Fallen aspen. Credit: Samuli Junttila

Aspens and standing dead trees, which are important to forest biodiversity, can be reliably identified from openly available aerial imagery using methods developed by researchers from the University of Helsinki and the University of Eastern Finland.

Aspen (Populus tremula) is an important keystone species. Only a small percentage of the trees in Finnish forests are aspens, and actual aspen stands are even rarer. Aspens nevertheless host more than a thousand species: mammals, birds, insects, fungi and mosses. The rapidly decomposing litter of aspens boosts nutrient cycling in forests, and many species thrive only on aspens’ gnarled and base-rich bark.

A study conducted by a research group led by Associate Professor Samuli Junttila at the University of Helsinki introduces a method for reliably distinguishing, for the first time, aspen trees, which support and embody biodiversity, from other trees using open aerial image data. Previously, extensive and accurate mapping of aspen distribution has been expensive and laborious. "Our research offers an up-to-date, scalable, inexpensive and reliable method for monitoring forest biodiversity," Junttila says.

The Global Ecosystem Health Observatory research group led by Junttila develops new methods for obtaining accurate information on the forest environment and its ecology by combining remote sensing techniques with AI solutions. The methods created by Junttila’s group have previously helped, for example, in monitoring the destruction caused by the European spruce bark beetle and tree mortality.

Doctoral Researcher Anwarul Chowdhury from the University of Eastern Finland is the lead author of the article published in October 2025 in Remote Sensing Applications: Society and Environment. He is very pleased with the accuracy of the neural network model created in the study.

"This study demonstrates how our methods can produce reliable data for the practical needs of forest management and conservation in all forests in Finland," Chowdhury says.

The model designed by the researchers proved effective in all seasons, as it reliably identified aspens with and without leaves. The model was more likely to identify taller than smaller aspens—on average, fully grown, taller trees were identified with a 71% probability and even better when the trees had no leaves. This information is important, as tall, old aspens are particularly significant for biodiversity.

In the future, the research group would also like to improve the accuracy of identifying juvenile aspens.

"Going forward, open laser scanning data could be combined with aerial images to make the model better identify young aspens as well," Chowdhury muses.

Even dead trees affect biodiversity

In November, another study by researchers Anis Rahman, Einari Heinaro and Mete Ahishali from Junttila’s research group was published in the International Journal of Applied Earth Observation and Geoinformation, developing a more accurate technique for identifying standing dead trees from aerial images.

Dead trees are also important to biodiversity because many specialized and even threatened species depend on them. However, identifying dead trees from aerial images is challenging, especially under dense canopy structures. The group combined machine-learning algorithms and complex adaptive filters to produce better results when surveying dead trees compared with the forest remote sensing models intended for general use.

"Aspens and standing dead trees are important indicators of biodiversity. Their automated mapping from extensive open datasets is a big step forward in monitoring the biodiversity of forest environments," Junttila says.

More information: Anwarul Islam Chowdhury et al, Mapping large European aspens (Populus tremula L.) using national aerial imagery and a U-Net convolutional neural network, Remote Sensing Applications: Society and Environment (2025). DOI: 10.1016/j.rsase.2025.101755

Anis Ur Rahman et al, Dual-task learning for dead tree detection and segmentation with hybrid self-attention U-Nets in aerial imagery, International Journal of Applied Earth Observation and Geoinformation (2025). DOI: 10.1016/j.jag.2025.104851

Citation: AI pinpoints aspens and standing dead trees in forests using aerial imagery (2025, December 10) retrieved 10 December 2025 from https://phys.org/news/2025-12-ai-aspens-dead-trees-forests.html

This document is subject to copyright. Apart from any fair dealing for the purpose of private study or research, no part may be reproduced without the written permission. The content is provided for information purposes only.