11 The RADARSAT Geophysical Processor System
237
standing and modeling these phenomena requires information of sea ice motion,
thickness, and concentration.
We plan to estimate and monitor five sea ice parameters in the RGPS system: ice
motion, ice age/thickness, date of melt onset/freeze-up, open water fraction, and histograms of backscatter. Below, we describe the scientific relevance of these geophysical
variables. A fundamental RGPS measurement is that of ice motion, obtained by tracking common features in successive SAR images. In the RGPS, a Lagrangian view of the
motion field is produced from a sequence of SAR imagery. From these observations,
the ice age and ice thickness histograms of the thin ice fraction of the ice cover in the
winter are estimated. In the summer, we estimate the open water fraction from
backscatter and area change information. The backscatter histogram of each cell, an
areal parcel on the ice cover, is recorded during each observation. The times of melt
onset in the spring and freeze-up in the late fall are estimated from changes in the
records of the backscatter signature of the ice cover. These fields of geophysical variables will support process studies, model validation and the development of climatologies of sea ice processes. A brief description of the importance of these five parameters and their methods of measurement are provided in the rest of this section. More
detailed discussion of the algorithms can be found in Sect. 11.2.
11.1.1
Ice Motion
Large scale circulation of sea ice determines the advective part of the ice balance and
provides a velocity boundary condition on the ocean surface. Smaller scale processes
involve the detailed motion of individual floes, aggregate of floes, and the formation of
leads. Small-scale ice motion controls the abundance of thin ice and therefore the many
surface processes dependent on thin ice, such as turbulent heat flux to the atmosphere.
Lead formation during periods of divergent motion produce open water and thin-ice
areas that dominate the heat flux into the atmosphere and salt flux into the ocean. In
contrast to this are pressure ridges that form during periods of ice convergence. Small
scale motion and deformation of the ice cover under wind and current stress are measurable in SAR data, as has been demonstrated by numerous studies (Curlander et al.
1985; Fily and Rothrock 1987; Kwok et al.1990). Time-series SAR data provide a smallscale view of sea ice motion obtained by tracking common features in successive observations provided by the high resolution SAR data.
11.1.2
Ice Age/Thickness Distribution
Estimation of the thickness distribution is the primary motivation for tracking the ice
age distribution. Among the many properties of sea ice strongly dependent upon its
thickness are compressive strength, rate of growth, surface temperature, turbulent and
radiative heat exchange with the atmosphere, salt content, and brine flux into the
oceanic mixed layer (Untersteiner 1986). Our present knowledge of the Arctic ice thickness distribution is derived largely from analysis of sonar data from submarine cruises. More recently, moorings with upward looking sonars have also been used to sample the thickness distribution at fixed locations. These and other remote sensing tech-
237
standing and modeling these phenomena requires information of sea ice motion,
thickness, and concentration.
We plan to estimate and monitor five sea ice parameters in the RGPS system: ice
motion, ice age/thickness, date of melt onset/freeze-up, open water fraction, and histograms of backscatter. Below, we describe the scientific relevance of these geophysical
variables. A fundamental RGPS measurement is that of ice motion, obtained by tracking common features in successive SAR images. In the RGPS, a Lagrangian view of the
motion field is produced from a sequence of SAR imagery. From these observations,
the ice age and ice thickness histograms of the thin ice fraction of the ice cover in the
winter are estimated. In the summer, we estimate the open water fraction from
backscatter and area change information. The backscatter histogram of each cell, an
areal parcel on the ice cover, is recorded during each observation. The times of melt
onset in the spring and freeze-up in the late fall are estimated from changes in the
records of the backscatter signature of the ice cover. These fields of geophysical variables will support process studies, model validation and the development of climatologies of sea ice processes. A brief description of the importance of these five parameters and their methods of measurement are provided in the rest of this section. More
detailed discussion of the algorithms can be found in Sect. 11.2.
11.1.1
Ice Motion
Large scale circulation of sea ice determines the advective part of the ice balance and
provides a velocity boundary condition on the ocean surface. Smaller scale processes
involve the detailed motion of individual floes, aggregate of floes, and the formation of
leads. Small-scale ice motion controls the abundance of thin ice and therefore the many
surface processes dependent on thin ice, such as turbulent heat flux to the atmosphere.
Lead formation during periods of divergent motion produce open water and thin-ice
areas that dominate the heat flux into the atmosphere and salt flux into the ocean. In
contrast to this are pressure ridges that form during periods of ice convergence. Small
scale motion and deformation of the ice cover under wind and current stress are measurable in SAR data, as has been demonstrated by numerous studies (Curlander et al.
1985; Fily and Rothrock 1987; Kwok et al.1990). Time-series SAR data provide a smallscale view of sea ice motion obtained by tracking common features in successive observations provided by the high resolution SAR data.
11.1.2
Ice Age/Thickness Distribution
Estimation of the thickness distribution is the primary motivation for tracking the ice
age distribution. Among the many properties of sea ice strongly dependent upon its
thickness are compressive strength, rate of growth, surface temperature, turbulent and
radiative heat exchange with the atmosphere, salt content, and brine flux into the
oceanic mixed layer (Untersteiner 1986). Our present knowledge of the Arctic ice thickness distribution is derived largely from analysis of sonar data from submarine cruises. More recently, moorings with upward looking sonars have also been used to sample the thickness distribution at fixed locations. These and other remote sensing tech-