Showing where beetles will kill trees if left unchecked
Bark beetles are insects about a centimetre long. They lay their eggs under the bark of trees and as the larvae hatch, they start eating. A single larva, or even a dozen, don’t do that much damage. However, when there are more, they have the ability to kill entire forests. During her own startup experience, Lisa Broekhuizen worked on forest pest risk assessment in European forestries. Today, at Space4Good she continues this work, helping foresters prevent damage from bark beetles.
At first, bark beetles target weakened or dead trees. From there, they reproduce and as they gain in number, the beetles start targeting healthy trees. As populations grow, entire forests can be destroyed. While technology hasn’t advanced to the point where we can see these centimetre long bugs from space, we can do the next best thing. We look where they are likely to be.
The large destruction by these beetles is not natural, at least not in most European forests. Over time humans have selectively logged certain trees. By changing the forest composition, larger numbers of pest insects were able to spread. Additionally, climate change has resulted in longer summers and warmer winters, meaning more generations of beetles are developing per year and more adults surviving the winter resulting in an exponential population increase in these insects in early spring.
Bark beetles reproduce under particular conditions. By taking into consideration forest age, vegetation moisture, temperature and tree characteristics we can ascertain likely areas of bark beetle breeding. Many of these we know from foresters, and now we can also detect this from space. Based on all the past bark beetle activity that the foresters have gathered, we trained an algorithm to identify the highest risk areas. This is combined with satellite data indicating forest health developments from Sentinel 2.
Every couple of days we have a new data input, which we compare with data of the past five years. By combining the data in our models, we see if the forest is developing as it should, or if there is a reason for alarm. We can then go on to predict where there is a high risk for forest destruction in the future and flag it as a potentially high-risk zone. Foresters can then investigate these areas specifically to best mitigate the infestations.
However, like most things, it’s not that simple. A sick tree, or a tree missing a branch, does not mean there is an insect outbreak. To combat this, we combine our findings with a phonology module. This looks into the development of the beetles and is a good measure for population size. We investigate temperature over time to further determine the high-risk periods. We can then indicate this so that the foresters can take this information into account in the field.
Over the bark beetle season (March — November) 2020 the model was deployed in Switzerland. Although the beetles are no longer reproducing, some survive under fallen leaves and eggs remain developing under the bark. We use this precious time to assess how well the model performed, and make adjustments for next year where we aim to continue our work monitoring these beetles from space.
Originally published at https://www.space4good.com on November 26, 2020.