Operational algorithm for ice–water classification on dual-polarized RADARSAT-2 images

• Published in: The cryosphere Vol. 11; no. 1; pp. 33 - 46
• Main Authors: Zakhvatkina, Natalia, Korosov, Anton, Muckenhuber, Stefan, Sandven, Stein, Babiker, Mohamed
• Format: Journal Article
• Language: English
• Published: Katlenburg-Lindau Copernicus GmbH 11.01.2017; Copernicus Publications
• Subjects: Accuracy; Algorithms; Automation; Backscatter; Backscattering; Charts; Classification; Comparative analysis; Data; Discrimination; Dual polarization radar; Ice; Ice charts; Ice cover; Image classification; Mathematical models; Methods; Noise reduction; Observations; Polarization; Radar; Radar data; Radar imaging; Radarsat; Remote sensing; SAR (radar); Satellites; Sea ice; Sensors; Support vector machines; Synthetic aperture radar; Thermal noise; Wavelet transforms; Norway; Arctic region; Baltic Sea; Fram Strait
• ISSN: 1994-0424; 1994-0416; 1994-0424; 1994-0416
• DOI: 10.5194/tc-11-33-2017

Synthetic Aperture Radar (SAR) data from RADARSAT-2 (RS2) in dual-polarization mode provide additional information for discriminating sea ice and open water compared to single-polarization data. We have developed an automatic algorithm based on dual-polarized RS2 SAR images to distinguish open water (rough and calm) and sea ice. Several technical issues inherent in RS2 data were solved in the pre-processing stage, including thermal noise reduction in HV polarization and correction of angular backscatter dependency in HH polarization. Texture features were explored and used in addition to supervised image classification based on the support vector machines (SVM) approach. The study was conducted in the ice-covered area between Greenland and Franz Josef Land. The algorithm has been trained using 24 RS2 scenes acquired in winter months in 2011 and 2012, and the results were validated against manually derived ice charts of the Norwegian Meteorological Institute. The algorithm was applied on a total of 2705 RS2 scenes obtained from 2013 to 2015, and the validation results showed that the average classification accuracy was 91 ± 4 %.