Seals from space: automated Antarctic ecosystem monitoring via high-resolution satellite imagery

The Antarctic is constantly evolving as the ecosystem recovers from past exploitation (e.g. whaling, seal harvesting), adapts to climate change, and responds to current anthropogenic impacts including fishing (krill and Patagonian toothfish), shipping and tourism. Due to its vastness, relatively little is known about the ecology of the region and its species, and how best to mitigate and control anthropogenic impacts in this region. Traditional field methods are costly and limited in geographical extent due to areas of interest being difficult to access by ships. Remote sensing provides a low-cost, non-invasive method that can be used for ecological monitoring.

The overall aim of the challenge was to create an automated system for classifying sea ice and mapping seals which can then be used to transform the satellite raster images into images with vectorised features of ice and seals. Using these two outputs (seal counts and ice classification/environmental features) we can explore ecological questions such as what habitat features do seal prefer and how is the habitat changing over time. This report presents the outputs of a week-long collaboration between the Alan Turing Institute and the British Antarctic Survey (BAS), to scope an automated system to classify sea ice, count seals, and explore the environmental factors influencing seal density.

Citation information

Data Study Group team. (2020, February 17). Data Study Group Final Report: British Antarctic Survey. Zenodo. http://doi.org/10.5281/zenodo.3670637

Additional information

Aditya Acharya, The Alan Turing Institute
Anna Laws, University of Exeter
Christopher Tegho, Calipsa
Eleanor Miller, University of Cambridge
Fred Shone, Arup London
Laura Marcela Guzman Rincon, University of Warwick
Laurie Baker, University of Glasgow
Lucas Deecke, University of Edinburgh
Prem Gill, British Antarctic Survey
Tim Hurst, University of Edinburgh and Heriot Watt University
Timothy Pollington, University of Warwick
Victor Sanchez-Silva, University of Warwick