NASA scientists scanned over 10 billion individual trees in Africa's drylands using commercial, high-resolution satellite pictures and artificial intelligence (AI) to estimate the amount of carbon stored outside of the continent's vast tropical forests.

The result is the first thorough assessment of tree carbon density in Africa's Saharan, Sahelian, and Sudanese zones.

1 Billion Metric Tons of Carbon

The researchers discovered there are many more trees than previously believed dispersed over semi-arid regions of Africa, but they also store less carbon than some models have anticipated.

According to the latest study, approximately 0.84 petagrams of carbon, or 1 billion metric tons, are trapped in African drylands.

For grasses and shrubs that develop seasonally, this "carbon residence time," as scientists refer to it, is quite brief, but for trees that grow for years, it is considerably longer. NASA notes that it is essential to know exactly what is growing in a landscape to estimate its carbon storage capacity.

Outside the immense tropical forests that cover much of the center of the continent, Africa's landscapes vary from arid grasslands with a few trees to savannahs with a few scattered trees to more humid regions brimming with many trees as well.

Scientists have found it challenging to estimate the number of trees in these places due to the dispersed tree cover, and their estimations have frequently been either too high or too low. But these observations are necessary to understand the global carbon cycle and to support conservation efforts.

"Our team gathered and analyzed carbon data down to the individual tree level across the vast semi-arid regions of Africa or elsewhere - something that had previously been done only on small, local scales," Compton Tucker, lead scientist on the project and an Earth scientist at NASA's Goddard Space Flight Center, said in a press release statement.

Read Also: NASA's Upcoming Roman Space Telescope to Rewind Universe, Solve Cosmic Mysteries

Using AI and Satellite Photos

The scientists sorted through more than 326,000 commercial satellite photos from the QuickBird-2, GeoEye-1, WorldView-2, and WorldView-3 satellites using powerful machine learning and artificial intelligence methods for the new study.

The team used NASA's Center for Climate Simulation and its Explore/ADAPT Science Cloud to organize and get the photos ready for machine learning processing.

Martin Brandt of the University of Copenhagen created training data for AI using 89,000 individual trees. Using high-resolution images of Africa's drier, less verdant landscapes, Ankit Kariyaa, a colleague at Copenhagen, adjusted a neural network so that computers could detect the individual trees in 50-centimeter scale photos.

With the help of a viewer app the team created, the African tree carbon data are publicly available. Every tree in the research area may be seen, along with how much carbon it stores.

Scientists and students researching the carbon cycle may find this data helpful, as may decision-makers striving to strengthen conservation efforts and farmers interested in learning how much carbon is stored on their land.

The study was published in the journal Nature.

Related Article: [WATCH] NASA, ESA's Solar Orbiter Catches Mercury Traveling Across the Face of the Sun