基于Beer-Lambert方法的机载LiDAR草地叶面积指数估算研究  

Estimating Grassland Leaf Area Index Using Airborne LiDAR with the Beer-Lambert Model

作　　者：李苏和 郭利彪 谭维贤[1,2] 黄平平 兰春阳[1,2] 吴雨晴 LI Suhe;GUO Libiao;TAN Weixian;HUANG Pingping;LAN Chunyang;WU Yuqing(School of Information Engineering,Inner Mongolia University of Technology,Hohhot 010080,China;Key Laboratory of Radar Technology and Application of Inner Mongolia Autonomous Region,Hohhot 010080,China)

机构地区：[1]内蒙古工业大学信息工程学院,内蒙古呼和浩特010080 [2]内蒙古自治区雷达技术与应用重点实验室,内蒙古呼和浩特010080

出　　处：《中国草地学报》2024年第9期43-53,共11页Chinese Journal of Grassland

基　　金：内蒙古教育厅基本科研业务费项目(JY20220198);内蒙古自然科学基金面上项目(2021LHMS04002);国家自然科学基金联合基金项目(U22A2010)。

摘　　要：草地叶面积指数(Leaf area index,LAI)是天然草地的重要结构参数,能够用来监测草地的生长状况和生产力水平,对草地资源可持续利用和科学管理具有重要意义。以内蒙古锡林郭勒盟典型草原为研究对象,首先使用无人机激光雷达(Airborne light detection and ranging,Air-LiDAR)草地冠层观测数据,通过解析点云数据构建冠层高度模型(Canopy height model,CHM),随后进行研究区草地冠层间隙率计算,最后基于Beer-Lambert方法进行0.05 m、0.10 m、0.15 m、0.20 m 4个不同空间分辨率采样尺度下的LAI估算,并选择CHM低值、中值、高值3个不同子区域分别进行不同冠层高度下LAI的检验和分析。结果表明:(1)草地叶面积指数与冠层高度模型数值呈正相关、与冠层间隙率呈负相关。(2)机载LiDAR草地LAI估算的最优采样尺度为0.15 m,CHM不同高度子区域LAI结果检验R^(2)和RMSE分别为:低值区为0.66和0.04;中值区为0.54和0.34;高值区为0.54和1.17,表明无人机LiDAR可捕获草地冠层观测采样存在的异质性差异分布特征。(3)不同空间分辨率0.05~0.20 m间隔采样LAI结果表明,对于CHM低值、植被分布稀疏区域不同分辨率LAI无显著空间尺度变化差异,但CHM高值、较密植被分布群落LAI会随空间分辨率表现出尺度性差异。综上所述,本研究设计完成的无人机LiDAR草地LAI估算模型,参数机理具体、流程方法可操作性强,具有较好的数值检验精度,可为激光雷达在草地植被叶面积指数遥感反演及应用提供技术参考。Grassland leaf area index(LAI)is a key structural parameter for monitoring grassland ecosystem.The temporal growth patterns and spatial distribution of natural grass vegetation are significantly influenced by regional meteorological and grazing practices,posing challenges in accurately obtaining ecological parameters specific to grassland.The study presented a novel approach for estimating grassland LAI using airborne LiDAR data and the Beer-Lambert method.The study area was the typical grassland located in the Xilin Gol League of Inner Mongolia,China.First,the grassland canopy height model(CHM)was reconstructed using airborne LiDAR cloud point data.Next,the grassland community canopy gap fraction(P_(gap))was calculated as an indicator of grass volume density distribution indicator relative to the CHM.Finally,using the Beer-Lambert method,LAI was estimated at four different spatial resolution sampling scales,distributed across three different canopy sub-regions(lower,medium,and higher).The results showed that:(1)A strong positive correlation was found between LAI and CHM values,while a negative correlation was observed with P_(gap).This alignment indicated that the model effectively captured the actual grass volume density and spatial distribution.(2)The optimal spatial sampling scale for grassland LAI estimation using airborne LiDAR was determined to be 0.15 m.The R^(2) values for LAI estimation within different CHM sub-regions ranged from 0.54 to 0.66,and the RMSE ranged from 0.04 to 1.17.These results demonstrated that the ability of airborne LiDAR to capture the heterogeneity distribution characteristics of grassland communities.(3)While grassland LAI estimates were relatively insensitive to spatial sampling scale within the lower CHM sub-regions,variations were observed at higher CHM and denser vegetation areas.In summary,the developed grassland LAI estimation model,based on airborne LiDAR and the Beer-Lambert model,provides an accurate and operational sensing technique for grassland ecosystem monitoring.Thi

关 键 词：草地植被 无人机激光雷达 间隙率 叶面积指数 空间异质性 

分 类 号：P237[天文地球—摄影测量与遥感]