A couple of research scientists have created a satellite map of anthropogenic impact on the land. According to the Colorado State University Press Office, the Atlas provides new insights into the state of the environment.
The coronavirus pandemic has forced researchers to switch jobs or temporarily abandon projects due to health safety protocols or the inability to travel. But for Patrick Keys and Elizabeth Barnes, husband, and wife of Colorado State University (CSU) scientists, the past year has been a fruitful one. The fact is that they teamed up with Neil Carter, an assistant professor at the University of Michigan, to work on an article published in the journal Environmental Research Letters. In it, the authors of the study talk about creating a satellite map of anthropogenic impact on the regions of the planet.
Scientists have used machine learning to create an atlas that shows exactly where in the world the dramatic landscape changes have occurred. The map shows the impacts of deforestation, mining, expansion of road networks, urbanization, and agricultural growth.
The Human Footprint Index (HFI) is a widely used tool for interpreting humanity’s increasing pressures on Earth. Until now, the HFI creation process has required significant data and modeling, and updated versions of the index are often many years behind today. In the new paper, scientists presented a global machine learning-based HFI (ml-HFI) that can be regularly updated using only satellite imagery. In the study, scientists presented the latest HFI map and documented changes in human influence on nature over the past 20 years (from 2000 to 2019).
“The map we have developed can help people understand the important issues of biodiversity conservation and sustainability in general,” said Keyes, lead author and research fellow at CSU’s School of Global Environmental Sustainability.
Maps of this type can be used to track progress towards the United Nations Sustainable Development Goal 15 (SDG 15), Life on Earth, which aims to conserve biodiversity.
For their map, the researchers created a convolutional neural network (CNN), which is commonly used to interpret images. This is similar to how Facebook works when the site offers to tag friends in a photo.
In developing the algorithm, scientists used existing data that classified human impact on the planet. These include factors such as roads and buildings, as well as pastures for livestock and deforestation. The convolutional neural network then learned how to interpret satellite images based on this existing data accurately.