IMHE OpenIR  > 数字山地与遥感应用中心
基于时空数据融合模型的西藏“一江两河”地区植被覆盖度构建研究
Alternative TitleResearch on Vegetation Fractional Coverage Constructing of Middle Reaches of Brahmaputra River in Tibet Based on Spatial and Temporal Data Fusion Model
Language中文
陈阳
Thesis Advisor范建容
2013
Degree Grantor中国科学院研究生院
Place of Conferral北京
Degree Name硕士
Degree Discipline地图学与地理信息系统
Keyword植被覆盖度 时空数据融合模型 云去除 高时空分辨率 植被生长峰值阶段 亚像元混合模型
Abstract植被覆盖度不仅是衡量地表植被生长状况的一个重要指标,而且还是土壤侵蚀的重要影响因素。西藏“一江两河”地区生态环境脆弱、土壤侵蚀严重,对人类活动跟气候变化响应较为敏感,因此对该区域的植被覆盖度进行研究极具科学价值。基于研究区高时空分辨率遥感数据缺乏的现状,本研究旨在探索高时空分辨率的植被覆盖度数据的构建方法,主要聚焦于两个问题:一是如何构建高时空分辨率的遥感数据;二是在高时空分辨率遥感数据的基础上,怎样准确估算出植被覆盖度。本研究取得的成果主包括一下几个方面: ⑴基于多源多时相遥感数据,采用基于规则的面向对象分类方法实现了研究区地表覆被分类,分类总体精度为89.28%。 ⑵针对已提出的各类云去除方法在实际应用中存在的局限性,本研究将时空数据融合模型引入到云去除方法中,提出了一种基于时空数据融合模型的TM影像云去除方法。在修复后的影像中,替换区域与无云区域色调基本一致;通过无云区TM合成数据间接对替换云及其阴影区数据的精度评价,结果表明:替换数据各波段整体精度优于83%,平均精度为87.15%。 ⑶针对增强时空适应反射率融合模型(ESTARFM)在归一化植被指数(NDVI)预测应用上存在的不足,引入通过地表覆被类型选择相似像元的思想对模型进行改进并实现了研究区植被生长峰值阶段归一化植被指数(NDVI)的构建。 ⑷实现了研究区高时空分分辨率植被覆盖度的构建。基于地表覆被类型与植被生长峰值阶段的归一化植被指数(NDVI)数据,综合应用等密度、非密度亚像元混合模型估算研究区植被覆盖度的总体精度约85%。 本研究通过时空适应反射率融合模型(ESTARFM)有效地将TM的高空间分辨率与MODIS的高时间分辨率特点结合在一起,实现了研究区植被生长峰值阶段归一化植被指数(NDVI)的构建。通过以生长峰值阶段的归一化植被指数(NDVI)为输入数据,提高了亚像元混合模型在植被覆盖度估算时对输入数据的抗扰动性。经野外实测数据验证,植被覆盖度的估算精度约85%,这表明:本研究所提出的方法能够实现对缺少高时空分辨率遥感数据区域的植被覆盖度估算。
Other AbstractThe vegetation fractional coverage is not only an important indicator of vegetation growth, but also is one of the most important factors of soil erosion. With fragile ecology evironment and serious soil erosion, middle reaches of Brahmaputra River in Tibet, responds more sensitive to human acitivites and climate change. So the study of vegetation fractional coverage in the region is of great scientific value. Lacked of remote sensing data with high spatial and temporal resolution, the study aims to find a method of constructing vegetation fractional coverage with high spatial and temporal resolution. This study focused on two points: First, how to build remote sensing data with high temporal and spatial resolution; Second, based on high spatial and temporal resolution remote sensing data, how to estimate vegetation fractional coverage accurately. In this study, the main results achieved are as follows: (1) Based on multi-source multi-temporal remote sensing data, land use and land coverage data of study area was obtained by object-oriented classification method, with the overall accuracy of 89.28%. (2) To solve the limitation of the existing models for cloud removal in practical application, in this study, a new method was proposed based on spatial and temporal data fusion models. In the result image, the color of the replaced area is consistent with the color of area uncontaminated by clouds and shade; verified indirectly based on the data of target TM image and composed image without cloud and cloud shade, the accuracy of all bands of the replaced data are better than 83% and the mean accuracy is 87.15%. (3) Selecting similar pixel by land coverage type, the author improved the performance of enhanced spatial and temporal adaptive reflectance fusion model (ESTARFM) in normalized difference vegetation index (NDVI) prediction and then constucted the NDVI data in the peak stage of vegetation grouwth of study area. (4) The vegetation fractional coverage of study area with high spatial and temproral roslution was constructed by the method proposed by author. Based on the land coverage type data and NDVI data in the peak stage of vegetation growth, the vegetation fractional coverage of study area with about 85% overall accuracy was estimated by density sub-pixel model and non-density sub-pixel model. In this study, normalized difference vegetation index (NDVI) data with high sptatial and temproral resolution was constucted by conbining advantages both of TM and MODIS using spatial and temporal adaptive reflectance fusion model (ESTARFM). The normalized difference vegetation index (NDVI) data, in the peak stage of vegetation growth, as input data improved the anti-disturbance to input data of sub-pixel mixed model estimates of vegetation coverage. Validated by the data measured in field, the accuracy of estimated coverage is about 85%, which suggest that it is viable to estimate vegetation fractional coverage in regions, especially lacking of remote sensing data with high spatial and temporal resolution.
Document Type学位论文
Identifierhttp://ir.imde.ac.cn/handle/131551/6978
Collection数字山地与遥感应用中心
Recommended Citation
GB/T 7714
陈阳. 基于时空数据融合模型的西藏“一江两河”地区植被覆盖度构建研究[D]. 北京. 中国科学院研究生院,2013.
Files in This Item:
File Name/Size DocType Version Access License
陈阳:基于时空数据融合模型的西藏“一江两(4858KB) 开放获取CC BY-NC-SAView Application Full Text
Related Services
Recommend this item
Bookmark
Usage statistics
Export to Endnote
Google Scholar
Similar articles in Google Scholar
[陈阳]'s Articles
Baidu academic
Similar articles in Baidu academic
[陈阳]'s Articles
Bing Scholar
Similar articles in Bing Scholar
[陈阳]'s Articles
Terms of Use
No data!
Social Bookmark/Share
File name: 陈阳:基于时空数据融合模型的西藏“一江两河”地区植被覆盖度构建研究.pdf
Format: Adobe PDF
This file does not support browsing at this time
All comments (0)
No comment.
 

Items in the repository are protected by copyright, with all rights reserved, unless otherwise indicated.