Caroline Kanaskie

Timely, accurate detection of southern pine beetle (SPB) outbreaks is key to managing this tree-killing beetle. The problem is locating these outbreaks early enough: when the beetles first attack, symptoms are only visible along the trunk of the tree.

Stressed trees are known to show spectral signatures that we can't see with just our eyes. Can SPB attack be detected by multispectral cameras, which capture infrared and near infrared light? These wavelengths of light are beyond the visible range, but have been used to identify tree stress in other situations.

On the ground, we recorded visible tree decline symptoms and SPB stage under the bark, following SPB outbreaks.

From above, drones captured visible-light and multispectral imagery.

In the lab, we processed the imagery and calculated spectral indices. A random forest algorithm classified the trees in the imagery.

Principal Investigators:

Dr. Jeff Garnas (UNH)

Dr. Michael Routhier (UNH)

Dr. Matthew Ayres (Dartmouth)

Project manager:

Caroline Kanaskie (UNH)

UAV imagery processing:

Dr. Benjamin Fraser (UNH)

Field technician:

Isabella Barbini (UNH)