Geology Reference
In-Depth Information
Rayleigh scattering
Gas molecules
(size <<
λ
)
8
L
S
L
C
Diamete
r
Smoke, dust, etc.
(size ~
λ
)
Mie scattering
Lp
2
1
5
L
T
Nonselective
scattering
Water vapor
Sand particles
(size >>
λ
)
Water
vapor
6
3
θ
r
4
θ
0
7
Figure 7.37
Three types of scattering in the atmosphere,
caused by different sizes of the scattering elements.
Neighboring pixel
IFOV pixel
Figure 7.38
Interaction of solar radiation with atmosphere and
surface. The components of radiation that reaches the satellite
point of observation are shown. Number and symbols are indi-
cators of radiative processes and components of the received
signal as explained in the text.
than near‐infrared light (0.8
μ
m). This preferential scatter-
ing is responsible for blue sky. Rayleigh scattering com-
prises part of the atmospheric emitted radiation.
In Mie scattering, the size of the scattering elements is
comparable to the wavelength of the radiation. For optical
radiation, Mie scattering is usually caused by water vapor,
dust, and other aerosol particles ranging from a few
tenths of a micrometer to a few micrometers in diameter.
Compared to Rayleigh scattering, Mie scattering is greater
in intensity and tends to affect longer wavelengths. While
Rayleigh scattering is nearly isotropic, Mie scattering has a
dominant forward direction. This property makes Mie
scattering responsible for the beautiful sunrise and sunset
colors of the sky when the atmosphere holds a large amount
of smoke and dust. These elements scatter more of the
shorter wavelength violet and blue color in different direc-
tions (Rayleigh scattering) and allow only the longer wave-
length orange and red colors to reach our eyes through the
forward Mie scattering. Lastly, the nonselective scattering
occurs when there are particles in the atmosphere several
times the diameter of the radiation being transmitted. In
this type of scattering all wavelengths of light are scattered,
thus the name nonselective. The water drops in clouds sat-
isfy this condition and that is why clouds appear white. In
general, atmospheric scattering is minimal for the relatively
long wavelength of thermal infrared radiation.
The radiative processes that contribute to the observed
radiation in the optical region are summarized in
Figure 7.38. The figure shows the three components of
radiation that are received by an optical sensor (
L
S
): the
terrestrial reflection/radiation (
L
T
), the cloud reflection
(
L
C
), and the path radiation (
L
P
). For surface applica-
tions of remote sensing the terrestrial radiation is the
desired contribution to be retrieved. The solar radiation
received at the IFOV encompasses direct solar radiation
(1), indirect radiation through scattering by the atmos-
phere (3), and the “leakage” contribution from the neigh-
boring areas (7), which is called adjacent effect. Some
solar and scattered radiation that reaches neighboring
pixels (4) is scattered back to the atmosphere (6). This,
along with the scattered and emitted radiation by atmos-
pheric elements (5) constitutes the path radiance (
L
P
).
Clouds receive Sun irradiance and reflect approximately
19% of it, forming the
L
C
component. In general, in the
optical and infrared regions reflection and emission occur
from cloud top only and therefore become weakly related
to cloud contents. Temperature of cloud top can be
inferred from TIR observations.
Based on the above qualitative description, the radi-
ance received by an optical sensor can be formulated as
the summation of the three components:
LH LL
S
C
(7.100)
P
where
H
is the total downwelling radiance,
α
is the surface
reflectance, and
τ
is the atmospheric transmittance. The
first term in the above equation represents
L
T
. However, this
representation does not include the leakage contribution
from neighboring pixels, but this effect is usually small and
can be neglected, yet it is better to be accounted for.
In order for the optical observations by a satellite sen-
sor to represent the intrinsic directional reflectance
characteristics of the surface, the observations must be
corrected to remove the influences of the path radiation
(atmospheric influences) and the cloud reflection. Several
algorithms have been developed to perform the atmos-
pheric correction. In an original study,
Lindsay and
Rothrock
[1994] estimated the seasonal cycle of clear‐sky
hemispherically integrated surface albedo of sea ice in the
Arctic from measurements made with the AVHRR
onboard the NOAA‐10 and NOAA‐11 satellites. Details
of the method include calculations of the TOA reflectance
of a Lambertian surface, an account for the nonisotropic
reflectance of the ice and atmosphere, and a correction for
