单立木三维(3-D)重建与全面解析叶面积指数(LAI)的空间分布对于认识我国南方林下土壤侵蚀过程与防治具有重要意义。本研究选择江西省东部德兴市新岗山丘陵区的中国–德国–瑞士三方合作生物多样性试验基地，使用地面激光扫描仪(TLS)，在野外实地通过扫描获取一棵5年生杉木的点云数据，并基于点云数据反演了该杉木LAI的空间分布，讨论了冠层结构及其异质性对林下土壤侵蚀的影响。结果表明：地面激光扫描仪是单立木3-D重建的有力工具，本研究提出的提取LAI径向分布的体元模型切分方式行之有效。样木LAI在纵向及径向上分布皆不均匀，纵向上，LAI(ΔH=0.1m)变化范围为0.02 ~ 0.18 m²/m²；径向上，树干处的LAI可达到树冠边缘处的8 ~ 25倍。东南SE、东北NE、西北NW、西南SW四个方位的LAI分布表现为下坡向(SW、NW)大于上坡向(SE、NE)。整棵杉木的LAI为2.45 m²/m²，TLS应用于林木野外扫描时受风速的影响比较大，基于体元模型计算LAI时，体元尺寸的选择非常关键。

Building the 3-D model of individual tree and analyzing completely the spatial distribution of leaf area index (LAI) are of great significance to understand the process and prevention of soil erosion under the forest in the south of China. A case study was conducted in the experiment site of the joint Chinese-German-Swiss research project “BEF-China”, which was located in the hilly region of the eastern part of Jiangxi Province, China. Initially, the point cloud data of a 5-year-old Chinese fir tree was obtained by using terrestrial laser scanner (TLS), then the spatial distribution of LAI was extracted and the impact of crown structure, its spatial heterogeneity on soil erosion under the forest was analyzed. The results showed that TLS was a powerful tool for reconstructing the 3-D model of individual trees. An effective segmentation algorithm to extract the radial distribution of canopy elements was presented. The spatial distribution of LAI of the whole tree was uneven: LAI (ΔH=0.1m) ranged from 0.02 to 0.18 m²/m² in the vertical direction, while radially LAI values, close to the trunk, could reach 8 – 25 times of those on the crown edge. LAI in the downhill was greater than that in the uphill. LAI of the whole fir tree was 2.45 m²/m². In addition, it was found that TLS was sensitive to wind in our field measurement and the voxel size was a crucial factor in predicting LAI with the voxel-based method.

S157.1

中国科学院知识创新工程重要方面项目群(KZCX2-YW-Q1-07和15)资助