地面激光雷达LiDAR可以快速获取高精度、高密度的点云数据,在植被结构参数获取方面的应用越来越广泛。为了定量分析地面激光雷达点云数据获取单木结构参数的能力和精度,本文提出利用光线跟踪结合植被真实结构模拟地面3维激光扫描仪的单木点云数据(以RIEGL VZ-1000为例),并结合间隙率模型反演单木叶面积指数LAI。在点云模拟过程中,充分考虑了脉冲特性,包括光斑大小、波束发射角以及回波探测强度。重点分析了光斑大小和最小探测强度对LAI反演的影响,并采用根河实测单木数据进行了验证。结果表明,光斑大小和最小探测强度的设定对于LAI反演结果存在很大影响,该结论对于提高地面激光雷达点云数据反演植被结构参数精度具有重要意义。

Terrestrial Laser Scanner (TLS) technology can quickly acquire three-dimensional information of targets with high precision. Given that TLS is a new data collection technique, it has been gradually applied to characterize the structural attributes of forest canopy. However, the inversion accuracy of Leaf Area Index (LAI) is highly dependent on the intrinsic configuration of the sensor, such as beam size and echo detection energy. In this paper, a computer simulation model was proposed to simulate point clouds from TLS and to analyze quantitatively the influence of beam characteristic on LAI inverted from TLS data.
A realistic tree was generated with OnyxTREE BROADLEAF software. Moreover, a computer model was proposed to simulate the interactions of lasers with a single tree and to acquire the point clouds from a TLS Riegl VZ-1000 based on the ray tracing algorithm. This model consisted of the ray intersection with triangular patches of photorealistic trees, the coordinate system conversion, and the acceleration of the algorithm. The beam size at exit, beam divergence, and echo detection algorithm were considered in the computer simulation method. One laser beam was divided into multiple bins, and each bin was treated as a separate pulse with its location, propagation direction, and an initial energy changing into a Gaussian shape. We inverted the crown-level Leaf Area Index (LAI) by using gap fraction analysis with the simulated point clouds, and the influence of beam characteristics (such as beam diameter and minimum echo detection intensity) on the LAI inversion was analyzed. Finally, we conducted the validation with the measured points of a birch tree located in Root River. We analyzed the influences of beam characteristics, such as beam size, beam divergence, and echo detection energy, on LAI inversion. The inversion results indicate that beam size and detection limit greatly influence LAI inversion. The points are increased with the decrease of the corresponding gap fraction because several points can be returned from one beam when the beam width and divergence were considered, particularly when significant differences are achieved at the edge of leaves. A larger beam size means that components in the edge portion are intercepted more easily. Thus, the deviation of LAI inversion would be greater. When the detection intensity threshold was small, echo information could be returned even if only part of the spot edge was intercepted. Thus, gap fraction is undervalued. However, when the energy threshold setting was large, the returned energy may be below the threshold value and cannot be recorded, thereby resulting in overestimation of the gap fraction and underestimation of LAI. Therefore, the points caused by beam size and echo detection must be filtered, and suitable points must be chosen before inverting LAI with the gap fraction model.
The simulation model based on the ray tracing algorithm was presented to explore the laser beam interceptions with an individual tree generated by using OnyxTree software. The LAI was retrieved via gap fraction analysis with zenith slicing method. The beam characteristics, such as beam size, echo detection energy, and beam divergence, were considered. The simulation model enables efficient and cost-effective research that can avoid environmental and instrumental error. This model contributes to an improved understanding of the intersections of laser beams with the tree crown well, and the LAI inversion of an individual tree is facilitated.