Binary sea ice lead maps derived from Sentinel-1 SAR images during MOSAiC expedition

• Main Author: Murashkin, Dmitrii
• Format: Dataset
• Language: English
• Published: PANGAEA 2023
• Subjects: Arctic Ocean; Binary Object; Binary Object (File Size); Binary Object (MD5 Hash); CT; DATE/TIME; LATITUDE; leads; LONGITUDE; MOSAiC; MOSAiC20192020; MOSAiC expedition; Multidisciplinary drifting Observatory for the Study of Arctic Climate; Polarstern; PS122/1; PS122/1-track; PS122/2; PS122/2-track; PS122/3; PS122/3-track; PS122/4; PS122/4-track; PS122/5; PS122/5-track; SAR; Sea ice; Sentinel-1; Synthetic Aperture Radar (SAR); Sentinel; 1A/B; Underway cruise track measurements; Arctic; The Cryosphere; The Cryosphere Discussions
• Online Access: https://doi.pangaea.de/10.1594/PANGAEA.962904; https://doi.org/10.1594/PANGAEA.962904

During winter, a significant part of the Arctic Ocean is covered with sea ice. Sea ice works as an insulation layer between the ocean and the atmosphere. The presence of leads is an important feature of the Arctic sea ice cover. Leads are areas with open water or thin ice, which are usually of elongated shape. Leads regulate the heat, gas, and moisture fluxes between the ocean and the atmosphere and are places of increased sea ice production, during periods of freezing conditions. Here binary lead maps derived from Sentinel-1 SAR images covering the MOSAiC expedition, which took place in the Central Arctic in 2019-2020, are presented. The data set contains 1596 classified Sentinel-1 scenes acquired in the extra wide swath (EW) mode with 400 km swath width and 40 meters pixel spacing. The maps are produced with a binary classification algorithm based on a convolutional neural network. Results are provided as geotiff images in the SAR geometry with the native SAR 40 meters pixel size. Georeferencing is done with ground control points corresponding to the ground control points provided with Sentinel-1 images.