There are many more trees in the West African Sahara Desert than you might think, according to a study combining artificial intelligence and detailed satellite imagery.

Researchers have counted more than 1.8 billion trees and shrubs in the 1.3 million square kilometers (501,933 square miles) area that covers the westernmost part of the Sahara Desert, the Sahel and this known as the subhumid zones of West Africa.

“We were very surprised to see that a number of trees actually grow in the Sahara Desert, because until now most people thought there were hardly any,” says Martin Brandt. , Professor in the Department of Geosciences and Natural Resource Management at the University of Copenhagen and lead author of the study in Nature.

“We have counted hundreds of millions of trees in the desert alone. It would not have been possible without this technology. Indeed, I think this marks the beginning of a new scientific era.

The researchers used detailed NASA satellite images and deep learning, an advanced artificial intelligence method. Normal satellite imagery is unable to identify individual trees, they literally remain invisible. In addition, a limited interest in counting trees outside forest areas has led to the dominant opinion that this particular region has almost no trees. This is the first time anyone has counted trees in a large arid region.

Trees and the global carbon footprint

New knowledge about trees in drylands like this is important for several reasons, Brandt says. For example, they represent an unknown in the global carbon footprint.

“Trees outside forest areas are generally not included in climate models and we know very little about their carbon stocks. They are essentially a white spot on the cards and an unknown component of the global carbon cycle, ”he says.

In addition, the new study contributes to a better understanding of the importance of trees for biodiversity and ecosystems and for the people living in these areas. In particular, improving knowledge about trees is also important for the development of programs to promote agroforestry, which plays a major environmental and socio-economic role in arid regions.

“So we are also interested in using satellites to determine tree species, as tree types are important in relation to their value to local people who use wood resources as part of their means of survival. subsistence, ”explains Rasmus Fensholt, professor of geosciences and natural resources. management department.

“Trees and their fruits are consumed by both livestock and humans, and when kept in fields, trees have a positive effect on crop yields as they improve the balance between water and nutrients. “

Thousands of trees identified in a few hours

Researchers at the Department of Computer Science at the University of Copenhagen developed the deep learning algorithm that made it possible to count trees over such a large area.

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 only needs a few hours, which would take thousands of humans several years.

“This technology has enormous potential when it comes to documenting changes on a global scale and ultimately contributing to global climate goals. We are motivated to develop this type of beneficial artificial intelligence, ”explains professor and co-author Christian Igel from the IT department.

The researchers will then extend 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.

Source: University of Copenhagen