The rot fungus (Heterobasidion parviporum) is the main cause of economic losses in the Norwegian forestry sector. The loss, or reduction in timber value for spruce, is estimated to be more than NOK 100 million per year. In a large, four-year project, researchers and the forestry industry are now well on their way to develop methods that can limit the spread of the fungus.
“The goal is to move towards precision forestry,” says one of the researchers in the project called ‘Precision’, Professor Erik Næsset.
Imagery and data from airborne laser scanning
The work of obtaining data has been ongoing since the start of the project in 2018. The researchers have collected hyperspectral aerial images, data from airborne laser scanning and images taken with both aircraft and drones.
“Since the end of 2018, a Komatsu harvester has collected huge amounts of data for individual, felled trees,” says Næsset.
Like all modern harvesters, it not only registers dimensions of the produced logs, but also a number of other properties of the felled trees, such as tree species, x and y coordinates and timber quality. The harvester has been equipped with a state-of-the-art GPS that increases the positional accuracy of felled trees.
In addition, the scientists have conducted control measurements of stumps after harvest, where rot was recorded for each stump, and the exact location of each stump was recorded.
Pushing the envelope
In the spring of 2020, two forestry students, Erik Armand Iversen and Ole Marius Tollefsen Moen, submitted their master's theses on the project. The former used hyperspectral images and data from airborne laser scanning to map root rot in spruce forests and improved the accuracy of detected root rot compared to previous studies. The latter investigated methods for predicting how far the root rot extends upwards in a given stem.
“Their theses pave the way for further research, and show that great progress has already been made,” says Næsset.
Accurate positioning of felled trees
Postdoctoral fellow Lennart Noordermeer has looked at the accuracy of the positioning system in the Komatsu harvester. His results show that felled trees can be positioned with an average error of 79 cm.
“Intensive wood measurements from a harvester can be linked to remote sensing data, as an alternative to traditional measurements made in the field,” Næsset comments.
PhD candidate Ben Allen continues to work on the detection of root rot using hyperspectral data. The preliminary results look very promising, and he will continue by including other types of remote sensing data. Furthermore, PhD candidate Ana Claudia Ferreira Aza is investigating how the risk of root rot affects optimal rotation time.
“Precision will continue until 2022 and we expect many exciting results in the time to come,” Næsset concludes.