172
M.R. DRINKWATER
deriving kinematic measurements of the Weddell ice cover. In contrast to the Arctic ice
motion studies using ERS SAR (Kwok et al. 1990), the capability to track Antarctic ice
using automated SAR motion tracking had not been demonstrated. The reason was
because in the Arctic, large fractions of multiyear ice provide high contrast targets
which can be successfully tracked in time using radar images. Weddell Sea ice in comparison exhibits large expanses oflow contrast, level first-year ice, and a lack of distinct features with which computer algorithms may successfully track the ice. Here it
is demonstrated that SAR images are effective at resolving features which may be used
throughout the year in tracking both perennial and seasonal Antarctic ice floes.
8.5.4.1
Perennial Ice Drift
Perennial ice in the Antarctic shares some similarities with Arctic multiyear ice, in that
it is relatively thicker than its seasonal counterpart and generally supports a high snow
load (Massom et al. 1997). Old ice may be expected to produce a relatively bright
backscatter signature at C-band during all seasons of the year, as demonstrated by
examples in Sect. 8.4.1. During 1992, a series of SAR image acquisitions were planned
in conjunction with the drift of ISW. ISW was deployed on a perennial ice floe in the
south-western Weddell Sea (Drinkwater and Lytle 1996) and positioned over the continental shelf-break at 71.5°S, 50 0 W (ISW Group 1993).
Figure 12 shows a result from tracking snow-covered perennial ice floes in the vicinity of ISW during the February - March drift phase of this 1992 experiment. Overlapping image pairs acquired along 1- and 3-day-spaced crossing orbits (during the soFig.12. Three-day ice
motion vectors superimposed onto an ERS-l
SAR image for March 15,
1992 (image copyright
ESA 1992). The scale
indicates the distance
(km) of the vector displacements and the
mean u and v components of ice velocity are
given at upper left. For
further details, see text
M.R. DRINKWATER
deriving kinematic measurements of the Weddell ice cover. In contrast to the Arctic ice
motion studies using ERS SAR (Kwok et al. 1990), the capability to track Antarctic ice
using automated SAR motion tracking had not been demonstrated. The reason was
because in the Arctic, large fractions of multiyear ice provide high contrast targets
which can be successfully tracked in time using radar images. Weddell Sea ice in comparison exhibits large expanses oflow contrast, level first-year ice, and a lack of distinct features with which computer algorithms may successfully track the ice. Here it
is demonstrated that SAR images are effective at resolving features which may be used
throughout the year in tracking both perennial and seasonal Antarctic ice floes.
8.5.4.1
Perennial Ice Drift
Perennial ice in the Antarctic shares some similarities with Arctic multiyear ice, in that
it is relatively thicker than its seasonal counterpart and generally supports a high snow
load (Massom et al. 1997). Old ice may be expected to produce a relatively bright
backscatter signature at C-band during all seasons of the year, as demonstrated by
examples in Sect. 8.4.1. During 1992, a series of SAR image acquisitions were planned
in conjunction with the drift of ISW. ISW was deployed on a perennial ice floe in the
south-western Weddell Sea (Drinkwater and Lytle 1996) and positioned over the continental shelf-break at 71.5°S, 50 0 W (ISW Group 1993).
Figure 12 shows a result from tracking snow-covered perennial ice floes in the vicinity of ISW during the February - March drift phase of this 1992 experiment. Overlapping image pairs acquired along 1- and 3-day-spaced crossing orbits (during the soFig.12. Three-day ice
motion vectors superimposed onto an ERS-l
SAR image for March 15,
1992 (image copyright
ESA 1992). The scale
indicates the distance
(km) of the vector displacements and the
mean u and v components of ice velocity are
given at upper left. For
further details, see text