Press Play and Listen to this article on Nyongesa Sande!!
There are more than 1.8 billion trees in the Sahara Desert. A combination of high-resolution satellite imaging and ‘deep learning’ has identified more than 1.8 billion trees across the West African Sahara, Sahel and sub-humid zone – significantly more trees than were previously thought to exist in the region. The collaboration between NASA and several geoscience departments across the world used 11,128 satellite images from four satellites to count individual trees across 1.3 million square kilometres.
The deep-learning approach has, for the first time, allowed researchers to identify individual trees across the dryland expanse. Because of the absence of closed canopies, many parts of the Sahara and the Sahel have previously been mapped with zero per cent tree cover. ‘You need high-resolution satellite images to be able to detect individual trees and not just to make estimations based on identified areas of canopy cover,’ says Martin Brandt from the University of Copenhagen.
AI has just revealed there are over 1.8 billion trees in the Sahara Desert
Traipsing through new satellite images, Brandt manually identified nearly 90,000 individual trees in the Sahara and the Sahel. Each identification increased the sophistication of a deep-learning computing system. The system then took over, using data from Brandt’s manual identifications to spot individual tree-like objects in the satellite images. The final map shows that tree density develops along the rainfall gradient. Trees are sparse in the hyper-arid Sahara Desert in the north, are scattered in arid and semi-arid lands, and form denser coverage in the sub-humid south.
More than anywhere else on Earth, the Sahel is on the frontline of climate change. Persistent droughts are hastening desertification, soils are eroding and agricultural yields are declining. For these reasons, the region’s scattered trees are precious. They combat desertification and soil erosion, locking in nutrients for agriculture, on which 80 per cent of inhabitants depend. Tree-derived products provide a source of income for communities, while fruits and leaves are a valuable source of food. ‘Many people in dryland, semi-arid and arid areas depend on trees for their entire livelihoods,’ says Brandt.
Enumerating the trees that dot these drylands is essential to our understanding of their value to local communities, but also to the rest of humanity. ‘Thirty per cent of our carbon emissions enter terrestrial carbon sinks and numerical simulation modelling indicates that about 40 per cent of this will enter sinks in arid and semi-arid regions,’ says Compton Tucker, an earth scientist at NASA and co-author of the study. ‘Mapping tree distribution in the Sahara and the Sahel – vast tracts of which were once thought to be barren and treeless – will build our understanding of the carbon-sequestration potential of the region.’
In 2007, more than 20 African governments came together to launch the Great Green Wall project, an ambitious initiative that aims to plant an 8,000-kilometre-long forest from Dakar, Senegal to Asmara, Eritrea – the entire length of the Sahel. Brandt and Tucker think that a more complete understanding of the area’s vegetation is key to the project’s success. ‘It’s not just about planting as much as possible: there should be a sustainable and ecologically reasoned plan behind planting initiatives. Data such as ours can help,’ says Brandt. The scientists’ next step will be to quantify the carbon-sequestration potential of these newly identified trees. ‘By doing that, we might further understand the mechanisms that control carbon sequestration,’ says Tucker.
Thousand of trees identified in just hours.
Researchers from the University of Copenhagen’s computer science department developed the deep learning algorithm that made the counting of trees over such a large area possible.
The researchers fed the deep learning model thousands of images of various trees to show it what a tree looks like. Then, based on the recognition of tree shapes, the model could automatically identify and map trees over large areas and thousands of images. The model needs only hours what would take thousands of humans several years to achieve.
“This technology has enormous potential when it comes to documenting changes on a global scale and ultimately, in contributing towards global climate goals. We are motivated to develop this type of beneficial artificial intelligence,” says professor and coauthor Christian Igel of the computer science department.
Researchers will next expand the count to a much larger area in Africa. And in the longer term, they plan to create a global database of all trees growing outside forest areas.
License and Republishing