Deep learning enables satellite-based monitoring of large populations of terrestrial mammals across heterogeneous landscape

Zijing Wu; Ce Zhang; Xiaowei Gu; Isla Duporge; Lacey F. Hughey; Jared A. Stabach; Andrew K. Skidmore; J. Grant C. Hopcraft; Stephen J. Lee; Peter M. Atkinson; Douglas J. McCauley; Richard Lamprey; Shadrack Ngene; Tiejun Wang

doi:10.1038/s41467-023-38901-y

Deep learning enables satellite-based monitoring of large populations of terrestrial mammals across heterogeneous landscape

Zijing Wu, Ce Zhang, Xiaowei Gu, Isla Duporge, Lacey F. Hughey, Jared A. Stabach, Andrew K. Skidmore, J. Grant C. Hopcraft, Stephen J. Lee, Peter M. Atkinson, Douglas J. McCauley, Richard Lamprey, Shadrack Ngene, Tiejun Wang^*

^*Corresponding author for this work

School of Natural Sciences

Research output: Contribution to journal › Article › peer-review

41 Citations (Scopus)

25 Downloads (Pure)

Abstract

New satellite remote sensing and machine learning techniques offer untapped possibilities to monitor global biodiversity with unprecedented speed and precision. These efficiencies promise to reveal novel ecological insights at spatial scales which are germane to the management of populations and entire ecosystems. Here, we present a robust transferable deep learning pipeline to automatically locate and count large herds of migratory ungulates (wildebeest and zebra) in the Serengeti-Mara ecosystem using fine-resolution (38-50 cm) satellite imagery. The results achieve accurate detection of nearly 500,000 individuals across thousands of square kilometers and multiple habitat types, with an overall F1-score of 84.75% (Precision: 87.85%, Recall: 81.86%). This research demonstrates the capability of satellite remote sensing and machine learning techniques to automatically and accurately count very large populations of terrestrial mammals across a highly heterogeneous landscape. We also discuss the potential for satellite-derived species detections to advance basic understanding of animal behavior and ecology.

Original language	English
Article number	3072
Pages (from-to)	1-15
Number of pages	15
Journal	Nature Communications
Volume	14
Issue number	1
DOIs	https://doi.org/10.1038/s41467-023-38901-y
Publication status	Published - 27 May 2023

Bibliographical note

Copyright the Author(s) 2023. Version archived for private and non-commercial use with the permission of the author/s and according to publisher conditions. For further rights please contact the publisher.

Access to Document

10.1038/s41467-023-38901-y

Publisher version (open access)Final published version, 9.33 MBLicence: CC BY

Cite this

Wu, Z., Zhang, C., Gu, X., Duporge, I., Hughey, L. F., Stabach, J. A., Skidmore, A. K., Hopcraft, J. G. C., Lee, S. J., Atkinson, P. M., McCauley, D. J., Lamprey, R., Ngene, S., & Wang, T. (2023). Deep learning enables satellite-based monitoring of large populations of terrestrial mammals across heterogeneous landscape. Nature Communications, 14(1), 1-15. Article 3072. https://doi.org/10.1038/s41467-023-38901-y

@article{bb065bd5a73b4b88b07251acd882cc8b,

title = "Deep learning enables satellite-based monitoring of large populations of terrestrial mammals across heterogeneous landscape",

abstract = "New satellite remote sensing and machine learning techniques offer untapped possibilities to monitor global biodiversity with unprecedented speed and precision. These efficiencies promise to reveal novel ecological insights at spatial scales which are germane to the management of populations and entire ecosystems. Here, we present a robust transferable deep learning pipeline to automatically locate and count large herds of migratory ungulates (wildebeest and zebra) in the Serengeti-Mara ecosystem using fine-resolution (38-50 cm) satellite imagery. The results achieve accurate detection of nearly 500,000 individuals across thousands of square kilometers and multiple habitat types, with an overall F1-score of 84.75% (Precision: 87.85%, Recall: 81.86%). This research demonstrates the capability of satellite remote sensing and machine learning techniques to automatically and accurately count very large populations of terrestrial mammals across a highly heterogeneous landscape. We also discuss the potential for satellite-derived species detections to advance basic understanding of animal behavior and ecology.",

author = "Zijing Wu and Ce Zhang and Xiaowei Gu and Isla Duporge and Hughey, {Lacey F.} and Stabach, {Jared A.} and Skidmore, {Andrew K.} and Hopcraft, {J. Grant C.} and Lee, {Stephen J.} and Atkinson, {Peter M.} and McCauley, {Douglas J.} and Richard Lamprey and Shadrack Ngene and Tiejun Wang",

note = "Copyright the Author(s) 2023. Version archived for private and non-commercial use with the permission of the author/s and according to publisher conditions. For further rights please contact the publisher. ",

year = "2023",

month = may,

day = "27",

doi = "10.1038/s41467-023-38901-y",

language = "English",

volume = "14",

pages = "1--15",

journal = "Nature Communications",

issn = "2041-1723",

publisher = "Springer, Springer Nature",

number = "1",

}

Wu, Z, Zhang, C, Gu, X, Duporge, I, Hughey, LF, Stabach, JA, Skidmore, AK, Hopcraft, JGC, Lee, SJ, Atkinson, PM, McCauley, DJ, Lamprey, R, Ngene, S & Wang, T 2023, 'Deep learning enables satellite-based monitoring of large populations of terrestrial mammals across heterogeneous landscape', Nature Communications, vol. 14, no. 1, 3072, pp. 1-15. https://doi.org/10.1038/s41467-023-38901-y

TY - JOUR

T1 - Deep learning enables satellite-based monitoring of large populations of terrestrial mammals across heterogeneous landscape

AU - Wu, Zijing

AU - Zhang, Ce

AU - Gu, Xiaowei

AU - Duporge, Isla

AU - Hughey, Lacey F.

AU - Stabach, Jared A.

AU - Skidmore, Andrew K.

AU - Hopcraft, J. Grant C.

AU - Lee, Stephen J.

AU - Atkinson, Peter M.

AU - McCauley, Douglas J.

AU - Lamprey, Richard

AU - Ngene, Shadrack

AU - Wang, Tiejun

N1 - Copyright the Author(s) 2023. Version archived for private and non-commercial use with the permission of the author/s and according to publisher conditions. For further rights please contact the publisher.

PY - 2023/5/27

Y1 - 2023/5/27

N2 - New satellite remote sensing and machine learning techniques offer untapped possibilities to monitor global biodiversity with unprecedented speed and precision. These efficiencies promise to reveal novel ecological insights at spatial scales which are germane to the management of populations and entire ecosystems. Here, we present a robust transferable deep learning pipeline to automatically locate and count large herds of migratory ungulates (wildebeest and zebra) in the Serengeti-Mara ecosystem using fine-resolution (38-50 cm) satellite imagery. The results achieve accurate detection of nearly 500,000 individuals across thousands of square kilometers and multiple habitat types, with an overall F1-score of 84.75% (Precision: 87.85%, Recall: 81.86%). This research demonstrates the capability of satellite remote sensing and machine learning techniques to automatically and accurately count very large populations of terrestrial mammals across a highly heterogeneous landscape. We also discuss the potential for satellite-derived species detections to advance basic understanding of animal behavior and ecology.

AB - New satellite remote sensing and machine learning techniques offer untapped possibilities to monitor global biodiversity with unprecedented speed and precision. These efficiencies promise to reveal novel ecological insights at spatial scales which are germane to the management of populations and entire ecosystems. Here, we present a robust transferable deep learning pipeline to automatically locate and count large herds of migratory ungulates (wildebeest and zebra) in the Serengeti-Mara ecosystem using fine-resolution (38-50 cm) satellite imagery. The results achieve accurate detection of nearly 500,000 individuals across thousands of square kilometers and multiple habitat types, with an overall F1-score of 84.75% (Precision: 87.85%, Recall: 81.86%). This research demonstrates the capability of satellite remote sensing and machine learning techniques to automatically and accurately count very large populations of terrestrial mammals across a highly heterogeneous landscape. We also discuss the potential for satellite-derived species detections to advance basic understanding of animal behavior and ecology.

UR - http://www.scopus.com/inward/record.url?scp=85160200172&partnerID=8YFLogxK

U2 - 10.1038/s41467-023-38901-y

DO - 10.1038/s41467-023-38901-y

M3 - Article

C2 - 37244940

AN - SCOPUS:85160200172

SN - 2041-1723

VL - 14

SP - 1

EP - 15

JO - Nature Communications

JF - Nature Communications

IS - 1

M1 - 3072

ER -

Deep learning enables satellite-based monitoring of large populations of terrestrial mammals across heterogeneous landscape

Abstract

Bibliographical note

Access to Document

Other files and links

Fingerprint

Cite this