地理研究

地理研究 ›› 2006, Vol. 25 ›› Issue (4): 570-578.doi: 10.11821/yj2006040002

• 论文 • 上一篇    下一篇

应用“北京一号”遥感数据计算官厅水库库滨带植被覆盖度

杨胜天1, 李茜1, 刘昌明1,2, 智泓3, 王雪蕾1   

  1. 1. 遥感科学国家重点实验室北京师范大学地理学与遥感科学学院环境遥感与数字城市北京市重点实验室 北京100875;
    2. 北京市水科学研究所,北京100875;
    3. 中国科学院地理科学与资源研究所,北京00044
  • 收稿日期:2006-02-22 修回日期:2006-05-28 出版日期:2006-08-15 发布日期:2006-08-15
  • 作者简介:杨胜天(1965-),贵州人,教授,博士生导师。主要从事自然地理、遥感和环境科学研究。 E-mail:yangshengtian@bnu.edu.cn
  • 基金资助:

    国家自然科学基金项目(40471127)

Detecting vegetation fractional coverage of riparian buffer strips in Guanting Reservoir based on "Beijing-1"|remote sensing data

YANG Sheng-tian1, LI Qian1, LIU Chang-ming1,2, ZHI Hong3, WANG Xue-lei1   

  1. 1. State Key Laboratory of Remote Sensing Science,School of Geography,Beijing Key Laboratoryfor Remote Sensing of Environment and Digital Cities,Beijing Normal University,Beijing 100875,China;
    2. Institute of Geographic Sciences and Natural Resources Research,CAS,Beijing 100101,China;
    3. Beijing Hydraulic Research Institute,Beijing 100044,China
  • Received:2006-02-22 Revised:2006-05-28 Online:2006-08-15 Published:2006-08-15

PDF (PC)

赞

摘要：

“北京一号”卫星是拥有多方面技术优势的一颗对地观测小卫星。本文在运用“北京一号”、SPOT5、QuickBird遥感图像对官厅水库库滨带植被覆盖度进行综合监测的基础上,对三种不同分辨率的遥感图像进行基于统计的尺度转换,并应用尺度转换的结果修正了“北京一号”图像提取的植被覆盖度。经检验,运用SPOT5和QuickBird图像对“北京一号”图像进行像元分解,将统计结果与“北京一号”图像的提取信息建立统计模型,应用该统计模型可以有效地提高“北京一号”图像提取植被覆盖度的精度。对湿生植被进行样方调查,结果证明运用像元分解和统计模型的方法使“北京一号”提取植被覆盖度的精度较运用植被指数转换模型的计算精度提高了22.7%。应用该方法可以更有效地运用“北京一号”遥感数据进行连续、大面积的植被监测。

关键词: “北京一号”, 植被覆盖度, 像元分解, 统计模型, 尺度转换

Abstract:

"Beijing-1"satellite(DMC+4 microsatellite) was launched on October 27th 2005,which covers a large area and has outstanding ability of continual observation.The vegetation fractional coverage is an important parameter for describing vegetation quality and indicating ecosystem change.In this study,"Beijing-1",SPOT5 and QuickBird images were integrated to detect and analyze the vegetation fractional coverage in riparian buffer zones of Guanting Reservoir.On the basis of integrated detection,scale transforming method was used between three images to improve the precision of vegetation fractional coverage measured by "Beijing-1" image.The concrete methods are pixel decomposability method and statistic model establishment,that is,SPOT5 and QuickBird pixels were used to decompose "Beijing-1" image pixel,and then established the statistic models between decomposability results and vegetation fractional coverage measured by"Beijing-1" image.The results show that:(1) The correlation coefficients of vegetation fractional coverage between "Beijing-1",SPOT5 and QuickBird images are high,which suggests that using "Beijing-1" image to detect vegetation fractional coverage of riparian buffer zones of Guanting reservoir is feasible.(2) Using these statistic models can effectively improve the precision of vegetation fractional coverage measuring from "Beijing-1" image.For reed marshes,compared with the result of using vegetation fractional coverage calculation equation,the mean absolute error of vegetation fractional coverage measured by "Beijing-1" image was reduced by 22.7% after applying statistic models.This is a practice on scale transformation of remotely sensed data.(3) Expanding the application extent,pixel decomposability method and statistic model establishment are feasible in enhancing the application precision of lower spatial resolution remotely sensed data by using high spatial resolution remotely sensed data.The integrated application of multi-scale remotely sensed data is a significant approach to promote the precision of acquiring earth surface parameters on large scale.This has become a development trend of remote sensing technology.

Key words: "Beijing-1", vegetation fractional coverage, pixel decomposability, statistic model, scale transformation