地理研究

地理研究 ›› 2009, Vol. 28 ›› Issue (1): 10-19.doi: 10.11821/yj2009010003

• 地球信息科学 • 上一篇    下一篇

蒙古高原NDVI的空间格局及空间分异

张雪艳1,2, 胡云锋1, 庄大方1, 齐永清1   

  1. 1. 中国科学院地理科学与资源研究所, 北京 100101;
    2. 中国科学院研究生院, 北京 100049
  • 收稿日期:2007-11-24 修回日期:2008-06-25 出版日期:2009-01-25 发布日期:2009-01-25
  • 作者简介:张雪艳(1979-),女,浙江绍兴人,博士生。主要从事地理信息系统与遥感技术的应用研究。E-mail:xyzhang@lreis.ac.cn *通讯作者 : 胡云锋(1974-),江西赣州人,助理研究员。主要从事蒙古高原资源环境与可持续发展研究。E-mail:huyf@lreis.ac.cn
  • 基金资助:

    国家自然科学基金(40571130, 40701150);中国科学院国际合作重点项目(GJHZ06);中国科学院科学数据库(INF105-SDB-A3);中国科学院地理科学与资源研究所所长基金(06W60001SZ);中国科学院陆地水循环及地表过程重点实验室开放基金(WL2005001)

The spatial pattern and differentiation of NDVI in Mongolia Plateau

ZHANG Xue-yan1,2, HU Yun-feng1, ZHUANG Da-fang1, QI Yong-qing1   

  1. 1. Institute of Geographic Science and National Resources Research, CAS, Beijing 100101, China;
    2. Graduate University of Chinese Academy of Sciences, Beijing 100049, China
  • Received:2007-11-24 Revised:2008-06-25 Online:2009-01-25 Published:2009-01-25
  • Supported by:

    国家自然科学基金(40571130, 40701150);中国科学院国际合作重点项目(GJHZ06);中国科学院科学数据库(INF105-SDB-A3);中国科学院地理科学与资源研究所所长基金(06W60001SZ);中国科学院陆地水循环及地表过程重点实验室开放基金(WL2005001)

PDF (PC)

赞

被引次数

148 | 47

摘要：

基于GIMMS NDVI多年最大值合成数据,采用空间统计学方法,利用Moran’ s I系数分析、半变异函数分析以及分维分析等3种方法,对蒙古高原NDVI空间格局及空间分异特征进行研究。结果表明:(1)蒙古高原NDVI的空间分布在全局范围内呈现正的空间自相关,相似的NDVI值倾向于聚集在一起,这表明蒙古高原植被具有较好的整体性,地表植被无显著破碎化;(2)蒙古高原NDVI的空间分布虽然同时受到结构性因子和随机性因子的影响,但结构性因子占据绝对控制地位,结构性因子引起的空间变异占系统总变异的88.7%;(3)蒙古高原NDVI存在各向异性的分布特征,具有相似NDVI值的像元主要沿着西北-东南方向展布;全局NDVI空间自相关距离约为1178km,西北-东南方向与东北-西南方向的空间自相关距离比可达2.4 ∶ 1。

关键词: GIMMS NDVI, 蒙古高原, 空间格局, 空间分异, 空间统计

Abstract:

It is a key step to deepen the understandings about the structure, functions and processes of regional resources and environment by using spatial, quantitative, multi-scale and temporal analysis, and there are many obvious advantages to study on the spatial pattern, spatial difference and the related ecological & environment characteristics by using remote sensing technology. As an independent geographic and ecological unit, the status of Mongolia Plateau ecology system and its changes have great effects on the environment of North China and even the whole East Asia, which makes it necessary to develop certain research works in the whole Mongolia Plateau in both theory aspect and practice aspect. In this study, based on GIMMS NDVI dataset of 1982-2003, the new Mongolia Plateau NDVI Dataset using MVC method was constructed and the stability in the time scale was also assessed. Then three spatial statistic methods including Moran coefficient, semi variance function and fractal analysis, were selected to investigate the spatial pattern and spatial differentiation of NDVI in Mongolia Plateau. The results show that: (1) The NDVI in Mongolia Plateau shows positive spatial autocorrelation in the whole region and similar NDVI are apt to aggregate together, which means the vegetation cover of Mongolia Plateau is generally intact and little fragmental. The NDVI distribution pattern is indeed clear; (2) In spite of the effects of both structural factors and random factors, the spatial distribution pattern of NDVI in Mongolia Plateau is mainly controlled by the structural factors which induce 88.7% of the total spatial variations, while only 11.3% of the total spatial variations is induced by the random factors; (3) There exists a clear anisotropy about NDVI distribution pattern in Mongolia Plateau, and there are higher spatial autocorrelation in the Northwest-Southeast direction which means those points with similar NDVI value are mainly distributed along Northwest-Southeast direction. Generally, the radius of NDVI patch is about 1178km, and the length-width ratio of those patches is about 2.4. According to the research, temporal NDVI study based on spatial statistics method could not only explore the NDVI spatial pattern & spatial difference but also indicate the geographic essences of NDVI and the relevant spatial statistics parameters.

Key words: GIMMS NDVI, Mongolia Plateau, spatial pattern, spatial differentiation, spatial statistic