Semi-automated tracking of iceberg B43 using Sentinel-1 SAR images via Google Earth Engine

Abstract. Sentinel-1 C-band synthetic aperture radar (SAR) images can be used to observe the drift of icebergs over Southern Ocean with around 1–3 d temporal resolution and 10–40 m spatial resolution. The Google Earth Engine (GEE) cloud-based platform allows processing a large quantity images, saving time computational resources. In this study, we process data via GEE detect track iceberg B43 during its lifespan 3 years (2017–2020) in Ocean. First, all candidate employ an object-based image segmentation (simple non-iterative clustering – SNIC) traditional backscatter threshold method. Next, automatically choose trace location target by comparing centroid distance histograms (CDHs) detected subsequent days CDH reference iceberg. Using approach, successfully from Amundsen Sea Ross examine changes area, speed, direction. Three periods sudden losses area (i.e., split-offs) coincide low sea ice concentration, warm air temperature, high waves. This implies that these variables may related mechanisms causing split-off Since is generally surrounded compacted ice, correlates part motion wind velocity. Given bulk under water (∼30–60 freeboard ∼150–400 thickness), predominantly driven westward-flowing Antarctic Coastal Current, which dominates circulation region. Considering complexity modeling icebergs, there demand for database better understand behavior their interactions surrounding environments. semi-automated tracking based on storage capacity computing power purpose.