New tech hoped to be used alongside existing systems to help detect wildfires and provide earlier warning strategies
By
As the capabilities of artificial intelligence continue to grow, it seems as if there is little the technology can’t do. From being able to detect illegal logging in West Africa to monitoring volcanic activity across the globe, its applications in the world of environmental science are vast.
And in the latest development of its ever-evolving powers, researchers have now shown how a type of AI has ‘great potential’ for detecting wildfires in a new study of the Amazon rainforest.
Enjoying this article? Check out our related reads:
This type of artificial intelligence mimics the function of the human brain, using ‘Artificial Neural Networks’ to combine satellite imaging technology with deep learning.
The researchers’ findings, published in the International Journal of Remote Sensing, show a 93 per cent success rate when training the model via a dataset of images of the Amazon rainforest with and without wildfires. Afterwards, the model was presented with a further 40 images not included in its training. It correctly identified 23 out of 24 images with wildfires and all 16 where wildfires were not present – bolstering confidence that the technology is robust.
Researchers say that using this technology alongside other AI systems could help enhance early warning systems and improve wildfire response strategies. For example, the model could help alert local authorities about wildfires before they become out of control.
Such a development is critical in protecting the Amazon, a region which experienced more than 98,000 wildfires in 2023 alone. In 2024, 44.2 million acres of the Amazon were burned in Brazil – an area larger than the entire US state of California.
Other Amazonian countries also experienced a record number of fires in 2024, such as Bolivia, while Peru, Colombia, Ecuador and Guyana were home to huge fire surges in the same year.
Wildfires are increasing in the Amazon predominately for two major reasons: droughts (worsened through climate change), and the expansion of industrial agriculture (which leaves vast areas deforested, and leftover vegetation extremely dry and prone to igniting).
‘The ability to detect and respond to wildfires is crucial for preserving the delicate ecological balance of these vital ecosystems, and the future of this Amazon region depends on decisive rapid action,’ said lead author Professor Cíntia Eleutério of the Universidade Federal do Amazonas.
‘Our study’s findings could improve wildfire detection in the Amazonian ecosystem and elsewhere in the world, significantly assisting authorities in combating and managing such incidents.’
Researchers are also keen for others to explore alternate applications of this new AI technology, for example, in combatting deforestation around the world.
