154
M.R. DRINKWATER
location of Maud Rise. In contrast, after the peak ice extent the ice edge is typically in
a recessional mode, and the MIZ usually comprises brash ice and a mixture of small
wave-broken floes and wave-washed piles of rubble.
In contrast to the situation for nilas and young ice forms found in calm environments, pancakes were found to be almost ubiquitous at the ice margin (along the
Greenwich meridian) during ice edge advance in 1992. The occurrence of vast
expanses of pancakes and/or streamers with densely packed subresolution floes
results in an extremely characteristic MIZ signature (Drinkwater et al. 1993a; Gohin
1995; Early and Long 1997). Typically riding on the surface of waves, small wavewashed, porous pans often have a high salinity (Fig. 3b) and sufficiently high permittivity that there is relatively strong backscatter near nadir. The rate at which crO Ship
falls with incidence angle, however, appears dependent on the size and packing density of the pancakes, the amount of open water between them, and how wet or
deformed their surfaces are. Figure 3b shows the salinity characteristics of a pancake sampled on June 13,1992, and Fig. 3C contrasts the June 13 radar signature with
another site from the following day. The June 13 signature is greater by 10 dB
throughout the incidence angle range, with values exceeding -5 dB at 40 0 incidence
or more, due in part to the denser packing of pans and their rough, deformed, and
often rafted surfaces. In cases of smaller pancakes spaced by open water, as on June
14, croShip falls more rapidly. Gohin (1995) reports an intermediate MIZ signature (in
his Fig. 15) which appears a composite of backscatter curves originating over combinations of open water and dispersed floes, or even low-concentration patches of
pancakes.
Considerable azimuthal anisotropy is also noted in pancake ice margins in the
ERS scatterometer data (Early and Long 1997), and is likely due to the fact that the
pancakes damp out smaller gravity and capillary waves, leaving only long-wavelength swell waves. This results in a dominant wave-propagation direction sensed
by the scatterometer in some circumstances. Future attempts to fit the functional
form of the wind relationship may enable wave propagation direction to be derived.
Nilas
Large expanses of recently formed nilas « 10 cm thick) have rarely been observed
during surface ship borne experiments, largely because Weddell Sea experiments
have focused on the ice margin and swell-influenced regions of the seasonal ice pack.
Consequently, the total fraction of observed ice formed under calm, nondynamic
conditions is small, and the opportunity to sample large areas other than in recently opened leads or polynyas with scatterometers is rare. Some young ice was
observed during WWGS 92, which is reported in Drinkwater and Haas (1994). When
observed, nilas typically had little or no snow cover other than frost flowers. The
typical signature includes large backscatter coefficients at near nadir-angles, with a
high gradient of croShip, falling to values less than -20 dB at high incidence angles. A
large sample of calibrated SAR pixel values shown as a thick black global probability distribution function in Fig. 4 (from all image locations in Fig. 2) testifies that
the likelihood of finding large fractions of new ice and nilas with values below -15
dB is rare in the central Weddell Sea, and extensive nilas areas are generally only
found in locations of persistent divergence such as coastal or ice-shelf polynya systems.
M.R. DRINKWATER
location of Maud Rise. In contrast, after the peak ice extent the ice edge is typically in
a recessional mode, and the MIZ usually comprises brash ice and a mixture of small
wave-broken floes and wave-washed piles of rubble.
In contrast to the situation for nilas and young ice forms found in calm environments, pancakes were found to be almost ubiquitous at the ice margin (along the
Greenwich meridian) during ice edge advance in 1992. The occurrence of vast
expanses of pancakes and/or streamers with densely packed subresolution floes
results in an extremely characteristic MIZ signature (Drinkwater et al. 1993a; Gohin
1995; Early and Long 1997). Typically riding on the surface of waves, small wavewashed, porous pans often have a high salinity (Fig. 3b) and sufficiently high permittivity that there is relatively strong backscatter near nadir. The rate at which crO Ship
falls with incidence angle, however, appears dependent on the size and packing density of the pancakes, the amount of open water between them, and how wet or
deformed their surfaces are. Figure 3b shows the salinity characteristics of a pancake sampled on June 13,1992, and Fig. 3C contrasts the June 13 radar signature with
another site from the following day. The June 13 signature is greater by 10 dB
throughout the incidence angle range, with values exceeding -5 dB at 40 0 incidence
or more, due in part to the denser packing of pans and their rough, deformed, and
often rafted surfaces. In cases of smaller pancakes spaced by open water, as on June
14, croShip falls more rapidly. Gohin (1995) reports an intermediate MIZ signature (in
his Fig. 15) which appears a composite of backscatter curves originating over combinations of open water and dispersed floes, or even low-concentration patches of
pancakes.
Considerable azimuthal anisotropy is also noted in pancake ice margins in the
ERS scatterometer data (Early and Long 1997), and is likely due to the fact that the
pancakes damp out smaller gravity and capillary waves, leaving only long-wavelength swell waves. This results in a dominant wave-propagation direction sensed
by the scatterometer in some circumstances. Future attempts to fit the functional
form of the wind relationship may enable wave propagation direction to be derived.
Nilas
Large expanses of recently formed nilas « 10 cm thick) have rarely been observed
during surface ship borne experiments, largely because Weddell Sea experiments
have focused on the ice margin and swell-influenced regions of the seasonal ice pack.
Consequently, the total fraction of observed ice formed under calm, nondynamic
conditions is small, and the opportunity to sample large areas other than in recently opened leads or polynyas with scatterometers is rare. Some young ice was
observed during WWGS 92, which is reported in Drinkwater and Haas (1994). When
observed, nilas typically had little or no snow cover other than frost flowers. The
typical signature includes large backscatter coefficients at near nadir-angles, with a
high gradient of croShip, falling to values less than -20 dB at high incidence angles. A
large sample of calibrated SAR pixel values shown as a thick black global probability distribution function in Fig. 4 (from all image locations in Fig. 2) testifies that
the likelihood of finding large fractions of new ice and nilas with values below -15
dB is rare in the central Weddell Sea, and extensive nilas areas are generally only
found in locations of persistent divergence such as coastal or ice-shelf polynya systems.