地理研究 ›› 2015, Vol. 34 ›› Issue (11): 2113-2123.doi: 10.11821/dlyj201511010

• • 上一篇    下一篇

基于DEM-NDVI的高山植被带定量刻划

常纯1,2(), 王心源1,3(), 杨瑞霞1,3, 刘传胜1,3, 骆磊1,2, 甄静1,3, 项波1,3, 宋经纬1,2, 廖颖1,2   

  1. 1. 中国科学院遥感与数字地球研究所,数字地球重点实验室,北京 100094
    2. 中国科学院大学,北京 100049
    3. 联合国教科文组织国际自然与文化遗产空间技术中心,北京 100094
  • 收稿日期:2015-06-15 修回日期:2015-09-03 出版日期:2015-11-15 发布日期:2015-11-24
  • 作者简介:

    作者简介:常纯(1989- ),男,湖南衡阳人,硕士,研究方向为遥感影像处理与应用。E-mail:changchun0311@163.com

  • 基金资助:
    国家国际科技合作项目(S2013GR0477);国家自然科学基金(41271427)

A quantitative characterization method for alpine vegetation zone based on DEM and NDVI

Chun CHANG1,2(), Xinyuan WANG1,3(), Ruixia YANG1,3, Chuansheng LIU1,3, Lei LUO1,2, Jing ZHEN1,3, Bo XIANG1,3, Jingwei SONG1,2, Ying LIAO1,2   

  1. 1. Key Laboratory of Digital Earth Science, Institute of Remote Sensing and Digital Earth, CAS, Beijing 100094, China
    2. University of Chinese Academy of Sciences, Beijing 100049, China
    3. International Centre on Space Technologies for Natural and Cultural Heritage under the Auspices of UNESCO, Beijing 100094, China
  • Received:2015-06-15 Revised:2015-09-03 Online:2015-11-15 Published:2015-11-24

摘要:

根据不同高山植被类型具有不同归一化植被指数响应的特点,通过对NDVI的分析来定量刻划高山植被带的海拔分布。首先,利用DEM、NDVI构建DEM-NDVI散点分布图;然后,结合地面调查资料与WorldView-2高分遥感影像对DEM-NDVI散点分布图进行统计回归分析;最后,利用分析结果定量刻划高山植被垂直分带结构。将该方法应用于四川卧龙大熊猫保护区的卧龙关沟,结果表明:① NDVI随海拔升高而呈“Z”字形变化;② DEM-NDVI散点图比样本点DEM-NDVI分布图能更完全地表达高山植被NDVI随高程变化的特征;③ 卧龙关沟东北坡高山植被带海拔高度为3255~4415 m,西南坡高山植被带海拔高度为3193~4473 m,与地面调查得到的区域代表植被的分布高度基本一致。

关键词: 高山植被, 定量刻划, DEM, NDVI, 卧龙

Abstract:

Alpine vegetations, whose distribution vary with changing altitudes, are located in inaccessible areas under cold climate condition. This has caused great challenges in terms of quantitatively characterizing the spatial distribution of alpine vegetation. For example, extremely huge time and money costs are the inevitable handicaps in traditional ground transects and quadrat surveys. Moreover, ground transects, and quadrat survey methods may cause some errors when local ecological features are used to characterize the expanded region. Based on the theory that vegetation of different types have different normalized difference vegetation index (NDVI) responses, NDVI analysis in this study was used to quantitatively characterize the spatial distribution of alpine vegetation. Firstly, a DEM-NDVI scatter diagram was plotted based on DEM and NDVI data. Secondly, statistical regression analysis was conducted in combination of ground survey data, high resolution remote sensing images with the DEM-NDVI scatter diagram. The spatial distribution of alpine vegetation was then derived upon the analysis results. We applied the method to a test area, Wolongguangou, Wolong Giant Panda Reserve. The results demonstrated that: (1) NDVI presented a 'Z' pattern as the altitude increased. (2) The DEM-NDVI scatter diagram expressed the NDVI variation of alpine vegetation more completely when compared to a traditional sample-point DEM-NDVI plot. (3) The elevations of alpine vegetation ranged from 3255 m to 4415 m on the northeast slope and 3193 m to 4473 m on the southwest slope, respectively. This finding was favorably consistent with the ground survey results in the respective regions.

Key words: alpine vegetation, quantitative characterization, DEM, NDVI, Wolong