Drones as a cheap option for forest structure characterization
Virginia Garcia will give a seminar.
Different forestry management systems are evaluated at the EMEND project, Canada (https://emend.ualberta.ca/) to find out the most sustainable ways of producing wood. There, one birch forest area has been kept pristine as a baseline to evaluate the different management systems against a natural primary forest.
The quality and characteristics of what a natural forest is, can be defined by its structural parameters: tree height, trunk diameter, forest density, clearings, tree distribution, etc.
Laser scanners can record the full forest as a digital cloud of points, which can later be used to calculate the aforementioned forest structure parameters. However, terrestrial and airborne LiDAR are expensive tools and require intensive fieldwork and sophisticated data processing and analysis. Point clouds can also be reproduced from aerial flights, by photogrammetry, stereographic interpolation and tie point recognition. However, the number of retrieved points is much smaller in the case of aerial flights, and therefore, it is thought that the precision of the point cloud is smaller than for LiDAR.
A field campaign was performed in 2015 in EMEND baseline forest, where terrestrial LiDAR (Light Detection Aperture Range, or laser) and UAV (Unmanned Aerial Vehicles, or drones) data was collected at similar times. In this study, I compare the LiDAR dataset with un-expensive and uncomplicated UAV data to evaluate if commercial UAVs could reach similar accuracy than LiDAR, in the estimation of forest structural parameters. In addition, the difference in precision between techniques will be estimated.
This seminar will be given via Zoom.
Meeting ID: 682 3216 9029