Sea ice classification using wide-swath SAR data considering incidence angle depenency of backscatter intensity and surface texture

In this study we develop a novel sea ice classification scheme based on remote sensing Synthetic-aperture Radar (SAR) data, and use it to classify sea ice types over the spatial and temporal range of the Norwegian Young sea ICE cruise (N-ICE2015). Ice type classification will be conducted on wide-swath SAR datasets including RADARSAT-2 and Sentinel-1 data. We use a classification scheme that takes into account different rates of decrease in backscatter intensity with incidence angle variation for different classes. In addition, it examines texture features of different sea ice types, and also variations of surface texture with changing incidence angles, and incorporates this relationship into the classification process. Sea ice classifications using high-resolution SAR images collected over the same period and also field data retrieved from the N-ICE2015 expedition will be used for ground truthing. Earlier N-ICE2015 studies with high resolution SAR and deformation suggest high lead and pressure ridge formation. We will use our lower-resolution results to explore potential increase in the fraction of deformed and lead ice from January to June 2015 in the region north of Svalbard.