Marius Hauglin at the Norwegian University of Life Sciences (NMBU) is researching the digitalisation of forests in order to find a "dream tree" without having to go into the forest.
"For properties that make the trees especially suited for a specific purpose, such as tall, thin light poles or planks without knots, a 'digitalised forest' may provide significant advantages," says Hauglin.
For the forest is like people: few trees are the same and each has its own properties. And it is important to pick the right tree for the right purpose. Finding a dream tree by the click of a button is still some time away. However, finding out more about how a forest looks without having to physically go there is useful and may lead to significant savings for the forestry industry.
One forest, many different trees
For more than 100 years, the forest industry has counted trees and calculated the cubic mass in Norwegian forests. This has been done through sample surveys, where some selected areas are studied closely and used to represent the entire forest – a bit like political opinion polls.
This has made it possible to divide a forest property into stands with similar properties, such as height, species and growing conditions.
However, it has been difficult to say anything precise about an individual tree, except by going into the forest to measure it. However, with 3.1 billion Norway spruce trees, 1.4 billion Scots pine trees and about 4 billion birch trees, this would be an insurmountable task.
Scanning the forest in 3D
New digital tools open up new possibilities. Most of the Norwegian forest has been digitally mapped. The challenge now is to interpret the data.
It would be very advantageous to be able to tell what purposes the trees in a specific area would be suitable for. There may be tall, thin trees for light poles or short, gnarled pine for paper.
"This is where our research is focused. How do we pick a forest stand with the properties we are looking for? We have scanned the forest from the air in 3D for many years. The distance between the points has shorted, which gives greater digital clarity to individual trees. Seen from above, the points form a grid, but we also know how high up from the ground each point is. If we turn the point swarm by 90 degrees, the trees in the forest appear, and this allows us to distinguish the different species of tree with the naked eye. Thus, we have for a long time had a picture of the approximate look of the trees, but not something that is precise enough to indicate what industry may use it for," Hauglin explains.
Links "aerial photographs" with data from harvesters
The trees logged by harvesters are measured and assessed for visible "faults". The resulting measurements can be used to create a mathematical model of each logged tree.
"Put simply, we are trying to link everything registered by the harvesters with how the scanned tree looked in the "aerial photograph". This is because we assume that trees with similar properties look similar. Thus we can look for the same trees in the aerial photograph when we want more trees with the same properties. However, this linking of data is challenging and in practice this is quite difficult," says Hauglin.
What can we tell?
The forest is scanned before it is logged, and then data is registered about each tree when the forest is logged. Through sophisticated GPS measurements, the harvester knows where each tree stood, and this can then be used to link the data.
Forestry student Tallak Dieset is affiliated with the project, and this spring he will investigate whether it is possible to create models that can tell us something about the distribution of the quality of the tree, or in other words how much of it can be used for timber and how much for paper.
The information about quality will be linked with data from the laser scans, and if the trial is successful this will make it possible to estimate the share of timber in a stand in a digitalised forest.
"This is the first step towards the goal of finding the dream tree in a digitalised forest", Hauglin points out.
Benefits for the forestry industry
Though the forestry industry has logged trees for more than 100 years, it remains a relatively costly process. A greater degree of digitalisation will make it cheaper.
"The ongoing research will show whether this method is sufficiently precise, and whether it can replace some of the manual work involved in mapping the forest. If this is possible, it will represent a cost savings for forest owners," Hauglin explains.