Geology Reference
In-Depth Information
slightly warmer than MY ice, whereas airborne and sur-
face radiometer measurements show the opposite.
Most of the ice concentration retrieval algorithms from
passive microwave observations require tie points to solve
the equations that link the observations to the concen-
tration. These are “typical” values of radiometric obser-
vations or derived parameters that represent different ice
types and open water. Examples of the derived parame-
ters include polarization and gradient ratios. Tie points
depend on the season and the region of ice formation.
Moreover, within a given season and region the radiation
from a given ice type may vary significantly due to varia-
tions in ice surface conditions in response to weather
events. This is particularly notable in the case of thin ice.
One of the early attempts to develop a set of tie points of
passive microwave data for OW, FY, and MY ice was pre-
sented in
Cavalieri et al
. [1991]. The purpose was to use
those tie points in the NASA Team (NT) ice concentra-
tion algorithm (section 10.2.2). An extensive data set
acquired from SSM/I and aircraft underflights was con-
ducted in the Arctic in March 1988. The aircraft data
were used to verify the ice concentration output derived
from SSM/I data. The tie points were adjusted to yield
estimates close to the results from the airborne sensor.
They are presented in Table 8.6 for the brightness tem-
perature from the three channels used in the NT algo-
rithms [
T
b
(19
h
),
T
b
(19
v
), and
T
b
(37
v
)]. The FY and MY
ice data in the table agree with the data in Figure 8.17.
The tie points of OW also agree with other published
data such as those obtained for saline water in an out-
door tank [
Shokr and Kaleschke
, 2012]. Except for the
37 GHz vertical channel, values of
T
b
from sea ice in the
Antarctic, are less than the values from the Arctic. This
difference can be attributed to the colder temperature in
the Antarctic, but it can also be explained in terms of the
difference in emissivity. The snow cover on sea ice in the
Southern Ocean is thicker with larger grain size, which
causes more scattering and therefore less emissivity.
Zwally et al
. [1983] presented maps of 4 year average of
monthly
T
b
values over the Antarctic sea ice between
1973 and 1976 from the 19.53 GHz channel of ESMR.
The
T
b
of OW from these maps is rather high (around
138.3 K), compared to the values included in Table 8.6.
The ability of the polarization and gradient ratios to
discriminate between OW and sea ice was demonstrated
in
Rubinstein et al
. [1994] from using SSM/I observations
acquired over the ice edge in the Labrador Sea along the
Table 8.6
Brightness temperature tie points for OW, FY ice,
and MY ice in the Arctic and Antarctic.
T
b
(K)
Surface Type
Channel
Arctic
Antarctic
OW
T
b
(19
v
)
177.1
176.6
T
b
(19
h
)
100.8
100.3
T
b
(37
v
)
201.7
200.5
FY ice
T
b
(19
v
)
258.2
249.8
T
b
(19
h
)
242.8
237.8
T
b
(37
v
)
252.8
243.3
MY ice
T
b
(19
v
)
223.2
221.6
T
b
(19
h
)
203.9
193.7
T
b
(37
v
)
186.3
190.3
Source
:
Cavalieri et al
., 1991, Table 2, with permission from AGU.
0.3
Gradient ratio GR
37v19v
Polarization ratio PR
37
0.25
0.2
Open water
Sea ice
0.15
0.1
0.05
0
Ice edge location
-0.0
-55
-54
-53
-52
-51
-50
-49
-48
-47
-46
-45
Longitude (W)
Figure 8.18
Polarization and gradient ratios calculated from SSM/I data over the Labrador Sea area for 16
February 1992. Data were obtained along constant latitude 52°N, with the ice edge located around longitude
50°W [
Rubinstein et al
., 1994, Fig. 5.27, with permission from Wiley].
