青冈栎次生林森林生物量地面激光雷达精准估测方法研究

生物量精准估测是森林资源监测的重要研究内容。地面激光雷达技术弥补了传统人工抽样调查的不足。然而地面激光雷达获取的点云密度高,计算量大,目前缺乏有效的数据简化方法来实现枝、干、叶点云的分离;同时，树干点云缺失修复、三维曲面模拟的方法尚未明确,难以有效解决各分项生物量与地上总生物量模型的相容性,因此,森林生物量精准估测亟需系统研究。为此，本项目拟在中南林业科技大学芦头实验林场开展青冈栎次生林地面激光数据观测，揭示复杂林分环境条件下林木三维点云的有效分离机制及数据简化方法，提出树冠自遮蔽部分树干缺失点云修复和树干三维曲面模拟方法，提高树干、树冠三维曲面模拟的可信度;构建森林生物量精准估测模型，解决由扫描站点数量、位置及点云密度的不确定性对生物量估测所带来的预估偏差问题。研究结果可望改进地面激光雷达生物量建模方法,提高生物量计量精确度。为从样地尺度到区域生物量的精准估测提供理论依据。

Accurate estimation of forest biomass is an important research topic in forest resources monitoring. The existing studies mainly focus on the use of field observations to estimate forest biomass, but this is time-consuming, labor-intensive and costly. Terrestrial laser scanning (TLS) technology used to collect information of individual trees and forest stands provides a practical solution to accurately estimate stand biomass. However, the TLS method for biomass estimation is currently characterized by several defects, including: 1) It is difficult to automatically and accurately separate massive point clouds data into the components of tree stem, branches and leaves; 2) There is a lack of methods available for using TLS clouds data to reconstruct the missing point clouds of tree components and conduct 3-dimentional (3D) surface simulation; and 3) It is unknown what are the major sources of uncertainties for estimates of tree total aboveground biomass and its components. Therefore, this project would focus on seeking effective solutions for above three scientific research questions: 1) developing a simplified and effective mechanism of automatically and accurately separating the TLS point clouds data into the components of tree stem, leaves and branches, which will improve the efficiency and accuracy of forest parameters extracted; 2) proposing an effective point clouds reconstruction method for tree stem surface models to improve the reliability of 3D surface simulations of tree stems and tree crowns; and 3) Developing additive and compatible biomass models of tree and its components and further conducting an error and uncertainty analysis of the estimates for stand aboveground biomass by exploring the effects of scanning station numbers, locations and densities of point clouds on the estimation accuracies. The methods will be tested in the Lutou Experimental Forest Farm of Central South University of Forestry & Technology located in PingJiang County of Hunan Province, where a set of sample plots of secondary forest of Cyclobalanopsis glauca will be selected based on stand ages and structures and measured in the field using a traditional tree tally forest inventory method and their point clouds data will be acquired using TLS. It is expected that this research will provide the great potential to improve the accuracies of stand aboveground biomass and its components and thus will offer a theoretical basis for accurate estimation of stand biomass.