| A sea-ice tracking algorithm has been developed which identifies areas common to satellite image pairs using the principle of maximum cross-correlation. The method provides an automated and objective means of measuring ice motion between successive images. The algorithm is applied to determine ice motion in the North Water polynya located in northern Baffin Bay. The input data are both visible and infrared channels from the Advanced Very High Resolution Radiometer (AVHRR) onboard the National Oceanic and Atmospheric Administration's series of polar orbiting satellites.;A series of 14 AVHRR images, collected between January and April, 1997, were used as input to the algorithm. The resulting vector displacement field is taken to represent the ice flow field. The North Water polynya is a complex environment in which to resolve ice motion. Polynyas are characterized by young ice types and open water. Ice floes in the North Water exhibit large rotations, distortions, and displacements. The algorithm has difficulty resolving rotation and distortion. There are two sources of error when measuring ice motion using successive satellite images: inaccuracies in the geographic location of each pixel in either scene and the uncertainty in identifying common ice features in a sequential image. The ice displacement measurement error for this study is estimated to be 3 km. Shortcomings of the system are described and suggestions for improvement are made.;Atmospheric data from stations surrounding the region were also collected in order to investigate any relationship between geostrophic winds and ice motion. No relationship was found and the differences are thought to be primarily due to local wind effects, internal ice stresses, and surface currents. |
