基于机载激光雷达的亚热带森林生物量估算模型研究  

Estimation Model of Subtropical Forest Biomass Based on Airborne Lidar

作　　者：方志良 陈奇 任引[3] 王艳军 FANG Zhiliang;CHEN Qi;REN Yin;WANG Yanjun(National-Local Joint Engineering Laboratory of Geo-spatial Information Technology, Hunan University of Science and Technology, Xiangtan, Hunan 411100, China;Department of Geography & Environment, University of Hawai′i at Mānoa, Honolulu 96822, USA;Key Laboratory of Urban Environment and Health, Institute of Urban Environment, Chinese Academy of Sciences, Xiamen, Fujian 361021, China)

机构地区：[1]湖南科技大学地理空间信息技术国家地方联合工程实验室,湖南湘潭411100 [2]夏威夷大学马诺阿分校地理与环境系,美国檀香山96822 [3]中国科学院城市环境研究所城市环境与健康重点实验室,福建厦门361021

出　　处：《林业调查规划》2021年第6期1-8,共8页Forest Inventory and Planning

基　　金：国家自然科学基金(41771462).

摘　　要：利用机载激光雷达数据和地面调查数据研究和分析适合亚热带地区福建省龙岩市的森林生物量估算模型及其差异性。对比分析4种异速生长方程的树木生物量估算结果,结合机载激光雷达数据提取的15个特征变量(包括高度均值、方差、峰度、分位数、偏度和垂直高度百分比),构建了基于逐步回归、支持向量回归、随机森林和BP神经网络的森林生物量估算模型,分析不同的异速生长方程对森林生物量估算结果的差异性影响。结果表明,分省分树种的异速生长方程在样地尺度的生物量估算结果较准确,基于机载激光雷达的森林生物量估算模型中,随机森林方法精度较高。在整个森林生物量估算中,简单分析生物量估算模型误差的大小并不能说明生物量估计的准确程度,异速生长方程的误差需要更多地被考虑到森林生物量估算中。The forest biomass estimation model suitable for subtropical Longyan City,Fujian Province was analyzed by airborne lidar data and ground-based lidar data.The tree biomass estimation results of four different allometric equations were compared.Combined with 15 characteristic variables(including height mean,variance,kurtosis,quantile,skewness and percentage of vertical height)extracted from airborne lidar data,a forest biomass estimation model based on stepwise regression,support vector regression,random forest and BP neural network was constructed,and the different effects of different allometric equations on the estimation results of forest biomass were analyzed.The results showed that the allometric equation of tree species by province was more accurate in the biomass estimation results of sample plot scale,and the random forest method had higher accuracy in the forest biomass estimation model based on airborne lidar.In the whole forest biomass estimation,simple analysis of the error size of biomass estimation model could not explain the accuracy of biomass estimation.The error of allometric equation needed to be more considered in forest biomass estimation.

关 键 词：森林生物量 估算模型 机载激光雷达 异速生长方程 亚热带地区 

分 类 号：S718.556[农业科学—林学] V556.7[航空宇航科学与技术—人机与环境工程]