panchromatic images are readily available (e.g., his-
torical aerial photography) and often have high spatial
resolution (e.g., commercial satellites featuring a high-
resolution panchromatic band and several multispectral
bands with somewhat larger pixels). These gray-scale data
can be used to map channel planform, lateral migration,
and, to a lesser extent, variations in water depth (e.g.
Winterbottom and Gilvear, 1997). Incorporating spec-
tral information, however, allows channel morphology
to be measured with greater confidence and opens up
L p ( λ )
L B ( λ )
L s ( λ )
L c (
)
λ
AIR
WATER
Figure 3.6 Conceptual diagram of the components of the total
at-sensor radiance signal L T (
). Radiance reaching the sensor
originates from (1) reflectance from the streambed, L B (
λ
λ
); (2)
λ
volume reflectance within the water column, L C (
); (3) surface
λ
reflectance from the air-water interface L S (
); and (4)
scattering within the atmosphere, L P (
). Adapted from Bukata,
R.P. et al. (1995) Optical Properties and Remote Sensing of Inland
and Coastal Waters . Boca Raton, FL: CRC Press, p. 54.
λ
Search WWH ::




Custom Search