ABSTRACT Even at a well-managed harvesting site, vehicle trafficking occurs on at least 12% of the area and might cause ruts and compaction. The use of drones for inventory and mapping in forestry is still a new method. The purpose of this study was to develop a method for measuring the size and depth of wheel ruts caused by forest machines in harvested areas, using drones and Structure from Motion photogrammetry. In order to investigate the accuracy of drone photogrammetry, measurements from flight altitudes of 60 m and 120 m above ground level were compared with manual measurements. The same methods were used at a control site on farm land, taking into account the rut depth and the location of the sample surface (close to trees or in a fully open area). No statistically significant differences were found between manual measurements and remote measurements from 60 m or 120 m altitude at the harvesting site (R2 0.77–0.83). At the control site, an underestimation of 2.2 cm of the rut depth was found for remote measurements made from 120 m altitude. The data derived from drone images were able to reproduce the 3D model of surface features, such as bulges and ruts; these measurements were considered to be equivalent to manual measurements. For practical applications, a post-harvest survey using drones could contribute to verifying compliance with international forest certification standards or by private contractors to evaluate rut formation on their harvest sites.
Read full abstract