| Résumé | The present iceberg shape characterization ties the above and below water portions of the iceberg in a consistent manner, satisfies hydrostatic considerations, represents measured relationships between waterline length, waterline width, height, draft and mass, and can be used for probabilistic simulations. The approach involves the characterization of three dimensional iceberg shape in terms of the overall average shape and a random component based on the concepts of spatial statistics. The approach has a predictive capability that provides for the generation of a large number of complete iceberg shapes, each with the statistical attributes of measured data. The approach is illustrated through the analysis of two full iceberg profiles collected during the DIGS experiment conducted offshore Labrador in 1985. Many representative iceberg geometries were generated from the statistics of the DIGS icebergs, which were then reoriented and adjusted vertically in the water column to satisfy hydrostatic considerations. Index dimensions were calculated from the generated shapes and their interrelationships were compared with those derived from measured data. The approach yielded realistic iceberg shapes and should be useful for generating iceberg shapes for assessment of risk to Grand Banks installations. |
|---|
