From ice shelves to icebergs: Classification of calving fronts, iceberg monitoring and drift simulation

Abstract

Antarctica is surrounded by ice shelves and glaciers of different sizes. Satellite imagery shows different feature patterns (e.g. crevasses, rifts) at their surfaces, which control the shape and the size of icebergs that calve from their seaward edges. An edge detection method was used to map and classify the surface features, considering their orientation relative to the calving front. Calved icebergs can automatically be detected and then tracked on their way through the ocean using single and multi-polarized Synthetic Aperture Radar (SAR) images. Temporal gaps between subsequent SAR imagery can be closed by applying a simple wind-driven iceberg drift model.