A new approach for estimating living vegetation volume based on terrestrial point cloud data.

Living vegetation volume (LVV), one of the most difficult tree parameters to calculate, is among the most important factors that indicates the biological characteristics and ecological functions of the crown. Obtaining precise LVV estimates is, however, challenging task because the irregularities of many crown shapes are difficult to capture using standard forestry field equipment. Terrestrial light detection and ranging (T-LiDAR) can be used to record the three-dimensional structures of trees. The primary branches of Larix olgensis and Quercus mongolica in the Qingyuan Experimental Station of Forest Ecology at the Chinese Academy of Sciences (CAS) were taken as the research objects. A new rapid LVV estimation method called the filling method was proposed in this paper based on a T-LiDAR point cloud. In the proposed method, the branch point clouds are divided into leaf points and wood points. We used RiSCAN PRO 1.64 to manually separate the leaf points and wood points under careful visual inspection, and calculated that leaf points and wood points accounted for 91% and 9% of the number of the point clouds of branches. Then, the equation LVV = V₁N (where N is the number of leaf points, and V₁ is cube size) is used to calculate LVV. When the laser transmission frequency is 300,000 points/second and the point cloud is diluted to 30% using the octree method, the point cloud can be replaced by a cube (V₁) of 6.11 cm³ to fill the branch space. The results showed that good performance for this approach, the measuring accuracy for L. olgensis and Q. mongolica at the levels of α = 0.05 and α = 0.01, respectively (94.35%, 90.01% and 91.99%, 85.63%, respectively). The results suggest that the proposed method can be conveniently used to calculate the LVV of coniferous and broad-leaf species under specific scanning settings. This work is explorative because hypotheses or a theoretical framework have not been previously defined. Rather, we would like to contribute to the formation of hypotheses as a background for further studies. [ABSTRACT FROM AUTHOR]