Geology Reference
In-Depth Information
the four basic linear polarization components the back-
scattering for any desired combination of linearly trans-
mitted and received polarization combination can be
estimated. This is the essence of the polarization synthe-
sis technique [
Ulaby and Elachi
, 1990].
The fully polarimetric SAR technology was explored
for the first time in the sea ice applications using NASA's
airborne AIRSAR system that was commissioned in 1988
[
van Zyl et al
., 1987]. A notable summary of the potential
applications of data from this system is presented in
Drinkwater et al
. [1992]. The study demonstrated the
capability of the system for identifying early stages on
thin ice formation, potential use of a few polarimteric
parameters in ice classification, and variation of the
parameters with incidence angle for FY and MY ice types.
Drinkwater et al
. [1992, p. 429] noted that the “essence of
polarimtery is not simply in the added benefit of selecta-
ble polarizations, so much as in the additional phase
information embedded in the correlation between differ-
ently polarized waves.” Other studies that use the same
data to map sea ice types include
Rignot and Drinkwater
[1994]. The authors used the data to classify six types of
sea ice based not only on ice age but also on surface con-
ditions: MY ice, compressed FY ice, first‐year rubble and
ice ridges, FY rough ice, FY smooth ice, and thin ice. The
study revealed the potential of polarimteric data applica-
tions for ice classification.
Recently, a few space‐borne polarimetric SAR sys-
tems have become available including the Japanese L‐
band PALSAR on ALOS‐1 satellite (launched in January
2006 and decommissioned in April 2011), the Canadian
C‐band Radarsat‐2 (launched in December 2007 and
still operational), the German TerraSAR‐X (launched in
June 2007), and its successor TanDEM‐X (launched in
early 2010). The last two satellites are still operational.
They were developed under a public‐private partnership
between the German Aerospace Center (DLR) and
European Aeronautic Defence and Space Company
(EADS) Astrium. Polarimetric SAR systems for future
missions include Japan's ALOS‐2 PALSAR (launched in
2014) and Canada's Radarsat Constellation Mission
(RCM) (scheduled for launch in 2018). A brief theoreti-
cal background on radar polarimetry is presented in this
section and discussions of applications follow in sec-
tion 10.1.2. Readers who wish more information on the
theory and applications of SAR polarimetry may refer
to a few topics on the subject such as
Ulaby and Elachi
[1990],
Lee and Pottier
[2009], and
van Zyl
[2011].
Radar Scattering Mechanisms Detected by Polarimetric
SAR Observations
In addition to the direct observations
of the four linear polarization combinations from a
polarimetric sensor, a few derived parameters from the
observations can be used to inform about ice and surface
types as well as the polarimetric scattering mechanisms
triggered by the surface and volume characteristics of the
imaged surface. The derived parameters and their relation
to the scattering mechanisms are discussed later but the
three most commonly mechanisms of polarimetric mech-
anisms are introduced first.
For conventional SAR, the received signal is consid-
ered to be influenced by two scattering mechanisms: sur-
face and the volume. Both mechanisms can be associated
with high or low backscatter, therefore it is difficult to
resolve them from backscatter measurements. On the
other hand, the full polarimetric SAR signal, which
includes the magnitude and phase, reveal three radar
scattering mechanisms: (1) odd‐bounce scattering, (2)
double‐bounce scattering, and (3) random (or volume)
scattering. Figure 7.34 illustrates the three mechanisms.
Odd‐bounce scattering is triggered by a plane surface
with a small amount of roughness, a sphere, or reflector of
trihedral type. Double bounce implies a scattering mecha-
nism that takes place between two surfaces at right angle
with respect to each other. Both single and double bounce
mechanisms retain the polarization of the incident wave
though the latter changes its sign. Random scattering is
associated with multiple scattering from a surface with
structure or heterogeneous volume with many scattering
elements. Random scattering depolarizes the radar signal.
It is possible that more than one scattering mechanism may
emerge from the same ground cell.
ϕ
≅
180°
ϕ
≅
0°
0° <
ϕ
< 180°
Odd-bounce
scattering
Double-bounce
scattering
Random (volume)
scattering
Figure 7.34
Illustration of the three scattering mechanisms identified with radar polarimetric data. The values
of the phase shift
ϕ
between the two co-polarization returns is shown.
