摘要
油蒿群落覆盖度是判断毛乌素沙地荒漠化程度严重与否的直接指标,其信息的准确获取有助于更为合理的荒漠化监测与评价。本研究以Landsat ETM+为数据源,深入探讨了光谱混合分析过程中影像预处理、端元选取、光谱混合模型选择及光谱混合分析结果分析等关键问题,给出了恰当的解决方案,进而进行了油蒿群落覆盖度的提取,并通过野外实测数据对提取结果进行了验证。验证结果表明:基于光谱混合分析技术提取的油蒿群落分量与实测油蒿群落覆盖度线性相关显著,相关系数为0.88,因而研究区的油蒿群落覆盖度可以通过油蒿群落分量的线性变换得到。因此,光谱混合分析是提取毛乌素沙地油蒿群落覆盖度的有效技术。
Artemisia ordosica community coverage is a direct index to estimate the desertification severity in Mu Us Sandland. Acquiring its information is beneficial to carry out desertification monitoring and evaluation better. In this paper, we utilize spectral mixture analysis to retrieve Artemisia ordosica coverage information, based on Landsat ETM ＋ image. Some key issues in SMA process, including image pre-processing, endmember selection, spectral mixture model selection and SMA results analysis, are discussed in depth, and the suitable solutions are provided. Then the coverage of Artemisia ordosica community is retrieved, and the accuracy of the result is validated based on field survey data. The results show A significant linear relationship is found between Artemisia ordosica community fraction and measured Artemisia ordosica community coverage （the correlation coefficient is 0. 88）. So, Artemisia ordosica community coverage of the research region can be acquired though linear transformation of Artemisia ordosica community fraction. Therefore, SMA is an effective technology for retrieving Artemisia ordosica community coverage accurately in Mu Us Sandland.
出处
《遥感学报》
EI
CSCD
北大核心
2007年第6期923-930,共8页
NATIONAL REMOTE SENSING BULLETIN
基金
国家科技攻关项目(编号:2005BA517A07)
国家科技攻关项目(编号:2005BA517A04)
国家高技术发展计划项目(编号:2006AA12Z108)
国家科技支撑计划项目(编号:2006BAD26B0103)
关键词
油蒿群落
光谱混合分析
端元
毛乌素沙地
Artemisia ordosica
spectral mixture analysis
endmember
Mu Us Sandland
