Spices – Sea Ice Type Maps From Fram Strait, Barents And Kara Sea

• Main Authors: Babiker, Mohamed, Korosov, Anton, Park, Jeong-Won, Hamre, Torill, Yamakawa, Asuka
• Format: Dataset
• Language: English
• Published: Zenodo 2018
• Subjects: Sea ice, Arctic, Sentinel-1, SAR, synthetic aperture radar, Support Vector Machine; Arctic; Kara Sea; Fram Strait; Sea ice
• Online Access: https://dx.doi.org/10.5281/zenodo.1310968; https://zenodo.org/record/1310968

Sea ice type classification of Sentinel-1 SAR dual polarization EW images. Prior to classification, Sentinel-1 SAR images have been pre-processed to remove noise [1]. The procedure of classification is comprised of the following steps: 1. Haralick texture features are computed, 2. Principal Component Analysis (PCA) is applied, 3. K-means clustering is used to group the data into 15 clusters using factor scores of the PCA as input, 4. The 15 clusters are analysed by an ice expert and classified as open water and different types of sea ice. [1] J.-W. Park, A. A. Korosov, M. Babiker, S. Sandven, J.-S. Won, Efficient Thermal Noise Removal for Sentinel-1 TOPSAR Cross-Polarization Channel, IEEE Transactions on Geoscience and Remote Sensing, 2018; 56, 3, DOI:10.1109/TGRS.2017.2765248.