SLU news

New spectral features constructed from green light to early identify bark beetle-infested trees

Published: 25 September 2024
Large drone with camera mounted underneath. In the background, grass, forest, and a house. Photo.

A new groundbreaking technique using hyperspectral drone images now enables detection of the majority of trees infested by European spruce bark beetles before the new generation beetles leave the trunk. The technique is developed using hyperspectral drone imagery but aims to be applied by simpler sensors and applied for large-area monitoring.

Researchers at the Swedish University of Agricultural Sciences (SLU) uncovered the spectral characteristics that first responded to the decline of tree vitality, and used them to construct indices to identify more infested trees at an earlier stage. Using the new indices, 80 % of infested trees can now be detected before the new generation of beetles emerges from the trees, compared to the detection rates of 30% when using commonly used indices, such as NDVI.

The research indicates that the visible green wavelength at 530 nm, so-called green shoulder, is particularly effective in distinguishing between healthy and infested trees, rather than the green peak at 560 nm widely used in cameras. The indices were further improved by simplifying the use of reflectance from only three visible bands, which means the technique for early detection does not rely on expensive hyperspectral cameras.

The simplified method makes it possible to use simpler sensors, such as those found in cost-effective multispectral cameras and satellite images, for early detection of infested trees. This technique has the advantage of wide-scale coverage of satellite imagery and can be used over large forest areas.

– This new finding also deepens our understanding of how trees react to stress, says Langning Huo, researcher at SLU. We thought the first symptoms of an infestation were reduced chlorophyll and water content in the needles, but this research indicated an earlier change in the pigment of the chloroplasts (xanthophyll content) and highlighted the change in light use efficiency under stress. This gives us new insights out of the box to develop methods for earlier detection of stressed trees.

The new technique enables faster and more accurate monitoring of stressed trees, and early identification can help limit the spread of the next generation of bark beetles. The spruce bark beetle annually causes extensive damage to the forest, and with climate change, the risk of outbreaks increases. By improving detection techniques, forest management can be implemented more quickly to limit damage and the spread of beetles.

Facts:

Read the research paper: https://www.sciencedirect.com/science/article/pii/S0924271624002946

The research is supported by the SLU Forest Damage Center, including a research mobility for Langning Huo to stay at the Finnish Geospatial Research Institute (FGI). The research was conducted in collaboration with Prof. Eija Honkavaara’ team at FGI using data collected in Finland. A refined project started in 2023 in Sweden, and Luiz Cosimo, a doctoral student at SLU, is conducting research to test new findings and further refine the methodology. The new project is co-funded by SLU Forest Damage Center and the European Union within the project “Network for novel remote sensing technologies in forest disturbance ecology” under the Horizon 2020 framework. Project partners: Swedish University of Agricultural Sciences, Slovak Academy of Sciences, Finnish Geospatial Research Institute, and University of Eastern Finland.