Assessing branching structure for biomass and wood quality estimation using terrestrial laser scanning point clouds

Abstract

Terrestrial laser scanning (TLS) accompanied by quantitative tree-modeling algorithms can potentially acquire branching data non-destructively from a forest environment and aid the development and calibration of allometric crown biomass and wood quality equations for species and geographical regions with inadequate models. However, TLS’s coverage in capturing individual branches still lacks evaluation. We acquired TLS data from 158 Scots pine (Pinus sylvestris L.) trees and investigated the performance of a quantitative branch detection and modeling approach for extracting key branching parameters, namely the number of branches, branch diameter (bd) and branch insertion angle (bα) in various crown sections. We used manual point cloud measurements as references. The accuracy of quantitative branch detections decreased significantly above the live crown base height, principally due to the increasing scanner distance as opposed to occlusion effects caused by the foliage. bd was generally underestimated, when comparing to the manual reference, while bα was estimated accurately: tree-specific biases were 0.89 cm and 1.98°, respectively. Our results indicate that full branching structure remains challenging to capture by TLS alone. Nevertheless, the retrievable branching parameters are potential inputs into allometric biomass and wood quality equations.