AI finds hidden trees across Europe

New research from the University of Copenhagen’s Department of Geosciences and Natural Resource Management has used an advanced algorithm that combines satellite imagery and artificial intelligence to closely examine tree coverage in areas outside of traditional forests, shedding light on the importance of these hidden trees.

The study reveals that there are approximately 15 million hectares of tree coverage outside of traditional forested areas in Europe, which is equivalent to around a billion tons of hidden biomass. These trees are typically not included in national forest inventories, as these inventories traditionally focus on trees within designated forested areas.

This hidden tree coverage has several significant implications, particularly in countries like the Netherlands, the UK, Ireland, and Denmark, which have a substantial proportion of their tree cover located outside of traditional forested areas. In the Netherlands, for example, nearly 25% of the country’s tree cover is outside forests, with 8% growing in urban areas. The UK and Ireland also have a considerable proportion of their tree coverage outside of traditional forests. In contrast, countries with extensive forested areas, such as Finland, have a smaller proportion of tree coverage outside forests.

This research demonstrates the importance of including these hidden trees in climate models, biomass inventories, and biodiversity assessments. Trees outside traditional forests play a crucial role in biodiversity, microclimate regulation, habitat preservation, landscape aesthetics, and hydrological cycles, particularly in countries with limited forest resources.

The study relied on an advanced deep learning algorithm that can recognize trees down to three meters in height based on their crowns, using detailed satellite imagery. The algorithm achieved a remarkable precision of 92.4%. This level of precision and detail offers a significant advancement in monitoring forest resources and biomass, enabling more frequent and accurate assessments in a rapidly changing landscape.

This research builds on a series of studies conducted over the past few years that have used advanced technology, including detailed satellite imagery and artificial intelligence, to map tree cover in various regions. These advances make it easier to monitor forests and biomass, providing valuable insights for conservation and climate-related efforts.

The study was based on data from national forest inventories, which were used to calibrate and validate the algorithm’s results. This combination of advanced technology and field data allows for more comprehensive and precise assessments of tree coverage, even in areas outside traditional forests.

https://phys.org/news/2023-10-hidden-trees-europe-billion-tons.html