Skip to Main Content
Tree leaf orientation, including the distribution of the inclinational and azimuthal angles in the canopy, is an important attribute of forest canopy architecture and is critical in determining the within and below canopy solar radiation regimes. Characterizing leaf orientation is a key step to the retrieval of leaf area index (LAI) based on remotely sensed data, particularly discrete point data such as that provided by light detection and ranging. In this paper, we present a new method that indirectly and nondestructively retrieves foliage elements' orientation and distribution from point cloud data (PCD) obtained using a terrestrial laser scanning (TLS) approach. An artificial tree was used to develop the method using total least square fitting techniques to reconstruct the normal vectors from the PCD. The method was further validated on live tree crowns. An equation with a single parameter for characterizing the leaf angular distribution of crowns was developed. The TLS-based algorithm captures 97.4% (RMSE = 1.094 degrees, p <; 0.001) variation of the leaf inclination angle compared to manual measurements for an artificial tree. When applied to a live tree seedling and a mature tree crown, the TLS-based algorithm predicts 78.51% (RMSE = 1.225 degrees, p <; 0.001) and 57.28% (RMSE = 4.412 degrees, p <; 0.001) of the angular variability, respectively. Our results indicate that occlusion and noisy points affect the accuracy of normal vector estimation. Most importantly, this work provides a theoretical foundation for retrieving LAI from PCD obtained with a TLS.