Inside Science
/
Article

Artificial Intelligence Finds Hidden Roads Threatening Amazon Ecosystems

JAN 08, 2021
Researchers in Brazil are hunting for unofficial roads -- many of them illegal -- tied to rainforest destruction.
Artificial Intelligence Finds Hidden Roads Threatening Amazon Ecosystems lead image

Artificial Intelligence Finds Hidden Roads Threatening Amazon Ecosystems lead image

Paralaxis via Shutterstock

(Inside Science) -- It took years of painstaking work for Carlos Souza and his colleagues to map out every road in the Brazilian Amazon biome. Official maps of the 4.2 million-square-kilometer region only show roads built by federal and local governments. But by carefully tracing lines on satellite images, the researchers concluded in 2016 that the true length of all the roads combined was nearly 13 times higher.

“When we don’t have a good understanding of how much roadless areas we have on the landscape, we probably will misguide any conservation plans for that territory,” said Souza, a geographer at a Brazil-based environmental nonprofit organization called Imazon.

Now, Imazon researchers have built an artificial intelligence algorithm to find such roads automatically. Currently, the algorithm is reaching about 70% accuracy, which rises to 87%-90% with some additional automated processing, said Souza. Analysts then confirm potential roads by examining the satellite images. Souza presented the research last month at a virtual meeting of the American Geophysical Union.

The laborious work of mapping roads by hand was not wasted -- that data was needed to train the AI algorithm. Thanks to the algorithm, Souza and his colleagues should now be able to update their map every year with relative ease. The team also plans to share the algorithm with organizations in surrounding countries so roads can be mapped in other parts of the Amazon basin.

People build unofficial roads through the Amazon for a variety of reasons, said Souza. Logging companies build winding roads so they can access timber, while miners build straighter roads to reach gold deposits. Other roads are built for agriculture or as part of land speculation before selling public lands to private owners. Determining the legality of such roads is complicated; many are illegal, but some are permitted in logging and mining concessions, said Souza.

Large areas of road-free rainforest are important for protecting Amazonian biodiversity and isolated indigenous people, said Souza. Moreover, roads are often a harbinger of further destruction. Nearly 95% of deforestation in the Brazilian Amazon occurs within 5.5 km of a road or 1 km of a river, while about 95% of fires occur within 10 km of a road or river, according to prior research by Souza and his colleagues. Loggers and gold miners often abandon private roads when natural resources are exhausted, said Souza, whereupon farmers and ranchers make use of them for further development.

If policymakers don’t consider unofficial roads, they may underestimate the harm being done to the Amazon, said Souza. The new algorithm could help provide a complete and up-to-date picture, showing where to focus efforts at rainforest protection.

The unofficial roads, said Souza, cover a huge area. “If we don’t stop this, it’s going to be the next frontier of occupation and deforestation.”

More Science News
/
Article
Excitation-wavelength-dependent Raman shifts in a microbial rhodopsin–carotenoid complex reveal structural heterogeneity.
/
Article
A surrogate plant enables studies of complex submerged vegetation bending and swaying in flowing water.
/
Article
DIY triboelectric nanogenerators (TENGs) could connect physics students with this emerging technology and its many applications.
/
Article
Evaluating the critical bandwidth of pinnipeds can reveal differences in their ability to separate signals from noise.
/
Article
Over a fiber-optic line with state-of-the-art throughput, light carrying the classical information of internet users was joined by photons serving as quantum cryptographic keys.
/
Article
What had appeared to be an upper limit to the strength of the events is an artifact of measurement uncertainty and statistical bias. Those lessons could be relevant for data analyses in other disciplines.
/
Article
/
Article
Detecting quantum particles of gravity was long considered hopeless. A new perspective, which takes advantage of quantum technologies, suggests that an experimental realization may be closer than previously thought.