Annotated dataset of beluga whales detected in Very High Resolution (VHR) satellite imagery of the southern Kara Sea (Russia) in July 2016. ...

• Main Authors: Panigada, Viola, Cubaynes, Hannah C, Fretwell, Peter T
• Format: Dataset
• Language: English
• Published: NERC EDS UK Polar Data Centre 2024
• Subjects: "EARTH SCIENCE","BIOSPHERE","ECOSYSTEMS","MARINE ECOSYSTEMS","COASTAL"; "EARTH SCIENCE","BIOSPHERE","ECOLOGICAL DYNAMICS","SPECIES/POPULATION INTERACTIONS","POPULATION DYNAMICS"; "EARTH SCIENCE","BIOSPHERE","ECOLOGICAL DYNAMICS","SPECIES/POPULATION INTERACTIONS","USE/FEEDING HABITATS"; "EARTH SCIENCE","BIOLOGICAL CLASSIFICATION","ANIMALS/VERTEBRATES","MARINE MAMMALS"; Delphinapterus leucas; Beluga; Foraging ground; Remote Sensing; Satellite Imagery; Very High Resolution; Antarc*; Antarctic; Baydaratskaya bay; Beluga*; British Antarctic Survey; Kara Sea
• Online Access: https://dx.doi.org/10.5285/ae34b490-4838-4740-a78e-924d23dc5540; https://data.bas.ac.uk/full-record.php?id=GB/NERC/BAS/PDC/01829

This annotated dataset comprises locational data of beluga whales along the eastern shore of the Yugor Penisula and in the inner part of Baydaratskaya Bay in the southern Kara Sea (Russia) as detected from Very High Resolution (VHR) satellite imagery on July 4th and 9th, 2016. Images were manually scanned and whales detected through the use of grids. Additional metadata includes information on image type and model, and whale distinctive characteristics (e.g., fluke or blow). This work supports the 'training' of machine learning algorithms for automatic detection of whales from satellite imagery. This study was possible thanks to imagery support from MAXAR Technologies/Digital Globe Foundation for the VHR images and core funding from British Antarctic Survey, Natural Environment Research Council, as part of the 'Wildlife from Space' project. ... : Beluga whales were manually detected by scanning VHR satellite imagery obtained from the Maxar satellites WorldView-2 and -3. Single points were placed on the whales, and the attribute table was filled out to contain information on the certainty level (i.e. "definite", "probable", "possible") derived from the classification score assessed based on various criteria (i.e. body length, body width, body shape, body colour, fluke print, blow, contour, wake, after-breach, defecation, other disturbance, fluke, flipper, head callosities and mud trail), following Cubaynes et al. 2023. For additional details on attribute table definitions, please refer to the README affiliated with the shapefile data. For each annotated whale, there is information about the imagery analysed: the location, latitude and longitude (in decimal degrees and recorded using the same geographic coordinate system and projection as the satellite imagery), imagery ID, imagery date, type of satellite, spatial resolution, number of multispectral ...