QQ客服 官方微博 反馈留言

题名	基于GLAS 大光斑激光波形数据的山地森林高度反演
作者	马利群
学位类别	硕士
答辩日期	2011-05-14
授予单位	中国科学院研究生院
授予地点	北京
导师	李爱农
关键词	GLAS 大光斑 山地 森林高度 反演
其他题名	Inversion of forest height in mountainous region based on the GLAS Full-waveform data
学位专业	地图学与地理信息系统
中文摘要	森林参数是进行陆地生态系统碳循环和碳动态分析的基础，是生态学和全球变化研究的重要内容之一，而森林高度则是重要的森林参数之一。森林高度是描述森林垂直属性的重要信息，它与生物量等参数都存在着关联，这种生物约束机制为利用激光雷达反演的高度信息进行森林生物量建模提供了理论基础。传统的调查方法很难快速、准确、大面积、获取这些关键信息。激光测距雷达技术为快速获取空间信息提供了简单有效的手段。激光测距雷达LiDAR（Light Detection And Ranging）技术，特别是大光斑LiDAR技术为大面积获取森林垂直三维信息提供了可能。美国航空航天局（NASA）于2003年初发射了星载激光测高卫星ICESat（Ice，cloud，and land elevation satellite ），GLAS（Geoscience Laser Altimeter System waveforms）是搭载在卫星上的激光测高系统，该数据已经被成功应用于森林高度制图和陆面生物量反演。GLAS用波形记录激光信号，具有较强标识树冠结构的能力，精确描述大面积树冠信息的能力。本研究以中国西南地区为研究区，选择西南地区的森林生态系统为研究对象，利用GLAS数据、DEM数据、土地利用等数据源，在GLAS激光波形数据预处理的基础上，采用粒子群算法和最小二乘算法相结合的方法对其进行高斯分解，获取波形特征参数，结合波形数据的垂直分辨率对西南地区的森林高度进行了反演。同时本文根据山地特殊的地形特征，结合野外实地测量和已有数据集，定量评估地形对GLAS反演森林高度的影响，对森林高度反演结果进行地形校正。本文最后分析了西南森林高度空间特征及其与影响因子之间的关系。本研究第一次利用遥感的方法获取了西南山地林地高分辨率（30m）高度图，是进一步估算森林生态系统的材积蓄积量、陆上生物量、碳储量以及森林生态水文和动物栖息地等研究的重要数据基础。结果还有待进一步结合地面调查数据进行验证。本文的主要创新点有：（1）采用了阈值和连续多帧大于背景噪声阈值的方法确定信号起始位置；（2）采用粒子群算法和最小二乘法结合的方法对GLAS波形数据进行高斯分解；（3）根据西南地区有效激光点对中国西南地区的森林高度进行了反演，并对西南地区森林高度空间分布及其影响因子进行了分析。
英文摘要	Forest parameters are basement of analyzing carbon flux of terrestrial ecosystem and one of important contents of ecology and global climatic change. However, tree height is the important parameter. Tree height is crucial property describing forest vertical structures. Also, biomass, basal area, leaf area and so on have confidential relation with three heights. With the increasing of tree height, biomass and diameter can increase. This kind of tied mechanism provides theoretical principle for modeling forest biomass by use of tree height derived from LiDAR data. It is hard for the traditional forest surveying methods to derive information quickly, accurately, large acreage. LiDAR (Light Detection and Ranging) can get coordinates in space accurately and quickly, which is effective method to get space information. Especially Large footprint LiDAR facilitates deriving forest vertical structure extensively. The Geoscience Laser Altimeter System (GLAS) instrument aboard the Ice, Cloud, and Land Elevation (ICESat) satellite, launched on 12 January 2003. GLAS is the first LiDAR instrument designed for continuous global observation of the Earth. Researchers have started using GLAS data for forest studies and have applied to map forest height and inverse biomass. Measurements derived from GLAS waveforms were used to characterize the canopy vertical structure, and describing tree crown. This research based on forest ecological system in southwest of China. The paper estimated tree height by combined GLAS data, DEM, Landuse data, field data and meteorological data. The GLAS waveform has 544 bins with a bin size of 1ns or 15cm. This paper decomposed GLAS data based on the method of PSO (Particle Swarm Optimization, PSO) and Nonlinear Least square Method to gain waveform parameter and we can inverse tree height of southwest. Also, the footprint of GLAS is large and terrain in southwest of China is complex. The terrain can affect the result of research and remove or weaken the terrain influence on tree height and the paper estimated the terrain effect. Last, the research analyzed space distribution and factor of influence of tree height in southwest. The high-resolution of tree height distribution was gained firstly and it was 30m.This research provides technical support for getting forest structure in southwest and is easy to describing forest vertical information. The result of study has very important significance for estimating forest biomass, carbon storage, and hydrological role of forests and species diversity. However, the result of the research is to be validated. The novelties of the paper were (1) signal start and signal end were defined by threshold and several consecutive bins was greater threshold. (2) GLAS data were decomposed combined by PSO and Nonlinear Least square Method. (3) Tree height distribution of southwest was gained firstly and analyzed factor of influence of tree height.
语种	中文
公开日期	2013-01-24
内容类型	学位论文
源URL	[http://192.168.143.20:8080/handle/131551/4841]
专题	成都山地灾害与环境研究所_数字山地与遥感应用中心
推荐引用方式 GB/T 7714	马利群. 基于GLAS 大光斑激光波形数据的山地森林高度反演[D]. 北京. 中国科学院研究生院. 2011.

个性服务

查看访问统计

相关权益政策

暂无数据

收藏/分享

除非特别说明，本系统中所有内容都受版权保护，并保留所有权利。