Abstract
Using digital SEASAT synthetic aperture radar (SAR) imagery, high-precision densely sampled maps of ice motion have been derived by tracking ice features to determine the small-scale spatial variability of ice deformation. The digital SAR imagery was processed to remove geometric distortions and located on the Earth to an accuracy of about 100 m utilizing an algorithm based on the spacecraft orbital data and the characteristics of the SAR data collection system, independent of attitude information or ground reference points. Radiometric enhancement of the imagery using a variable linear stretch algorithm was performed to remove a system-related gradient and improve the identification of sea ice features. Using ice features common to an overlapping pair of images, vector plots of ice motion were then produced. Examples of ice motion are shown in the marginal ice zone and in the central ice pack where mean displacements of 15.3 km/day and 5.0 km/day were measured, respectively. Considering errors in Earth location with those in feature identification an overall error of 150-200 m in displacement measurements was estimated. The ice motion vector plots indicate a high degree of spatial deformation, demonstrating the potential value of spaceborne SAR data for production of precision large-scale maps of ice displacement with a spatial resolution of ice deformation on scales much less than 100 km.
