Quantitatively characterizing the non-random spatial distributions of foliage elements including coniferous needles is critical to map the radiation regime and retrieve the biophysical parameters of a given forest canopy from three-dimensional (3-D) perspective. Different experimental setups bring various challenges to the process of retrieving forest canopy clumping index (CI) using terrestrial laser scanning (TLS). In this paper, through developing a voxel-based gap size (VGS) algorithm, we compared the TLS-based forest canopy CIs with the ones obtained using the digital hemispherical photography (DHP)-based and tracing radiation and architecture of canopy (TRAC)-based approaches. Moreover, we investigated the effects of incident directions of solar beams, voxel size, and woody canopy components on the final retrieval accuracy of forest canopy CIs. Our results showed that: (1) TLS-based CIs accounted for 81% (N = 30, p < 0.001) of variations in the DHP-based method. (2) the anisotropic nature of forest canopy CIs suggested that a relatively comprehensive TLS data of a forest canopy was required to investigate the 3-D spatial variations of forest gap size distributions and CIs. (3) The user-defined laser sampling spacing was a reliable reference value to determine the voxel size when using the VGS algorithm. (4) It was recommended to separate woody canopy components when computing the forest canopy CI, especially for forest plots with higher proportions of woody material. (5) The effects of the penumbra on TLS-based forest canopy CIs were much more limited compared with the traditional optical instruments (i.e., DHP or TRAC). This work provides a solid foundation to dramatically improve the retrieval accuracy of leaf area index (LAI) using TLS.

定量表征包括针叶树针在内的树叶元素的非随机空间分布对于绘制辐射状况并从三维（3-D）角度检索给定林冠层的生物物理参数至关重要。不同的实验设置给使用地面激光扫描（TLS）检索森林冠层结块指数（CI）的过程带来了各种挑战。在本文中，通过开发基于体素的间隙大小（VGS）算法，我们比较了基于TLS的森林冠层CI和使用基于数字半球摄影（DHP）并跟踪辐射和冠层架构（TRAC）获得的基于方法。此外，我们研究了太阳光束入射方向，体素大小，和木质冠层成分对森林冠层CI最终检索精度的影响。我们的结果表明：（1）基于TLS的配置项占81％（N  = 30，p <0.001）基于DHP的方法的变化。（2）森林冠层CI的各向异性性质表明，需要一个相对全面的森林冠层TLS数据来调查森林间隙大小分布和CI的3-D空间变化。（3）使用VGS算法时，用户定义的激光采样间隔是确定体素大小的可靠参考值。（4）在计算林冠CI时，建议将木质林冠成分分开，尤其是对于木质材料比例较高的林地。（5）与传统的光学仪器（如DHP或TRAC）相比，半影对基于TLS的森林冠层CI的影响要有限得多。这项工作为使用TLS显着提高叶面积索引（LAI）的检索准确性提供了坚实的基础。