It is not an easy task to count the trees on Earth’s surface. NASA has now ventured into mapping the trees on earth using supercomputers and Artificial Intelligence. The Scientists from NASA’s Goddard Space Flight Center Greenbelt, Maryland, has shown a new methodology for mapping the location and size of the trees that grow outside the forest area.
In their observation, NASA scientists have come across billions of trees in arid and semi- arid regions. The team mapped the crown diameter (width of the tree when viewed from above), using supercomputers and machine learning algorithms. They mapped more than 1.8 billion trees across an area of more than five lakh square miles. The team mapped the crown diameter taking into account rainfall and land use. The scientists also looked at how much carbon the trees store.
Carbon is a primary building block for all life on Earth. Trees are called carbon “Sinks” as they use carbon for growth and store it in their roots, leaves, branches and trunks. When the trees are burnt, carbon is released into the atmosphere as carbon dioxide, which is one of the main causes of climate change.
The NASA team used a powerful computing algorithm at the University of Illinois Blue Waters. This is one of the world’s fastest supercomputers. The team first trained the model by manually marking about 90,000 trees across a variety of terrain. After this they allowed it to “learn” about the shapes and shadows that indicated the presence of trees. Study’s lead author Martin Brandt (assistant professor of geography at the University of Copenhagen) said that coding the training data took more than a year. He is credited with marking 89,899 trees and Ankit Kariryaa of the University of Bremen led the development of deep learning computer processing.
Jesse Mever programmer at NASA Goddard who led the processing on Blue Waters) said that data would help in preservation, restoration and other purposes. They are helpful in developing a baseline.