57A037
High-resolution monitoring of sea-ice drift and deformation
with SAR satellite data
Thomas Hollands, Wolfgang Dierking
Corresponding author: Thomas Hollands –
thomas.hollands@awi.de
Most methods monitoring sea-ice drift and deformation
with SAR satellite data generate motion fields at a resolution at about 1 km or
coarser. However, for an expanded understanding of sea-ice processes and their
motion/deformation energy balances it would be an asset to retrieve motion data
at higher spatial and temporal resolution. Most recent approaches are based on
the identification of corresponding patterns in temporal consecutive datasets,
assuming temporal pattern stability/constancy. The position difference between
corresponding patterns is taken as the drift vector, presuming a constant
direction and magnitude of motion over time. Since the algorithm depends on the
identifiability of corresponding patterns, one aspect of this work focuses on
the comparison of miscellaneous SAR products and their performance as to sea
ice. Existing products cover various polarizations and multiple radar frequency
bands (L-, C- and X-band) which differ in their penetration depth and, hence, in
their radar signature. Based on a comparison of algorithm performance the most
suitable data product for the implemented method is identified. Deformation
processes change existing patterns and thereby hurt the ‘constancy
constraint’. To minimize this effect it is necessary to increase the
temporal resolution depending on the existing forces within the relevant region.
One option to improve the temporal resolution is to employ wide swath data,
covering larger regions at coarser spatial resolution. Thus it is important to
understand how characteristics of sea-ice drift change between different spatial
resolutions. In the frame of this work drift patterns derived from a number of
SAR imaging modes at different resolutions are compared with one another. Any
difference between drift parameters obtained for different imaging modes is not
only important for the portability of estimated drift fields between different
resolutions, but as well an important aspect for the comparison of estimated
values with simulated drift patterns from existing sea-ice models. Calculated
motion fields and identified discontinuities (missing pattern correspondence)
are used to delineate deformation patterns and classify them. We will report on
the recent status of the work, present examples of ice velocity fields derived
from different SAR imaging modes and discuss pros and cons of the investigated
imaging modes.
Go Back