基于激光扫描单木结构的树间三维竞争模式研究：由有限测树因子到形态重构

Tree competition is a behavior synthetically triggered by a series of physiological and biochemical effects such as chlorophyll photosynthesis and carbon sequestration, respiration and transpiration, and soil nutrient absorption and cycle, and its study is of great significance for forest understanding, specifically on the topics such as tree cultivation, forest ecology and forest management. However, people have not yet achieved a comprehensive grasping of its internal triggering mechanisms and external functioning rules. This is mainly due to that the conventional devices for forest inventories can only offer limited kinds of tree measurement factors such as tree distance, and the resulting 2D competition relationships cannot fully characterize the competition effects between trees. To handle this problem, this project attempts to introduce multi-types of terrestrial laser scanning based remote sensing techniques, e.g. the backpack laser scanning system developed by the team, in order to acquire the 3D point clouds of the forests of interest in a flexible and efficient way. Then, the structures of individual trees and the morphologies of forest communities can be reconstructed, and their 3D structure feature parameters such as crown-layered projection area can be derived. Next, the 3D competition modes between each individual target tree and the trees around it in multi-layers will be explored, and the related models will be expanded to tree communities of both same and different tree species. Based on these works, it is expected that the study of tree competition can be improved into the 3D level, and the competition effects such as 3D allocation of tree biomass can be better explained. Overall, thru introducing the state-of-the-art laser scanning techniques into forest inventory as a potential breakthrough point of tree competition study, this project is dedicated to pushing forward the field of tree competition into 3D study level.

树间竞争是光合固碳、呼吸蒸腾、土壤养分吸收循环等生理生化过程的综合作用呈现，其研究对树种培育、森林生态、森林经理等林学广义领域的认知具有重要的意义。然而目前对其内在决定机理与外在表现机制的掌握尚未做到深入全面，这主要受制于传统的林业调查手段通常只能提供有限的测树因子如树间距离，由此建立的二维竞争关系无法全面揭示树间的竞争作用机制。针对这一问题，本课题将引入自主研发的背包式激光扫描系统为代表的多型地基激光遥感技术，通过其组合高效地获取森林群落的三维点云以实现形态重构，提取树木三维结构表征参数如树冠分层投影面积，探索并揭示对象木-多圈层竞争木、单木-群落尺度统一的种内与种间的三维竞争模式，以期实现将树间竞争机制的研究推进到三维层面，更好地解释如树木生物量三维分配等竞争效应。通过上述工作，力争藉由地基激光扫描这一新型林业调查技术的切入，推动我国在树间竞争方面的研究与国际保持同步。

树间竞争是光合固碳、呼吸蒸腾、土壤养分吸收循环等生理生化过程的综合作用呈现，其研究对树种培育、森林生态、森林经理等林学广义领域的认知具有重要的意义。然而目前对其内在决定机理与外在表现机制的掌握尚未做到深入全面，这主要受制于传统的林业调查手段通常只能提供有限的测树因子如树间距离，由此建立的二维竞争关系无法全面揭示树间的竞争作用机制，这也是本领域国际上共同关注的前沿主题之一。.本研究引入自主研发的背包式激光扫描系统为代表的多型地基激光遥感技术，通过其组合高效地获取森林群落的三维点云以实现形态重构，提取树木三维结构表征参数如树冠分层投影面积，探索并揭示对象木-多圈层竞争木、单木-群落尺度统一的种内与种间的三维竞争模式，以期实现将树间竞争机制的研究推进到三维层面，更好地解释如树木生物量三维分配等竞争效应，主要内容包括：（1）单木识别及形态表征参数提取算法开发；（2）树间竞争关系分析与竞争交互机制归纳建模；（3）面向单木-群落尺度统一的种内与种间三维竞争模式分析及新型竞争指标、模型建立。.在项目的支持下，我们探索开发了一系列原创性的解决方案和高效算法：（1）基于激光雷达的树木扫描数据处理及树间竞争互利分析新方案，涵盖树木识别、树木结构参数提取、树种分类和树间竞争互利分析新算法；（2）基于激光雷达分析树木竞争互利的森林生态科学增强认知新框架，涵盖树木生长习性、地形滤波、三维森林生态学新模式；（3）针对影响森林竞争模式的复杂环境及其驱动机制延伸探讨，涵盖树木易受影响的石漠化、沙漠化、雪驱动等新机制；（4）面向林木竞争互利分析深入的新型遥感技术拓展探索，涵盖偏振激光雷达、高光谱激光雷达等新技术。.项目组共发表SCI期刊论文15篇、中文核刊论文4篇、学术会议论文2篇，获批实用新型专利2项，撰写学术专著1部；项目支持下，课题组林沂研究员获得北京市科学技术奖技术发明奖一等奖（4/15，省部级）等学术奖励。