EGU24-10882, updated on 08 Mar 2024
EGU General Assembly 2024
© Author(s) 2024. This work is distributed under
the Creative Commons Attribution 4.0 License.

Satellite-based monitoring of the world’s largest terrestrial mammal migration using deep learning 

Zijing Wu1, Tiejun Wang2, Ce Zhang3,4, Isla Duporge5,6,7, Xiaowei Gu8, Lacey Hughey9, Jared Stabach9, Andrew Skidmore2,10, Grant Hopcraft11, Peter Atkinson12,13, Douglas McCauley14, Richard Lamprey2, Shadrack Ngene15, and Peng Gong1
Zijing Wu et al.
  • 1Department of Geography, The University of Hong Kong, Hong Kong, China (
  • 2Department of Natural Resources, Faculty of Geo-Information Science and Earth Observation, University of Twente, Enschede, The Netherlands
  • 3School of Geographical Sciences, University of Bristol, Bristol, UK
  • 4UK Centre for Ecology & Hydrology, Lancaster, UK
  • 5Department of Ecology and Evolutionary Biology, Princeton University, Princeton, NJ, USA
  • 6The National Academies of Sciences, Washington, D.C., USA
  • 7U.S. Army Research Laboratory, Army Research Office, Durham, NC, USA
  • 8School of Computing, University of Kent, Canterbury, UK
  • 9Conservation Ecology Center, Smithsonian National Zoo and Conservation Biology Institute, Front Royal, VA, USA
  • 10School of Natural Sciences, Macquarie University, Sydney, NSW, Australia
  • 11Institute of Biodiversity, Animal Health, and Comparative Medicine, University of Glasgow, Glasgow, UK
  • 12Geography and Environmental Science, University of Southampton, Southampton, UK
  • 13Lancaster Environment Center, Lancaster University, Lancaster, UK
  • 14Department of Ecology, Evolution and Marine Biology, University of California, Santa Barbara, CA, USA
  • 15Wildlife Research and Training Institute, Naivasha, Kenya

Accurate, reliable, and up-to-date information on wildlife populations is crucial for biodiversity conservation in the face of unprecedented biodiversity loss worldwide. However, monitoring wildlife populations at large scales remains challenging. Advances in satellite remote sensing, particularly very-high-resolution satellite data, offer new opportunities for monitoring wildlife from space, and new machine learning techniques present great potential for detecting wildlife with remarkable speed and accuracy. Here, we introduce a deep learning pipeline for automatically detecting and counting large migratory ungulate herds (wildebeest and zebra) at the individual level in the Serengeti-Mara ecosystem from submeter-resolution satellite imagery. We apply the pipeline to implement the first-ever population census of large-size ungulates in the Serengeti-Mara ecosystem through a satellite survey and generate the total count of the whole population. The model shows robust performance across diverse landscapes with an overall F1-score of 84.75% (Precision: 87.85%, Recall: 81.86%) on an independent test dataset containing 11,594 animals and achieves good transferability spatially and temporally. This research showcases the capability of satellite remote sensing and deep learning techniques to accurately locate and count very large populations of terrestrial mammals in open landscapes. It provides a new perspective on monitoring wildlife populations and animal migration, which will facilitate the understanding of animal behavior and ecology as well as improve the conservation of the whole ecosystem in the face of rapid environmental changes.

How to cite: Wu, Z., Wang, T., Zhang, C., Duporge, I., Gu, X., Hughey, L., Stabach, J., Skidmore, A., Hopcraft, G., Atkinson, P., McCauley, D., Lamprey, R., Ngene, S., and Gong, P.: Satellite-based monitoring of the world’s largest terrestrial mammal migration using deep learning , EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-10882,, 2024.