摘要
激光雷达扫描技术是森林参数清查的新型重要手段,无人机激光扫描(ULS)与手持移动激光扫描(HMLS)可分别从空中与地面视角获取森林三维信息,在关键参数测量上优势互补,但其单一应用仍存在局限:ULS对胸径(DBH)估测误差较大,HMLS对树高(TH)的提取精度不足。因此,融合多源激光雷达数据以提升参数提取精度,进而提高森林地上生物量(AGB)估算的可靠性,已成为当前研究的重点。本研究以杉木人工林为对象,综合利用ULS、HMLS及其融合数据(ULS+HMLS)进行单木结构参数提取与AGB估算,系统分析不同数据源在单木及样方尺度的表现。结果表明:1)ULS在TH提取上表现最优(R^(2)=0.98,RMSE=0.25 m),HMLS在DBH提取中精度最高(R^(2)=0.98,RMSE=1.20 cm),融合数据在两项参数上均保持了较高精度。2)单木AGB估算精度受DBH主导,HMLS与融合数据表现更优(R^(2)均为0.98)。3)在样方尺度,融合数据AGB估算精度最高(R^(2)=0.92,RMSE=11.07 t·hm^(-2)),ULS因单木分割完整性不足导致误差累积,精度相对较低。研究证实,融合多源激光雷达数据能够有效结合不同平台的观测优势,为提升人工林生物量估算精度与可靠性提供有效技术途径。
LiDAR scanning technology has emerged as a significant novel tool for forest parameter inventory.Unmanned aerial vehicle-based laser scanning(ULS)and handheld mobile laser scanning(HMLS)acquire forest structural information from complementary aerial and ground perspectives,offering distinct advantages in measuring key parameters like tree height(TH)and diameter at breast height(DBH).However,each platform has inherent limitations:ULS often exhibits larger errors in DBH estimation,while HMLS provides insufficient accuracy in TH retrieval due to restricted viewing geometry.Therefore,integrating multi-source LiDAR data to improve the accuracy of key stand parameter extraction and enhance the reliability of aboveground biomass(AGB)estimation has become a research priority.This study focused on Chinese fir plantations,utilizing ULS,HMLS,and their fused data(ULS+HMLS)for individual tree structural parameter extraction and AGB estimation,systematically evaluating the performance of different data sources at both individual tree and plot scales.The results indicate that,1)ULS achieved the best performance in TH estimation(R^(2)=0.98,RMSE=0.25 m),while HMLS provided the highest accuracy for DBH retrieval(R^(2)=0.98,RMSE=1.20 cm).The fused data maintained high accuracy for both parameters.2)Individual tree AGB estimation accuracy was dominated by DBH precision,with HMLS and the fused data performing better(both R^(2)=0.98).3)At the plot scale,the fused data yielded the highest AGB estimation accuracy(R^(2)=0.92,RMSE=11.07 t·hm^(-2)).ULS showed relatively lower accuracy due to error accumulation stemming from incomplete individual tree segmentation.This study indicates that fusing multi-source LiDAR data effectively combines the observational strengths of different platforms can provide a robust technical approach for improving the accuracy and reliability of biomass estimation in plantations.
作者
杨博文
曾宏达
方艺辉
张惠光
熊景峰
张晓萍
YANG Bowen;ZENG Hongda;FANG Yihui;ZHANG Huiguang;XIONG Jingfeng;ZHANG Xiaoping(Institute of Geography,School of Carbon Neutrality Future Technology,Fujian Normal University,Fuzhou 350117,China;Key Laboratory for Humid Subtropical Eco-geographical Processes of the Ministry of Education,School of Carbon Neutrality Future Technology,Fujian Normal University,Fuzhou 350117,China;School of Geographical Sciences,School of Carbon Neutrality Future Technology,Fujian Normal University,Fuzhou 350117,China;Fujian Forest Investigation and Planning Institute,Chief Engineer Office,Fuzhou 350003,China)
出处
《亚热带资源与环境学报》
2026年第1期165-175,共11页
Journal of Subtropical Resources and Environment
基金
福建省林业科技攻关项目(2026FKJ1)
福建省公益类科研院所项目(2023R1002001)
福建省教育厅项目(K202300690)。
