Item

Abstract

Leads are not permanently open, but are partially frozen and covered by a thin layer of ice up to about 25 cm thick. This thin ice layer is a hotspot for ocean ventilation as well as for the exchange of heat and moisture between the ocean and the atmosphere. Usually, satellite altimetry is used to determine sea level and its changes. In order to monitor the sea level in the polar oceans, methods have been published in recent years that can detect leads by analysing the shape and backscatter properties of altimeter radar echoes (i.e. waveforms). Here we present an extension of an unsupervised waveform classification of Cryosat-2 SAR observations to identify thin ice surfaces and delineate them from ice-free areas as well as from thicker ice. The unsupervised classification approach identifies similar patterns among a subset of randomly collected waveforms and groups them into a specific number of classes without the use of training data. The classification results are visually compared with thin ice thickness estimates from MODIS-observed ice-surface temperatures and Sentinel-1A/B SAR imagery for co-located datasets. In addition, the waveform derived shape and backscatter parameters are analysed with respect to changing thin ice thickness, revealing strong linear dependencies. The analyses can be used to improve altimeter range estimation and thus to allow for a more reliable determination of the sea surface height in the ice-covered oceans as well as a deeper understanding of the Arctic ice cover.