机载LiDAR航带旁向重叠对针叶林结构参数估测的影响

doi:10.11707/j.1001-7488.20180613

摘要/Abstract

摘要： [目的]研究机载LiDAR航带旁向重叠对针叶林林分平均高和森林叶面积指数（LAI）估测的影响，为机载LiDAR点云数据区域森林结构参数估测提供参考。[方法]野外分别测定30块樟子松、33块长白落叶松样地的林分平均高和LAI，对原始LiDAR点云数据进行去噪、点云分类、高程归一化和重叠点移除等处理，从重叠点移除前、重叠点和重叠点移除后的点云数据中分别提取一系列样方点云高度分位数（H_P1、H_P5、H_P10、…、H_P99、H_max和H_mean）和激光穿透指数（LPI），借助留一交叉验证建立并评价樟子松和长白落叶松林分平均高和LAI估测模型的精度，通过对比分析估测模型的决定系数（R²）和均方根误差（RMSE）揭示机载LiDAR航带旁向重叠对针叶林林分平均高和LAI估测的影响。[结果]对樟子松林分平均高估测而言，重叠点移除前林分平均高的最高估测精度（R²=0.873，RMSE=0.940）出现在H_P90处，重叠点林分平均高的最高估测精度（R²=0.892，RMSE=0.866）出现在H_P80处，而重叠点移除后林分平均高的最高估测精度（R²=0.892，RMSE=0.868）出现在H_P55处；对长白落叶松林分平均高估测而言，重叠点移除前、重叠点和重叠点移除后林分平均高的最高估测精度均出现在H_P99处，R²分别为0.725、0.719和0.741，RMSE分别为1.196、1.209和1.161。对樟子松LAI估测而言，重叠点移除前估测结果R²为0.666，RMSE为0.220，重叠点估测结果R²为0.551，RMSE为0.255，重叠点移除后R²提高到0.794，RMSE降低为0.172；对长白落叶松LAI估测而言，重叠点移除前估测结果R²为0.654，RMSE为0.110，重叠点估测结果R²为0.640，RMSE为0.112，与樟子松估测结果一致，重叠点移除后长白落叶松LAI估测精度大幅度提高，R²变为0.762，RMSE变为0.091。[结论]无论是林分平均高还是森林LAI，相邻航带旁向重叠点移除后的估测精度均高于重叠点移除前和重叠点，且樟子松的估测精度高于长白落叶松。对林分平均高而言，樟子松和长白落叶松达到最高估测精度时所对应的点云高度分位数不同。机载LiDAR点云数据相邻航带旁向重叠点的移除可有效提高森林结构参数的估测精度，在未来机载LiDAR点云数据预处理时应加入重叠点移除操作。

关键词: 机载激光雷达, 航带旁向重叠, 针叶林, 平均树高, 叶面积指数

Abstract: [Objective] In order to provide advice for the research of forest structural parameters estimation with airborne LiDAR data, the effect of side-overlap between adjacent swaths on forest stand mean height and leaf area index estimation was studied in this paper.[Method] In this study, the forest stand mean height and leaf area index from 30 field plots of scotch pine and 33 field plots of larch pine, were measured. Firstly, the raw LiDAR data were processed through a series of procedures including noise points removal, point cloud classification, elevation normalization and side-overlap points cut off. Then the quantile heights of LiDAR points (H_P1, H_P5, H_P10, …, H_P99, H_max and H_mean) and laser penetration index (LPI), extracted from the LiDAR data with overlap points, overlap points and LiDAR data without overlap points, were used to establish and assess the model accuracy of forest stand mean height and LAI estimation by using the leave one out cross validation. The coefficient of determination (R²) and root mean square error (RMSE) were used to assess the effect of side-overlap between LiDAR adjacent swaths on the estimation of forest stand mean height and LAI.[Result] For forest stand mean height estimation, the best result of scotch pine (R²=0.873, RMSE=0.940) from LiDAR data with overlap points was obtained with H_P90. The best result (R²=0.892, RMSE=0.866) from overlap points was achieved with H_P80. And the best result (R²=0.892, RMSE=0.868) from LiDAR data without overlap points was obtained at H_P55. For larch pine, all the best result were achieved with H_P99 for LiDAR data with overlap points, overlap points and LiDAR data without overlap points. The R² were 0.725, 0.719 and 0.741, and RMSE were 1.196, 1.209 and 1.161 respectively. For forest LAI estimation, the R² and RMSE of scotch pine were 0.666 and 0.220 when the overlap points included in LiDAR data. The R² and RMSE from overlap points were 0.551 and 0.255. In addition, the R² was improved to 0.794 and RMSE was induced to 0.172 after the overlap points were removed from LiDAR data. For larch pine, the R² and RMSE were 0.654 and 0.110 when the overlap points were included in LiDAR data. The R² and RMSE were 0.640 and 0.112 when the overlap points were used. While, the R² improved to 0.762 and RMSE was 0.091 when the overlap points were removed from LiDAR data.[Conclusion] Both for forest stand mean height and LAI estimation, the result acquired from LiDAR data without overlap points were better than the result obtained from LiDAR data with overlap points. Additionally, the result of scotch pine were better than the result of larch pine. For forest stand mean height estimation, the best result from scotch pine and larch pine achieved with different quantile height parameters. The removal of overlap points between LiDAR adjacent swaths could effectively improve the estimation accuracy of forest structural parameters. Therefore, the overlap cut off should be included in the pre-processing of airborne LiDAR point cloud data to improve the forest structural parameters estimation accuracy in the future research.

Key words: airborne LiDAR, side-overlap between adjacent swaths, coniferous forest, mean tree height, leaf area index(LAI)

中图分类号:

S757
TP79

尤号田, 邢艳秋, 彭涛, 丁建华. 机载LiDAR航带旁向重叠对针叶林结构参数估测的影响[J]. 林业科学, 2018, 54(6): 109-118.

You Haotian, Xing Yanqiu, Peng Tao, Ding Jianhua. Research on the Effect of Side-Overlap between Airborne LiDAR Adjacent Swaths on the Coniferous Forest Structural Parameters Estimation[J]. Scientia Silvae Sinicae, 2018, 54(6): 109-118.

参考文献

焦义涛,邢艳秋,霍达,等. 2015. 基于机载LiDAR点云估测林分的平均树高. 西北林学院学报,30(3):170-174.
(Jiao Y T,Xing Y Q,Huo D,et al. 2015. Study on mean canopy height estimation from airborne LiDAR point cloud data. Journal of Northwest Forestry University,30(3):170-174.[in Chinese])
骆社周,王成,张贵宾,等. 2013. 机载激光雷达森林叶面积指数反演研究. 地球物理学报,56(5):1467-1475.
(Luo S Z,Wang C,Zhang G B,et al. 2013. Forest leaf area index (LAI) inversion using airborne LiDAR data. Chinese Journal of Geophysics,56(5):1467-1475.[in Chinese])
穆喜云,张秋良,刘清旺,等. 2015. 基于机载激光雷达的寒温带典型森林高度制图研究. 北京林业大学学报,37(7):58-67.
(Mu X Y,Zhang Q L,Liu Q W,et al. 2015. Typical forest height mapping in cold temperate zone using airborne LiDAR data. Journal of Beijing Forestry University,37(7):58-67.[in Chinese])
庞勇,李增元,陈尔学,等. 2005. 激光雷达技术及其在林业上的应用. 林业科学,41(3):129-136.
(Pang Y,Zhao F,Li Z Y,et al. 2005. LiDAR remote sensing technology and its application in forestry. Scientia Silvae Sinicae, 41(3):129-136.[in Chinese])
庞勇,赵峰,李增元,等.2008. 机载激光雷达平均树高提取研究. 遥感学报,12(1):152-158.
(Pang Y,Li Z Y,Chen E X,et al. 2008. Forest height inversion using airborne LiDAR technology. Journal of Remote Sensing,12(1):152-158.[in Chinese])
王兵,任晓旭,胡文. 2011. 中国森林生态系统服务功能及其价值评估. 林业科学,47(2):145-153.
(Wang B,Ren X X,Hu W.2011. Assessment of forest ecosystem services value in China. Scientia Silvae Sinicae,47(2):145-153.[in Chinese])
邢艳秋,霍达,尤号田,等. 2016. 基于机载LiDAR单束激光穿透指数的白桦林LAI估测. 应用生态学报,27(11):3469-3478.
(Xing Y Q,Huo D,You H T,et al. 2016. Estimation of birch forest LAI based on single laser penetration index of airborne LiDAR data. Chinese Journal of Applied Ecology,27(11):3469-3478.[in Chinese])
尹艳豹,唐守正,郎璞梅,等. 2011. 多回波机载LiDAR数据提取林地DEM的判别分析方法. 林业科学,47(12):106-113.
(Yin Y B,Tang S Z,Lang P M,et al. 2011. Determination of DEM based on multiple-echo data of airborne LiDAR in mountainous wooded area using discriminant analysis. Scientia Silvae Sinicae,47(12):106-113.[in Chinese])
尤号田, 邢艳秋, 冉慧,等. 2014a. 基于LiDAR点云能量信息的樟子松郁闭度反演方法. 北京林业大学学报,36(6):30-35.
(You H T,Xing Y Q,Ran H,et al. 2014a. Inversion method for the crown density of Mongolian scotch pine from point cloud data of small-footprint LiDAR. Journal of Beijing Forestry University, 36(6):30-35.[in Chinese])
尤号田,邢艳秋,王铮,等. 2014b. 点云密度对激光雷达估计森林样方平均树高的影响. 东北林业大学学报,42(5):143-148.
(You H T,Xing Y Q,Wang Z,et al. 2014b. Effects of LiDAR point density on tree height estimation in plots level. Journal of Northeast Forestry University,42(5):143-148.[in Chinese])
张瑞英,庞勇,李增元,等.2016. 结合机载LiDAR和LANDSAT ETM+数据的温带森林郁闭度估测. 植物生态学报,40(2):102-115.
(Zhang R Y,Pang Y,Li Z Y,et al. 2016. Canopy closure estimation in a temperate forest using airborne LiDAR and LANDSAT ETM+ data. Chinese Journal of Plant Ecology,40(2):102-115.[in Chinese])
Austin M P,Meyers J A. 1996. Current approaches to modelling the environmental niche of eucalypts:implication for management of forest biodiversity. Forest Ecology and Management,85(1/3):95-106.
Babcock C,Finley A O,Cook B D,et al. 2016.Modeling forest biomass and growth:coupling long-term inventory and LiDAR data. Remote Sensing of Environment,182:1-2.
Ellsworth D S,Reich P B. 1993. Canopy structure and vertical patterns of photosynthesis and related leaf traits in a deciduous forest. Oecologia,96(2):169-178.
Hansen A J,Neilson R P,Dale V H,et al. 2001. Global change in forests:responses of species,communities,and biomes interactions between climate change and land use are projected to cause large shifts in biodiversity. BioScience,51(9):765-779.
Jensen J L,Humes K S,Vierling L A,et al. 2008. Discrete return lidar-based prediction of leaf area index in two conifer forests. Remote Sensing of Environment,112(10):3947-3957.
Korhonen L,Korpela I,Heiskanen J,et al. 2011. Airborne discrete-return LiDAR data in the estimation of vertical canopy cover, angular canopy closure and leaf area index. Remote Sensing of Environment,115(4):1065-1080.
Luo S Z,Wang C,Zhang G B,et al. 2013. Forest leaf area index (LAI) estimation using airborne discrete-return LiDAR data. Chinese Journal of Geophysics,56(3):233-242.
Naesset E.1997. Determination of mean tree height of forest stands using airborne laser scanner data. ISPRS Journal of Photogrammetry and Remote Sensing,52(2):49-56.
Olthof I,King D J. 2000.Development of a forest health index using multispectral airborne digital camera imagery. Canadian Journal of Remote Sensing,26(3):166-176.
Solberg S. 2008. Comparing discrete echoes counts and intensity sums from ALS for estimating forest LAI and gap fraction//Proceedings of SilviLaser 2008:8^th International Conference on LiDAR Applications in Forest Assessment and Inventory. Edinburgh,UK,446-455.
Solberg S. 2010. Mapping gap fraction, LAI and defoliation using various ALS penetration variables. International Journal of Remote Sensing,31(5):1227-1244.
Sumnall M J,Fox T R,Wynne R H,et al. 2016. Estimating leaf area index at multiple heights within the understorey component of Loblolly pine forests from airborne discrete-return lidar. International Journal of Remote Sensing,37(1):78-99.
Woodgate W,Mathias D,John D A,et al. 2015. An improved theoretical model of canopy gap probability for leaf area index estimation in woody ecosystems. Forest Ecology and Management,358:303-320.

[1]	蒋铮, 于倩楠, 乔明锋, 肖娟, 张子良, 尹华军. 云杉幼树根系分泌物对2种草本植物种子萌发和幼苗生长的影响[J]. 林业科学, 2019, 55(6): 160-166.
[2]	刘海轩, 许丽娟, 吴鞠, 徐程扬. 城市森林降温效应影响因素研究进展[J]. 林业科学, 2019, 55(4): 144-151.
[3]	耿林, 李明泽, 范文义, 王斌. 基于机载LiDAR的单木结构参数及林分有效冠的提取[J]. 林业科学, 2018, 54(7): 62-72.
[4]	李欢, 李明泽, 范文义, 王斌. 基于机载激光雷达的林隙结构参数提取[J]. 林业科学, 2018, 54(10): 98-107.
[5]	乔沙沙, 周永娜, 刘晋仙, 景炬辉, 贾彤, 李毳, 杨欣, 柴宝峰. 关帝山针叶林土壤细菌群落结构特征[J]. 林业科学, 2017, 53(2): 89-99.
[6]	刘志理, 金光泽. 光学仪器法测定叶面积指数季节变化的误差分析[J]. 林业科学, 2016, 52(9): 11-21.
[7]	段祝庚, 肖化顺, 袁伟湘. 基于离散点云数据的森林冠层高度模型插值方法[J]. 林业科学, 2016, 52(9): 86-94.
[8]	王振海, 殷秀琴, 张成蒙. 土壤动物在长白山臭冷杉凋落物分解中的作用[J]. 林业科学, 2016, 52(7): 59-67.
[9]	杨承栋. 我国人工林土壤有机质的量和质下降是制约林木生长的关键因子[J]. 林业科学, 2016, 52(12): 1-12.
[10]	周靖靖, 赵忠, 刘金良, 赵君, 赵青侠, 刘俊. 不同方法提取的快鸟影像信息估算刺槐林有效叶面积指数的精度比较[J]. 林业科学, 2015, 51(9): 24-34.
[11]	王会霞, 王彦辉, 杨佳, 谢滨泽, 石辉. 不同绿化树种滞留PM_2.5等颗粒污染物能力的多尺度比较[J]. 林业科学, 2015, 51(7): 9-20.
[12]	张俊艳, 成克武, 臧润国, 丁易. 海南岛热带针-阔叶林交错区群落环境特征[J]. 林业科学, 2014, 50(8): 1-6.
[13]	曹林, 代劲松, 庞勇, 赵兵, 徐建新, 李增元. 集成LiDAR和辐射模型的植被遮挡下城市屋顶太阳能估算[J]. 林业科学, 2014, 50(2): 99-110.
[14]	刘志理, 戚玉娇, 金光泽. 小兴安岭谷地云冷杉林叶面积指数的季节动态及空间格局[J]. 林业科学, 2013, 49(8): 58-64.
[15]	林杰;张金池;顾哲衍;吴玉敏. 基于叶面积指数的植被覆盖管理措施因子C的遥感定量估算[J]. , 2013, 49(2): 86-92.