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

Area of Science:

• Forestry Science
• Remote Sensing
• Ecology

Background:

• Living vegetation volume (LVV) is crucial for assessing tree health and ecological function.
• Estimating LVV is challenging due to complex crown structures and limitations of traditional forestry equipment.
• Terrestrial light detection and ranging (T-LiDAR) offers a 3D approach to capture tree structures.

Purpose of the Study:

• To develop a novel, rapid method for estimating living vegetation volume (LVV) using T-LiDAR point cloud data.
• To address the difficulties in accurately measuring LVV for irregularly shaped tree crowns.
• To validate the proposed method on Larix olgensis and Quercus mongolica species.

Main Methods:

• A new 'filling method' was proposed based on T-LiDAR point clouds.
• Branch point clouds were manually separated into leaf and wood points using RiSCAN PRO 1.64.
• LVV was calculated using the equation LVV = V1N, where V1 is a cube size of 6.11 cm³ and N is the number of leaf points, under specific scanning parameters (300,000 points/second, 30% dilution via octree method).

Main Results:

• Leaf points constituted 91% and wood points 9% of the branch point clouds.
• The filling method demonstrated good performance with high measuring accuracy for L. olgensis (94.35% at α=0.05) and Q. mongolica (91.99% at α=0.01).
• Accuracy levels were also reported as 90.01% and 85.63% respectively for the species.

Conclusions:

• The proposed filling method provides a convenient and accurate way to estimate LVV for both coniferous and broad-leaf species.
• The method is effective under specific T-LiDAR scanning settings.
• This explorative work contributes to forming hypotheses for future research in tree volume estimation.