Application of LiDAR System for Monitoring and Management of Artificial Forest

• Main Authors: Wei, Chun-Hung, 魏浚紘
• Other Authors: Chen, Chaur-Tzuhn
• Format: Others
• Language: zh-TW
• Published: 2014
• Online Access: http://ndltd.ncl.edu.tw/handle/43528302266555183642

博士 === 國立屏東科技大學 === 生物資源研究所 === 102 === The techniques of artificial forest management are focused on thinning operation and space adjusting, and the criteria of traditional thinning operation are based on quantitative and qualitative data. How to select thinning timbers and monitor the effects of thinning are the important subjects in artificial forest management. In this study, using 3D scanning Light Detection and Ranging (LiDAR) systems, which are the non-destructive methods, to investigate and monitor individual trees in artificial forest sample areas. Using ground-based LiDAR and traditional field survey data to obtain individual tree characteristic values which included tree height, diameter at breast height, canopy height, tree location map and evaluated the feasibility of ground-based LiDAR data in artificial forest areas. Results showed that the accurate information from the ground-based LiDAR system in permanent sampling plots, and different scanning stations could decrease occlusions. By using high density point cloud data, the error on the individual tree location that was over 1 meter could be controlled to less than 11% and the height and diameter at different height of the tree could also be measured directly by the LiDAR systems. Different number of point cloud affected the accuracy of diameter. When the ratio of point cloud fell below 100%(2.84 points cm-2), the coordinates of individual trees would shift from the accurate position on the circle fitting because of insufficient points. Results also showed that by the characteristics obtained from the LiDAR spectrum, the systems could automatically separate leaves, branches and stems and could also establish tree volume model non-destructively with tree height and DBH data obtained from ground-based LiDAR. The tree competition index could be estimated by the individual trees location, diameter at different height, tree distance and DBH. The diameter at different height could be estimated by the cloud data of LiDAR that were to approach with actual value; however, due to self-occlusion of the crown, errors may occur when estimating branch diameters. Using threshold value to classify gap fraction, results showed that were accurate. Therefore, the results could be used to analyze the horizontal and vertical canopy structure, it was beneficial for the monitoring of plot dynamics in forest areas. Ground-based LiDAR could be used to determine tree measurements, tree competition and canopy gap which solve the problems of quantification and spatial research and also help with the stimulation of thinning operations. Results of this study were beneficial to artificial forest management and improved the forest resource inventory methods.