Conservation of polar bears during their habitat phase is challenged by the limitations of traditional forward-looking infrared (FLIR) technology, which achieves detection accuracy of 45%. A recent pilot study conducted in Churchill, Manitoba, introduced synthetic aperture radar (SAR) as a potential solution, and despite its challenges, he achieved 66% accuracy in polar bear identification.
Image credit: Sergey Uryadnikov/Shutterstock.com
This latest research was published in the journal Ursus In October 2023.
Ongoing research shows that machine learning can effectively address all SAR issues. Machine learning data analysis capabilities improve detection accuracy in SAR images, reducing false negatives and minimizing disturbances, especially through neural networks that distinguish polar bears from their surroundings.
This technology not only improves accuracy, but also reduces the risk of disturbing animals and provides the possibility of consistent burrow detection. This is a challenge that FLIR technology struggles with.
The integration of synthetic aperture radar (SAR) technology and machine learning not only promises significant advances in polar bear detection, but also has significant environmental implications. This innovation supports the protection of polar bear populations and their fragile ecosystems by increasing the accuracy and efficiency of wildlife monitoring.
Reducing disturbance during the denning period ensures the health of these iconic creatures while protecting their habitat. Furthermore, the adaptability of SAR extends to broader ecosystem monitoring, providing a comprehensive approach to environmental protection.
Researchers can use SAR to assess polar bear populations, identify individuals, and monitor their health without physical contact. Beyond polar bears, SAR and machine learning have the potential for comprehensive ecosystem monitoring and conservation, creating opportunities for environmental and ecosystem consulting, government conservation organizations, researchers, and others.
sauce:
George, B. other. (no date) Evaluation of the effectiveness of airborne synthetic aperture radar in detecting polar bears: A pilot study, bio one complete. Available at: https://bioone.org/journals/ursus/volume-2023/issue-34e6/URSUS-D-22-00018/On-evaluating-the-efficacy-of-air-borne-synthetic-aperture-radar/10.2192/ URSUS-D-22-00018.short?tab=Article link (Accessed: October 19, 2023).