Geology Reference
In-Depth Information
No dominant scattering;
A
is considered for
this range of H
H
≅
0
A
≅
0
A
≅
1
H
≅
1
λ
1
>>
λ
2
and
λ
3
;
λ
2
≅
λ
3
λ
2
>>
λ
3
λ
1
≅
λ
2
≅
λ
3
;
λ
2
and
λ
3
≅
0
One secondary
mechanism
effective
Tw o secondary
mechanisms
effective
Figure 7.35
Illustration of the physical meanings of the entropy and anisotropy parameters in terms of the number
of the effective mechanisms.
Entropy is an indicator of the randomness of the scatter-
ing that comprises the received signal (caused by statistical
disorder of the scattering elements within the volume).
Small values signify the presence of a dominant scattering
mechanism. When
H
= 0, only one mechanism is dominant
(deterministic scattering). Large values, on the other hand,
signify the coexistence of more than one scattering mecha-
nism with approximately same weight. When
H
= 1, all
three mechanisms contribute equally to the observed scat-
tering (fully random scattering). Between these two ends,
the entropy does not provide information about the num-
ber of scattering mechanisms. In addition to the dominant
scattering mechanism from a resolution cell, secondary
scattering mechanisms may take place. In this case the ani-
sotropy provides the required information about the rela-
tive weight of each mechanism
λ
2
and
λ
3
. Low values mean
that the two secondary mechanisms are equally effective,
and high values mean that only one of them is effective.
Figure 7.35 illustrates the meanings of
H
and
A
(which
are complementary) in relation to the three eigenvalues of
the coherence matrix. The last parameter, alpha angle, is
related to the type of scattering mechanism and it varies
between 0° and 90°. When
α
=0°, it means that scattering
is triggered by a smooth surface, sphere or trihedral (odd
bounce). When
α
=90° the scattering is generated through
a double‐bounce mechanism (e.g., scattered high‐rise
buildings in a flat area). For
α
=45° the scattering is gener-
ated from the volume (random dipole scattering).
It would be appropriate to conclude this section with
a note regarding the progress on the work of using
the compact polarization versus fully polarimetric data.
The CP SAR system is currently operating onboard the
Indian satellite RISAT‐1 and the Japanese ALOS-2 sat-
ellite. It will be available from a small satellite called
μ
SAT, being developed also in Japan. In 2018 it will also
be available from the Canadian Radarsat Constellation
Mission (RCM). The RCM will transmit a right‐circular
polarization and receives two mutually coherent orthogo-
nal liner polarizations (RH and RV). This compact polar-
ization will be the primary imagng mode from RCM.
Although the CP mode generates half of the space of
information compared to the FP SAR, its advantage lies
in its operation with only one antenna (i.e., it comsumes
half the power of the FP SAR) and with the hybrid (cir-
cular) polarization it can generate measurements in the
wide swath ScanSAR mode (up to a 500 km swath).
Reseach on comparing the CP SAR data (using simu-
lated data) and the Radarsat fully polarimetric data is
progressing in Canada in preparation for the anticipated
data from the RCM.
Touzi
[2009] provided a critical comparison between
fully polarimetric (with linear polarization) from
Radarsat‐2 and simulated hybrid compact polarimetric
SAR. The study concluded that the FP should remain the
chioce if the high quality of cross‐polarization return is
the priority. That is because FP SAR proides cross‐polar-
ization measurements of much higher signal‐to‐noise
ratio than the CP SAR. However, the hybrid CP provides
meaningful phase information that might be worthy
exploiting.
Dabboor and Geldsetzer
[2014] calculated 23
parameters from simulated CP SAR data and evaluated
their potential for sea ice of type discrimination. The
study identified the three parameters for the optimal clas-
sificaation of sea ice and open water.
7.7. radiative Processes in relevant media
Remote sensing observations of sea ice are influenced by
radiative processes from three relevant media: atmosphere,
seawater, and snow. In order to retrieve ice parameters
from the observations, the influences of these media should
be accounted for. This section addresses a few key issues of
EM wave interactions with these media. The information
is rather basic and descriptive to serve the purpose of the
readers who seek brief knowledge about the impact of the
three media on the remote sensing observations of sea ice.
However, the text includes references to sources with more
in‐depth information on the subjects.
7.7.1. Atmospheric Influences
In addition to the radiance that originates from the sur-
face, the radiation received at the TOA has contributions
from the atmosphere. The gaseous and particulate matter
