Identification of Extreme Ice Features in the Canadian Arctic

Abstract

Research on automatic detecting, tracking and characterizing extreme ice features in the Arctic is based on analyzing and processing satellite synthetic aperture radar (SAR) and optical images. Algorithms to identify ridges from very high-resolution optical data have an accuracy of 86.4% when compared to manual extraction and ridge height has been estimated from shadow. SAR signatures of various ice features have been analyzed and the results indicate that it is possible to identify rubble fields from other ice types.