联合“空-地”激光雷达定量分离山区林下植被的叶面积指数

The light detection and ranging (lidar) technology has been becoming a powerful tool due to its ability to directly acquiring the three dimensional (3-D) forest canopy structural information. However, it is still not good enough for the existing lidar-based method for retrieving forest canopy leaf area index (LAI) since they don’t comprehensively accounting the effects of the understory and topographic variations. This study aims to develop a generic lidar-based approach to remove the effects of the forest background including understory and bare earth in terms of the LAI estimation in a mountain area. Four aspects will be considered in this study: (1) we will first investigate the mechanism interactions and relationship between laser returned energy, 3-D forest structure, and topographic variations; (2) a generic method suitable for complex terrain will be developed to quantitatively retrieve the LAI of understory including the approach for retrieving directional gap fraction and its spatio-temporal variations; (3) a numerical model will be developed to retrieve the LAI for both over- and under-story, and their seasonal variations will be further explored; (4) three different validation approaches including the computer-based simulation, indoor forest scenario measurements, and field-based measurements of forest structural parameters will be used in the current study to test the suitability of the proposed model and validate its result. The proposed model for retrieving the LAI for over- and under-story will provide a powerful and effective tool for lidar-based precision forestry study.

由于具有直接获取三维结构信息的能力，激光雷达已逐渐成为森林冠层叶面积指数定量反演的有力工具，但现有的激光雷达叶面积指数反演方法对于地形起伏变化和林下植被的影响考虑还不够全面。为此，本项目将发展适用于复杂地形条件下的激光雷达定量分离林下植被并反演其叶面积指数的算法和模型，该模型主要包括2个核心部分：（1）激光光束回波、森林三维结构和地形起伏变化三者关系的定量表达方法；（2）复杂地形条件下激光雷达识别和分离森林冠层和林下植被的方法模型，主要包括森林方向性孔隙率的定量反演和时空变化 ；（3）复杂地形条件下基于激光雷达数据定量反演森林冠层和林下叶面积指数，并研究其季节变化规律；（4）利用计算机模拟、室内模拟观测和野外实际观测的方法对所发展的模型算法进行验证。所发展和建立的激光雷达定量分离和反演森林冠层和林下植被叶面积指数的方法，为提高激光雷达定量反演植被结构参数的可靠性提供有效工具。

本项目融合“空-地”激光雷达数据，在定量分析森林三维结构与激光雷达波形和点云数据之间关系的基础上，识别并定量分离了森林的林下植被，并进一步定量评估了林下植被对森林叶面积指数的影响。研究表明：（1）联合“空-地”激光雷达数据能够有效地改善森林植被激光雷达数据所获取数据的完整性，有利于定量刻画植被的三维结构特征；（2）植被冠层的三维结构特征对于激光雷达回波波形具有显著的影响，定量评估森林空间分布格局、森林单木高度分布，以及林下植被分布是森林叶面积指数反演具有重要意义；（3）联合“空-地”激光雷达和光学遥感数据识别和定量分离林下植被，可以有效改善中低郁闭度森林叶面积指数估算的精度。