EN中文

Journal of Geo-information Science >

2015 , Vol. 17 >Issue 11: 1333 - 1340

DOI: https://doi.org/10.3724/SP.J.1047.2015.01333

Analysis of NDVI and the Impact of Human Activity in China from May to September During 1982 to 2006

  • WANG Yuanxiang ,
  • TANG Shihao , * ,
  • ZHENG Zhaojun
Expand
  • National Satellite Meteorological Centre, Beijing 100081, China
*Corresponding author: TANG Shihao, E-mail:

Received date: 2015-03-24

  Request revised date: 2015-09-06

  Online published: 2015-11-10

Copyright

《地球信息科学学报》编辑部 所有
Fold

Abstract

Vegetation is an important part of the terrestrial ecosystem, and it is the natural bond for linking land, atmosphere and ecosphere. Vegetation changes with climate and human activity, thus, studying its sudden and trend changes is an important subject. Using the methods of 3-year moving t-test, Mann-Kendall test (MK-test) and anomaly analysis, the sudden and trend changes of the normalized difference vegetation index (NDVI of AVHRR GIMMS) in China from May to September during 1982 to 2006, as well as their causes were analyzed. The t-test and MK-test results show that there was a sudden change of NDVI in the eastern China in 1998, and there couldn’t be a sudden change of NDVI in the northeast China and the Tibetan Plateau during 1982 to 2006. The changing trend of NDVI indicates that there was an increasing trend during 1982 to 1998, and then a decreasing trend from 1999 to 2006 in the eastern China. The causes of the sudden and trend changes of NDVI reveal that with the massive urbanization process in China that started in the late 1990s, the city and built-ups increased and the arable area decreased, and the NDVI indicated a sudden change with the vegetation decreased in the eastern China. Moreover, the satellite instruments and climate factors couldn’t be the main causes of the decreasing vegetation in the eastern China.

Key words： NDVI; sudden change; causes

Cite this article

WANG Yuanxiang , TANG Shihao , ZHENG Zhaojun . Analysis of NDVI and the Impact of Human Activity in China from May to September During 1982 to 2006[J]. Journal of Geo-information Science, 2015 , 17(11) : 1333 -1340 . DOI: 10.3724/SP.J.1047.2015.01333

1 引言

植被生态系统对水土保持起着不可替代的作用,它在调节大气成分、抑制温室气体浓度上升和维持稳定的气候方面也起着重要的作用。植被是联系陆地、大气和生物的自然纽带,它的变化对全球能量和物质循环产生了重要的影响。植被变化(包括突变和变化趋势)是一个变量变化过程中的转折点,该变量呈增加或减少的趋势。趋势是客观存在的运行特性,是自然界的客观存在,包括上升和下降趋势,其转折点就是突变。
以“全球变暖”为突出标志的全球环境变化及其可能对生态系统产生的严重影响,已引起了科学家、各国政府与社会各界的极大关注。因为它影响到国家的政治和经济、区域政策的制定[1],全球变暖导致了大气示踪气体和荒漠化增加[2],故采用HOLDRIDGE生命地带系统与CHIKUGO模型,对全球变化后的中国陆地生态系统植被的地理分布及净第一生产力进行研究,发现中国植被在气温增加2 ℃或4 ℃,年平均降水量增加20%的情况下,东部的森林界限将向北移动3~5°N,尤其以东北大兴安岭的北方寒温针叶林受影响较大,可能都全部北移出境;南部沿海的热带雨林季雨林则占有较大 面积[3]。
对于植被的突变,在南极King George岛长期环境资料中已经得到较好的研究[4-6]。国内外学者使用AVHRR的NDVI资料对1982-2002年西非植被的年际和季节性变化进行趋势分析,发现该地的植被有明显的年际变率和季节性变率[7]。同时,对美国Arizona东南部植被的时空变化趋势进行了评价[8]。在过去的20-30年,中国主要生态系统(如森林、草地、灌木和农田)的净生产力呈显著增长的趋势,而淡水湖和城区的植被呈显著下降的趋势[9]。东北平原NDVI的最大值从1998年的0.41下降到2007年的0.37,表明植被减少,春夏季显著下降,秋季有些上升[10]。另外,用Mann-Kendall检验分析了1982-2006年内蒙古不同气候区植被的变化趋势,发现中西部为上升趋势,东部为下降趋势[11]。2001-2010年长江中下游地区年最大EVI整体呈减少趋势,夏冬季均呈减少趋势(以2月和8月最为显著),春秋季则呈增加趋势(以5月和10月最为显 著)[12]。1989-2003年,上海市的植被面积呈现持续下降趋势,浦东新区植被减少最多[13]。青藏高原2000-2007年NDVI研究,结果表明,林芝和山南南部的植被长势较好,其次是昌都、拉萨和那曲的东部,最后是那曲中南部地区,2003-2006年青藏高原植被显著减少,2006年以后有所增加[14]。喜马拉雅山自然保护区5.04%的植被明显减少,13.19%略为减少,26.39%略为增加,0.97%显著增加,54.41%维持稳定,减少区位于保护区的南部,增加区位于保护区的北部和雅鲁藏布江的南部,稳定区位于减少区和增加区之间[15]。1992-1998年天山北坡植被覆盖度和生物量持续增加,1987-1998年三江河流域从前山带到北部沙漠区植被覆盖度和生物量显著增加,尤其是绿洲外围的北部沙漠区和荒漠过渡带;从增加的幅度看,平原区增幅明显大于山区,后期大于前期[16];22年间西北地区植被以增加趋势为主,主要分布在新疆和河西走廊绿洲、黄河沿岸灌区以及青海草区,NDVI减少地区主要分布在西北东部[17]。尽管植被的突变和变化趋势已得到了较好的研究,但是大多数研究都集中在气候变化对植被突变和趋势的影响,而人类活动对其影响的研究较少,本文将分析NDVI变化,着重讨论人类活动之一(城市扩张)对植被变化的影响。

2 研究数据与分析方法

本研究使用了1982-2006年5-9月NDVI(Normalized Difference Vegetation Index)8天合成资料,分辨率为8 km×8 km,该资料来源于NOAA/AVHRR GIMMS(Global Inventory Modeling and Mapping Studies)数据集;LAI(Leaf area index)数据集和1980、2000、2010年3期MSS/TM/ETM+3种传感器的影像解译出来的LUCC(Land-use and land-cover change)数据(源自中国科学院地理科学与资源研究所)。同时还使用了1982-2006年5-9月中国753个气象台站的温度和降水资料(源于国家气象信息中心)。
对此,采用3年滑动t检验和MK检验方法检查NDVI的突变,并分析其变化趋势,3年滑动t检验和MK检验方法[18-19]如式（1）-（6）所示。
滑动t检验：
t = x ̅ 1 - x ̅ 2 s × 1 n 1 + 1 n 2 （1）
s = n 1 s 1 2 + n 2 s 2 2 n 1 + n 2 - 2 （2）
式中, x ̅ 1 , x ̅ 2 是2个子样本的平均值; n 1 ， n 2 和 s 1 ， s 1 ,分别是2个子样本的长度和标准偏差。
设定一个临界值ta,假如|t|<ta,则2个子样本没有差异;假如|t|>ta,则2个子样本可能有差异或出现突变。
MK检验： UF = k [ s k - E ( s ) ] k Var ( s k ) （3）
其中, s k = ∑ i = 1 k r i （4）
r i = + 1 当 x i > x j ( j = 1 ， 2 ， … .. ， n ) 0 否则 （5）
E ( s k ) = n ( n + 1 ) 4 （6）
Var ( s k ) = n ( n - 1 ) ( 2 n + 5 ) 72 （7）
式中, E ( s k ) 和 Var ( s k ) 是 s k 的平均值和方差。
对于样本的相反序列,可得到UBk,假如UFk和UBk相交,这个交叉点可能是突变点。如果交叉点接近3年滑动t检验中变量出现显著性差异的时间,那么,这个交叉点就能确定为突变点[20-21]。

3 NDVI的突变和变化趋势分析

3.1 东亚地区NDVI的气候平均

图1为1982-2006年5-9月东亚地区NDVI的气候平均。从图1可看出,NDVI的高值区主要出现在东北地区、华东地区和青藏高原的东南部,最大值0.8出现在东北地区。据此,本文分3个区域来研究NDVI的变化,即东北地区(NEC,120°~140°E,40°~55°N)、华东地区(EC,105°~120°E,20°~40°N)和青藏高原(TP,70°~105°E,25°~40°N)。图1中的3个方框分别代表东北地区、华东地区和青藏高原。
Fig. 1 The climatic average of May-September NDVI in eastern Asia during 1982-2006

图1 1982-2006年5-9月东亚地区NDVI的气候平均

3.2 NDVI的突变和变化趋势分析

本文用NEC NDVI、EC NDVI和TP NDVI代表东北地区(Northeastern China)、华东地区(Eastern China)和青藏高原(Tibetan plateau)5-9月NDVI的区域平均值。采用3年滑动t检验和MK检验检查1982-2006年NEC NDVI、EC NDVI和TP NDVI是否存在突变。由于3年滑动t检验对连续3年的数据进行滑动平均后进行计算,故其时间序列是从1984-2003年。
图2(a)和(b)是NEC NDVI 3年滑动t检验和MK检验的时间序列。从3年滑动t检验(图2(a))中可看出,NEC NDVI分别在1988和1994年超过临界线,即有显著性差异。从MK检验(图2(b))中可看出,UF和UB曲线分别在1984、2003和2004年相交。3年滑动t检验中NEC NDVI超过临界线(出现显著性差异)的时间与MK检验中UF和UB曲线相交的时间不一致,即通过2种检验方法所得到的转折时间不一致,因此,仅仅通过3年滑动t检验和MK检验不能判断NEC NDVI是否有突变出现。进一步检查1988、1994、1984、2003和2004年NEC NDVI的区域平均值,发现连续性较好,没有发现与其他资料的显著性差异,因此,NEC NDVI在1982-2006年期间可能没有突变出现。为了清楚地显示T检验、MK检验、距平变化,以及突变的情况,本文将这些分析列于表1中。
Tab. 1 T test, MK test, anomaly and sudden change

表1 T检验、MK 检验、距平变化及突变

区域 有无显著性差异 有无突变
东北地区 T检验 有(1988年,1994年) 无
无
MK检验 有(1984、2003和2004年)
华东地区 T检验 有(1998年) 有
有
有
有
有
有
MK检验 有(1998年)
NDVI距平变化 1982-1998年正距平1999-2006年负距平
LAI距平变化 1982-1998年正距平1999-2006年负距平
青藏高原 T检验 无 无
MK检验 有(1988、1989、2002和2004年) 无
图2(c)和(d)是EC NDVI 3年滑动t检验和MK检验的时间序列。从3年滑动t检验(图2(c))可看出,EC NDVI在1998年超过临界线,即有显著性差异,表明EC NDVI在1998年可能存在突变。从MK检验(图2(d))可看出,UF和UB曲线在1998年相交,表明EC NDVI在1998年可能存在突变。3年滑动t检验中EC NDVI超过临界线(出现显著性差异)的时间与MK检验中UF和UB曲线相交的时间一致,即通过2种检验方法所得到的突变时间一致,因此通过3年滑动t检验和MK检验判断EC NDVI存在突变。
图2(e)和(f)是TP NDVI 3年滑动t检验和MK检验的时间序列。从3年滑动t检验(图2(e))中可看出,TP NDVI在1982-2006年间没有超过临界线,即没有显著性差异,表明TP NDVI在1982-2006年间没有突变。从MK检验(图2(f))可看出,UF和UB曲线分别在1988、1989、2002和2004年相交。3年滑动t检验中TP NDVI没有超过临界线(没有出现显著性差异),MK检验中UF和UB曲线尽管出现了4次相交,但通过3年滑动t检验和MK检验不能判断TP NDVI是否有突变出现。进一步检查1988、1989、2002和2004年TP NDVI的区域平均值,发现连续性较好,没有发现与其他资料的显著性差异,因此,TP NDVI在1982-2006年间可能没有突变出现。
Fig. 2 The time series of 3-year moving t-test (the two straight lines represent the threshold limit lines) and MK-test (the UF and UB curves are marked by the solid and dash curves, and the two straight lines are the threshold limit lines)

图2 1982-2006年5-9月NDVI 3年滑动t检验和MK检验的时间序列

通过3年滑动t检验和MK检验发现,华东地区在1998年NDVI存在突变,而东北地区和青藏高原NDVI可能没有出现突变。
从MK检验的UF曲线(图2(d))可看出,EC NDVI在1982-1998年为较平稳的正变化趋势(除了1989年为负值),1999-2006年转为明显的下降趋势(以1998年为突变点)。图3(a)是EC NDVI距平的时间序列,从图中可看出,1982-1998年大部分EC NDVI为正距平(除了1982和1989年为负距平),1999-2006年大部分EC NDVI为负距平(除了2004年为弱的正距平);为了进一步显示NDVI变化的可靠性,本文采用中国科学院地理科学与资源研究所的LAI资料,图3(b)是EC LAI距平的时间序列,从图中可以看出,1982-1998年大部分EC LAI为正距平(除了1982,1989和1996年为负距平),1999-2006年大部分EC LAI为负距平(除了2004年为弱的正距平);即EC NDVI和LAI有明显的年代际变化,并与MK检验UF曲线中的趋势变化相吻合。
Fig. 3 The time series of the anomaly of NDVI and LAI in eastern China from May to September during 1982-2006

图3 1982-2006年5-9月华东地区NDVI距平和LAI距平的时间序列

3.3 华东地区NDVI出现突变和变化趋势的驱动 因素

为了发现1998年华东地区NDVI出现突变和变化趋势的可能原因,本文从卫星仪器、气候因子和人类活动3方面分别进行分析。
GIMMS(Global Inventory Modeling and Mapping Studies)数据集是1981-2006年的NDVI(Normalized Difference Vegetation Index)产品。该数据集由NOAA 7、9、11、14、16 和17卫星上搭载的AVHRR(Advanced Very High Resolution Radiometer)仪器所获得的图像反演而来。该NDVI数据集经过定标、视图几何校正、火山气溶胶和不涉及植被变化的其他效果的处理[22],因此,资料本身连续性较好,前后资料未出现显著性差异。由于华东地区的NDVI出现了显著性差异,可能是由于某些原因引起的差异,而不可能是资料本身的原因,假如是资料本身的原因,不仅华东地区出现差异,其他地区(东北、青藏高原)也会出现差异。
因此,卫星仪器不是导致1998年华东地区NDVI出现突变的原因,那么说明出现突变的原因可能是气候因子或人类活动造成的。下面从气候因子方面进行分析,试图发现它们是否是1998年华东地区NDVI出现突变的原因。
图4(a)是华东地区1982-2006年5-9月温度距平的时间序列,从图中可看出,1982-1993年大多数温度距平为负(除了1990年为正距平),表明这段时期温度偏低;1994-2006年所有温度距平为正,表明这段时期温度偏高。转折点在1994年,比植被变化的转折点(1998年)偏早4年。
图4(b)是华东地区1982-2006年5-9月降水量距平的时间序列,从图中可看出,1982-1984年降水量距平为正,1985-1992年大多数降水量距平为负(除了1988年为弱的正距平),表明这段时期降水量偏少;1993-2006年大部分降水量距平为正(除了1997、2000、2001、2003和2004年为负距平),表明这段时期降水量偏多。转折点在1993年,与温度变化的转折点(1994年)相近,比植被变化的转折点(1998年)偏早5年。
Fig. 4 The time series of temperature anomaly and precipitation anomaly in eastern China from May to September during 1982-2006

图4 1982-2006年5-9月华东地区温度距平和降水量距平的时间序列

从气候因子的分析可看出,华东地区温度和降水量发生变化趋势的转折点分别在1994和1993年,比植被变化的转折点(1998年)偏早4-5年,因此,温度和降水量并非是导致华东地区植被发生突变的主要因素。
众所周知,中国大规模的房地产开发始于20世纪90年代后期并一直持续至今,尤其是华东地区人口密集,大量的房地产开发占用耕地面积,是否是导致植被出现突变并显著减少的主要原因呢?
图5(a)是1980年中国城镇和建设用地,从图中可以看出,城镇和建设用地浓密地分布在中国的东北和东部地区;图5(b)是2000年和1980年中国城镇和建设用地差值,从图中可看出,从1980-2000年,东北和华东地区城镇和建设用地有少量的增长;图5(c)是2010年和2000年中国城镇和建设用地差值,从图中可以看出,从2000-2010年,东北城镇和建设用地有少量的增长,而华东地区城镇和建设用地显著地增长。这表明东北和华东在2000年前20年城镇和建设用地增长得很少,但是2000年后10年华东地区增长得很快。因此,快速增长的城镇和建设用地导致1998年后华东地区植被的减少,而对东北地区影响较小。
为了进一步显示城市建筑变化的可靠性,本文也使用了中国科学院地理科学与资源研究所的土地覆盖资料。图5(d)提取的是华东地区1982-2006年城市建筑距平的时间序列,从图中可看出,1982-1997年大部分城市建筑距平为负,表明华东地区城市建筑较少;1998-2006年所有城市建筑距平为正,表明华东地区城市建筑较多。
Fig. 5 The urban and built-up areas (0.01 km-2)

图5 城镇和建设用地(0.01 km-2)

综上可知,卫星仪器和气候因子并非是导致华东地区植被发生突变并显著减少的主要原因,而20世纪90年代后期,并一直持续至今的大规模的房地产开发占用耕地面积,可能是导致华东地区植被发生突变并显著减少的主要因素。

4 结论

使用3年滑动t检验、MK检验和距平分析法,研究了中国地区(东北地区、华东地区和青藏高原)1982-2006年5-9月平均的NDVI的突变和趋势变化以及可能的原因。
3年滑动t检验和MK检验表明,1998年华东地区的NDVI出现了突变,而东北地区和青藏高原NDVI没有突变出现。NDVI趋势变化的分析表明,1982-1998年华东地区NDVI为较平稳的正变化趋势,1998-2006年转为明显下降的趋势(以1998年为转折点)。因此,华东地区NDVI存在明显的趋势 变化。
NDVI突变和变化趋势及其成因的研究表明,卫星仪器和气候因子并非是导致该地植被出现突变和趋势变化的主要原因。20世纪90年代后期至21世纪初,随着华东地区大规模的城市化建设和房地产的过度开发,导致耕地面积减少,华东地区植被于1998年出现了突变,并从偏多转为明显偏少的趋势,因此,人类活动是导致华东地区植被显著减少的主要原因。
房地产的过度开发,大量占用耕地,导致耕地面积减少,应引起国家相关部门的高度重视,否则人类的生存环境将遭到难以想象的巨大破坏。

The authors have declared that no competing interests exist.

References

Publishing order | Descend order by publishing year | Descend order by cited within
[1]
周广胜,王玉辉.全球变化与气候-植被分类研究和展望[J].科学通报,1999,44(24):2587-2593.lt;p>对全球变化与陆地生态系统关系研究的核心问题&mdash;&mdash;&mdash;气候 植被关系的研究进展进行综述和讨论 ,指出气候 植被分类研究3个阶段的特点 :( 1 )以现实自然植被类型与气候相关性为特征的气候 植被分类研究 ;( 2 )以对植物生理活动具有明显限制作用的气候因子为指标的气候 植被分类研究 ;( 3)综合反映植被的结构和功能变化的气候 植被分类研究 .在此基础上 ,提出了全球变化背景下中国气候 植被分类研究的方向 ,并强调了气候 植被分类研究必须考虑与大气环流模式的耦合.</p>

[2]
Wood F B.The need for systems research on global climate change[J]. Systems Research, 1988,5(3):225-240.Abstract Recent research has documented the immense complexity of global climate change. The climate system (1) is characterized by multiple inputs and feedback relationships involving many components of the earth system that (2) operate on variable time and geographic scales and (3) can exhibit nonlinear and counterintuitive behavior leading to (4) possibly abrupt changes in equilibrium conditions. Unprecedented rates of increase in atmospheric trace gases and deforestation justify intensive systems research to help verify current and projected global warming trends. This systems review identified numerous uncertainties that warrant attention.

DOI

[3]
张新时. 植被的PE(可能蒸散)指标与植被-气候分类(一)—几种主要方法与P E P 程序介绍[J].植物生态学与地植物学学报,1989,13(1):1-9.植物群落学研究的任务之一是关于群落的环境解释。植被-气候的相关定量分析是其中主要的一环。可能蒸散(PE)作为综合热量与水分两个最重要的生态因子的参数与联系植物及其环境的数量指标而引起了植物生态学家、地理学家与气候学家的重视。本文分篇介绍几种最重要与较成功的PE计算方法及植被-气候分类,并提供其微机计算与分类程序(PEP)以便使用,并期望促进这方面的研究。介绍的方法有:Penman,Thornthwaite,Holdr-idge与Kira(吉良竜夫)的公式或计算法。

DOI

[4]
Howell J F.Identifying sudden changes in data[J]. Monthly Weather Review, 1995,123:1207-1212.Abstract This note describes a method for locating sudden changes in mean data values. Positions of sudden changes are boundaries of variable-width blocks of data. These boundaries could correspond to synoptic frontal boundaries, the downstream edge of a wind gust, or, generally, any anomalous change in a locally averaged quantity. The algorithm described here is applied to artificial signals, century-long records of precipitation, and atmospheric turbulence data.

DOI

[5]
Valeria B, Fernando M, Irene R S, et al.Analysis of trends and sudden changes in long-term environmental data from King George Island (Antarctica): relationships between global climatic oscillations and local system response[J]. Climatic Change, 2013,116(3-4):789-803.The Western Antarctic Peninsula is one of the most rapidly warming regions on earth. It is therefore important to analyze long-term trends and inter-annual patterns of change in major environmental parameters to understand the process underlying climate change in Western Antarctica. Since many polar long-term data series are fragmented and cannot be analysed with common time series analysis tools, we present statistical approaches that can deal with missing values. We applied U-statistics after Pettit and Buishand to detect abrupt changes, dynamic factor analysis to detect functional relationships, and additive modelling to detect patterns in time related to climatic cycles such as the Southern Annular Mode and El Nio Southern Oscillation in a long-term environmental data set from King George Island (WAP), covering 20 years. Our results not only reveal sudden changes for sea surface temperature and salinity, but also clear patterns in all investigated variables (sea surface temperature, salinity, suspended particulate matter and Chlorophyll a) that can directly be related to climatic cycles. Our results complement previous findings on climate related changes in the King George Island Region and provide insight into the environmental conditions and climatic drivers of system change in the study area. Hence, our statistical analyses may prove valuable for other polar environmental data sets and contribute to a better understanding of the regional variability of climate change and its impact on coastal systems.

DOI

[6]
Meng M, Ni J, Zong M J.Impacts of changes in climate variability on regional vegetation in China: NDVI-based analysis from 1982-2000[J]. Ecological Research, 2011,26(2):421-428.Abstract<br/>Three methods were used to distinguish the characteristics of changes in climate variability and normalized difference vegetation index (NDVI) during the period from 1982 to 2000 in China. Great changes in climate variability and an increased trend in NDVI were observed. The changes in precipitation variability were greater than the changes in temperature variability in each month, which is attributed to changes in the monsoon system in East Asia. The abrupt changes in climate and NDVI were more significant in 1983 than in the other years due to the impacts of El Niño/Southern Oscillation (ENSO). Using these results, the influences of changes in climate variability on vegetation were studied in the whole nation, and eight regions were defined according to the vegetation division map of China. The results show that abrupt climate changes at a small scale cannot cause abrupt NDVI changes directly. At a nationwide level, over a longer time scale the persistence of above/below average temperature determines the changes in NDVI; at a shorter time scale, changes in the magnitude of precipitation influence NDVI significantly. Such regional climate variability affects vegetation in different ways owing to the diversity of vegetation types, climatic conditions and topography of the land.<br/>

DOI

[7]
Philippon N, Jarlan L, Martiny N, et al.Characterization of the interannual and intraseasonal variability of west African vegetation between 1982 and 2002 by means of NOAA AVHRR NDVI data[J]. Journal of Climate, 2007,20(7):1202-1218.Abstract The interannual and intraseasonal variability of West African vegetation over the period 1982鈥2002 is studied using the normalized difference vegetation index (NDVI) from the Advanced Very High Resolution Radiometer (AVHRR). The novel independent component analysis (ICA) technique is applied to extract the main modes of the interannual variability of the vegetation, among which two modes are worth describing. The first component (IC1) describes NDVI variability over the Sahel from August to October. A strong photosynthetic activity over the Sahel is related to above-normal convection and rainfall within the intertropical convergence zone (ITCZ) in summertime and is partly associated with colder (warmer) SST in the eastern tropical Pacific (the Mediterranean). The second component (IC2) depicts a dipole pattern between the Sahelian and Guinean regions during the northern summer followed by a southward-propagating signal from October to December. It is associated with a north鈥搒outh dipole in convection and rainfall induced by variations in the latitudinal location of the ITCZ as a response to the occurrence of the tropical Atlantic dipole. The analysis of the intraseasonal variability of the Sahelian vegetation relies on the analysis of the seasonal marches and their main phenological stages. Green-up usually starts in early July and shows a very low year-to-year variability, while senescence ends by mid-November and is prone to larger interannual variability. Six types of vegetative seasonal marches are discriminated according to variations in the timing of phenological stages as well as in the greening intensity. These types appear to be strongly dependent on rainfall distribution and amount, particularly those recorded in late August. Finally, year-to-year memory effects are highlighted: NDVI recorded during the green-up phase in year j appears to be strongly related to the maximum NDVI value recorded at year j 鈭 1.

DOI

[8]
King D M, Skirvin S M, Collins C D H, et al. Assessing vegetation change temporally and spatially in southeastern Arizona[J]. Water Resources Research, 2008,44:W05S15.1] Vegetation species cover and photographic data have been collected at multiple grass- and shrub-dominated sites in 1967, 1994, 1999, and 2005 at the U.S. Department of Agriculture Agricultural Research Service Walnut Gulch Experimental Watershed (WGEW) in southeastern Arizona. This study combines these measurements with meteorological and edaphic information, as well as historic repeat photography from the late 1880s onward and recent satellite imagery to assess vegetation change at WGEW. The results of classification and ordination of repeated transect data showed that WGEW had two main vegetation structural types, shrub dominated and grass dominated. Spatial distribution was closely linked to soil type and variations in annual and August precipitation. Other than the recent appearance of Eragrostis lehmanniana (Lehmann lovegrass) at limited sites in WGEW, little recruitment has taken place in either shrub or grass vegetation types. Effects of recent drought on both vegetation types were apparent in both transect data and enhanced vegetation index data derived from satellite imagery. Historic photos and a better understanding of WGEW geology and geomorphology supported the hypothesis that the shift from grass- to shrub-dominated vegetation occurred substantially before 1967, with considerable spatial variability. This work reaffirmed the value of maintaining long-term data sets for use in assessments of vegetation change.

DOI

[9]
Zhao X, Zhou D J, Fang J Y.Satellite-based studies on large-scale vegetation changes in China[J]. Journal of Integrative Plant Biology, 2012,54(10):713-728.Remotely-sensed vegetation indices,which indicate the density and photosynthetic capacity of vegetation,have been widely used to monitor vegetation dynamics over broad areas.In this paper,we reviewed satellite-based studies on vegetation cover changes,biomass and productivity variations,phenological dynamics,desertification,and grassland degradation in China that occurred over the past 2-3 decades.Our review shows that the satellite-derived index (Normalized Difference Vegetation Index,NDVI) during growing season and the vegetation net primary productivity in major terrestrial ecosystems (for example forests,grasslands,shrubs,and croplands) have significantly increased,while the number of fresh lakes and vegetation coverage in urban regions have experienced a substantial decline.The start of the growing season continually advanced in China's temperate regions until the 1990s,with a large spatial heterogeneity.We also found that the coverage of sparsely-vegetated areas declined,and the NDVI per unit in vegetated areas increased in arid and semi-arid regions because of increased vegetation activity in grassland and oasis areas.However,these results depend strongly not only on the periods chosen for investigation,but also on factors such as data sources,changes in detection methods,and geospatial heterogeneity.Therefore,we should be cautious when applying remote sensing techniques to monitor vegetation structures,functions,and changes.

DOI PMID

[10]
Huang F, Wang P.Vegetation change of ecotone in west of northeast China plain using time-series remote sensing data[J]. Chinese Geographical Science, 2010,20(2):167-175.lt;p>Multi-temporal series of satellite SPOT-VEGETATION normalized difference vegetation index (NDVI) and normalized difference water index (NDWI) data from 1998 to 2007 were used for analyzing vegetation change of the ecotone in the west of the Northeast China Plain. The yearly and monthly maximal values,anomalies and change rates of NDVI and NDWI were calculated to reveal the interannual and seasonal changes in vegetation cover and vegetation water content. Linear regression method was adopted to characterize the trends in vegetation change. The yearly maximal NDVI decreased from 0.41 in 1998 to 0.37 in 2007,implying the decreasing trend of vegetation activity. There was a significant decrease of maximal NDVI in spring and summer over the study period,while an increase trend was observed in autumn. The vegetation-improved regions and vegetation-degraded regions occupied 17.03% and 20.30% of the study area,respectively. The maximal NDWI over growing season dropped by 0.027 in 1998-2007,and about 15.15% of the study area showed a decreasing trend of water content. Vegetation water stress in autumn was better than that in spring. Vegetation cover and water content variations were sensitive to annual precipitation,autumn precipitation and summer temperature. The vegetation degradation trend in this ecotone might be induced by the warm-drying climate especially continuous spring and summer drought in the recent ten years.</p>

DOI

[11]
Cui Y P, Liu J Y, Hu Y F, et al.An analysis of temporal evolution of NDVI in various vegetation-climate regions in Inner Mongolia, China[J]. Procedia Environmental Sciences, 2012,13:1989-1996.Inner Mongolia is an important ecological barrier in northern China. The vegetation coverage and changes in the region will directly affect many important economic regions in China. In this paper, based on the long time series NOAA/AVHRR NDVI dataset and vegetation-climate regions map in Inner Mongolia, we utilized the Mann - Kendall non-parametric test to analyze vegetation change treads in different vegetation-climatic regions from 1982 to 2006. The results showed that different vegetation-climate regions had different vegetation change treads. Therein, Central and Western parts of Inner Mongolia had an ascending trend, while, the eastern part had a descending trend. Inner Mongolia span different climatic regions, and plants lived there had various adaptive capacities under different environments. Therefore, its change trends also were quite different. All these illustrate that, under the background of climate change, the sensibility of vegetation are significant different and some plants in different regions have showed their adaptive capacities.

DOI

[12]
周峰,许有鹏,吕慧华.基于MODIS-EVI数据的长江三角洲地区植被变化的特征[J].长江流域资源与环境,2012,21(11):1363-1369.<p>基于2001~2010年MODISEVI植被指数产品数据，结合国家标准气象站逐月气温、降雨及日照时数资料，对长江三角洲地区植被的时空变化特征进行分析。结果表明：（1）空间分布上，该区域西南部以林地为主，而东北部以农田为主，近10 a植被变化面积占总面积的32%，以农田的转出和城镇的转入为主；（2）区域年最大EVI整体呈减少趋势（-0028/10 a），不同季节下，夏冬季均呈减少趋势（以2月和8月份最为显著），春秋季则呈增加趋势（以5月和10月份最为显著）；（3）不同植被类型下，城镇和农田EVI呈不同程度减少趋势，以城镇EVI下降速度(-0076/10 a）最为显著（R2=077），而林地变化较弱；(4)研究区湿润气候环境下，农田和林地年最大EVI与日照时数和气温多呈正相关性，与降雨多呈负相关性，其中以林地EVI与2~4月份日照时数的正相关性较为显著，城镇EVI与气象因子的关系相对较弱，更多的是受城镇化等人类活动的影响</p>

[13]
韩贵锋,徐建华.上海城市植被变化轨迹及其成因分析[J].生态学报,2009,29(4):1793-1803.通过TM影像解译，获取上海市1989、1997和2003年共3个时期的植被分布，构造植被变化轨迹，分析轨迹的空间格局；选择距离、人口和景观等3类共12个影响因子，使用Logistic回归进行成因分析。研究结果表明：①14年来，上海市的植被面积呈现持续下降趋势，浦东新区植被减少最多；3个时期均为植被的轨迹占总面积的一半以上，其次是3个时期均为非植被的轨迹占总面积的1/5。早期植被转化为非植被的轨迹主要在城区周围，而近期的转化的轨迹发生在距城区较远的地区；植被与非植被的交替变化轨迹，说明了植被并不是一味地被破坏，而是出现了可逆的变化过程。②Logistic回归分析发现，各因子对植被变化轨迹的影响强度大小依次是，离道路的距离＞离行政中心的距离＞离植被-非植被边界的距离＞离商业中心的距离＝1990年人口密度＞离河流的距离＞离高速公路的距离＞土地利用多样性＞2003年与1990年人口密度差＝离上海市中心的距离。③回归模型的精度是满意的，其中二分类Logistic模型精度高于多分类Logistic模型。总体上，离道路的距离对植被变化影响最明显；变化轨迹表明在部分地区植被出现了植被-非植被反复变化的过程；植被分布不但具有空间依赖性，还具有较强的时间依赖性。

[14]
Zhuo G, Li X, Bu L, et al.Satellite dataset analysis of recent vegetation variation in Tibet region[J]. Sciences in Cold and Arid Regions, 2011,3(5):0426-0435.This research investigates the recent distribution variation trends of vegetation in the Tibet region using Normalized Difference Vegetation Index (NDVI) data from 2000 to 2007. It also discusses the causes of vegetation degradation in typical regions (such as Nagqu) based on climatic conditions, human activity, and other influencing factors. Results show that the areas with the best vegetation cover are in Nyingchi and the southern part of Shannan, followed by Chamdo, the Lhasa area, and the eastern part of Nagqu. Vegetation in various regions exhibits significant seasonal differences. The vegetation status has improved in some parts of the Tibet region in the past few years, while the areas with the most serious degradation are in the middle and southern parts of the Nagqu region. On average, distinct vegetation degradation occurred between 2003 and 2006 in the whole Tibet region but vegetation has been increasing since 2006. The vegetation cover in summer basically determines the annual vegetation status. An increase in precipitation and decrease in wind speed generally corresponds to an increase in vegetation cover. The reverse is also true: a decrease in precipitation and increase in wind speed correspond to the decrease in vegetation cover. NDVI is thus positively related to temperature and precipitation but has a negative relation with wind speed. Increasing temperature and decreasing precipitation have led to the present vegetation degradation in Nagqu, and vegetation in all of these regions has been affected by growth of human population, intensified urbanization, livestock overgrazing leading to the proliferation of noxious plants, extraction of underground minerals and alluvial gold, extensive harvesting of traditional Chinese medicinal plants [<i>e.g</i>., <i>Cordyceps sinensis</i> <i>Caladium</i> spp., and saffron crocus (<i>Crocus sativus</i>)], and serious rodent and other pest damage.

[15]
Zhang W, Zhang Y L, Wang Z F, et al.Vegetation change in the Mt. Qomolangma Nature Reserve from 1981 to 2001[J]. Journal of Geographical Sciences, 2007,17(2):152-164.lt;a name="Abs1"></a>Based on the NOAA AVHRR-NDVI data from 1981 to 2001, the digitalized China Vegetation Map (1:1,000,000), DEM, temperature and precipitation data, and field investigation, the spatial patterns and vertical characteristics of natural vegetation changes and their influencing factors in the Mt. Qomolangma Nature Reserve have been studied. The results show that: (1) There is remarkable spatial difference of natural vegetation changes in the Mt. Qomolangma Nature Reserve and stability is the most common status. There are 5.04% of the whole area being seriously degraded, 13.19% slightly degraded, 26.39% slightly improved, 0.97% significantly improved and 54.41% keeping stable. The seriously and slightly degraded areas, which mostly lie in the south of the reserve, are along the national boundaries. The areas of improved vegetation lie in the north of the reserve and the south side of the Yarlung Zangbo River. The stable areas lie between the improved and degraded areas. Degradation decreases with elevation. (2) Degeneration in the Mt. Qomolangma Nature Reserve mostly affects shrubs, needle-leaved forests and mixed forests. (3) The temperature change affects the natural vegetation changes spatially while the integration of temperature changes, slopes and aspects affects the natural vegetation change along the altitude gradients. (4) It is the overuse of resources that leads to the vegetation degeneration in some parts of the Mt. Qomolangma Nature Reserve.

DOI

[16]
罗格平,陈嘻,胡汝骥.基于AVHRR/NOAA影像的天山北坡近10a植被变化[J].冰川冻土,2003,25(2):237-242.提出一套基于遥感和GIS技术的干旱区植被变化和生态环境质量初步评价的研究方法,在天山北坡及其典型区域三工河流域应用取得满意结果.研究表明:1)1992—1998年天山北坡植被覆盖度和生物量持续增加,与同期温度和降水变化趋势一致,从植被角度响应了天山北坡气候向暖湿发展的态势.2)1987—1998年三工河流域从前山带到北部沙漠区植被覆盖度和生物量显著增加,尤其是绿洲外围的北部沙漠区和荒漠过渡带;从增加的幅度看,平原区增幅明显大于山区,后期大于前期;植被指数与6～7a温度、降水的均值呈现出明显的正相关,和6～7a蒸发潜力的均值表现出负相关,三工河流域植被变化显著地响应了6～7a时间尺度的气候变化.

[17]
邓朝平,郭铌,王介民,等.近20余年来西北地区植被变化特征分析[J].冰川冻土,2006,28(5):686-692.利用1982—2003年8 km分辨率的<i>NDVI</i>数据集,选中国西北地区森林、草原、灌溉农业、雨养农业区不同类型植被为研究区,分析了植被年、年际变化特征,并对植被覆盖空间变化进行动态研究.结果表明:森林、草原、灌溉农业区和以春小麦为主的雨养农业区<i>NDVI</i>年变化为单峰型曲线,以冬小麦为主的雨养农业区<i>NDVI</i>曲线呈双峰型;同一类型的植被<i>NDVI</i>受纬度或海拔高度的影响,绿峰出现时间存在1个月的位相差.22 a来森林植被<i>NDVI</i>多呈下降趋势,草原植被区为上升趋势;雨养农业区变化不大,灌溉植被区呈显著的上升趋势.西北东部雨养农业区植被波动频率和幅度最大,是受降水影响最敏感的地区;森林植被次之;有灌溉条件的绿洲植被,年际间波动最小.22 a间西北地区植被以增加趋势为主,增加面积约为20.5%,主要分布在新疆和河西走廊绿洲、黄河沿岸灌区以及青海草区,水分条件充足的绿洲是<i>NDVI</i>增加最显著的区域;<i>NDVI</i>减少地区面积为4.77%,主要分布在西北东部.

[18]
周顺武. 近50年拉萨夏季降水趋势和突变分析[J].高原气象,1999,4(86):35-39.根据１９５２～１９９８年拉萨站夏季（６～８月）月平均降水资料，首先通过线性趋势估计和多 项式拟合等方法分析了拉萨夏季降水长期趋势变化和周期变化骈利用滑动ｔ检验和Ｍａｎｎ－Ｋｅｎｄａｌｌ检验方法讨论了降水突变问题。结果表明：在过去４７ 年里，拉萨夏季降水有明显的下降趋势，降水突变有１９６６年，６０年代末至８０年代末降水明显偏少，９０年代降水出现回升势头。降水周期变化主要集中在高 频的短周期振荡。

[19]
符淙斌,王强.气候突变的定义和检测方法[J].大气科学,1992,16(4):482-493.气候突变现象及其理论的研究是近代气候学一个新兴的研究领域.本文是气候突变研究评述的第一部分,着重讨论了突变,主要是气候突变的定义和气候突变信号的各种检测方法.把气候突变归纳为四类,即均值突变、变率突变、转折突变和翘翘板(seasaw)突变.并通过Mann-Kendall法的检测,发现在本世纪20年代经历了一次全球范围的气候突变.

DOI

[20]
杜海波,吴正方,李明.长春市近57年气候变化及突变分析[J].农业与技术,2010,30(1):52-58.以长春市基准气象站 1951~2007年来的年、月平均降水量和气温资料为基础,利用Mann-Kendall突变检验法和滑动t-检验法,分析了长春市近57年来的气候变 化趋势和突变特点。结果表明:气温明显上升,降水量有减少的趋势;年和四季的平均气温和降水量均存在突变,气候从一个相对较湿冷的平均态转变到一个相对较 干暖的平均态。

DOI

[21]
张建军,周后福,翟菁.合肥气温和降水的突变特征分析[J].安徽农业科学,2007,35(9):2724-2726.基于合肥市1956～2005年的气象观测资料,利用滑动t-检验法、Yamamoto法和Mann-Kendall法对合肥市逐年平均气温的突变特征进 行比较分析;利用滑动t-检验法、Yamamoto法和滑动平均法对逐年降水量的突变特征进行比较分析.结果表明:合肥年平均气温在1986～1988年 和1993年附近发生了突变,年降水量在1968年附近发生了突变.

DOI

[22]
Tucker C J, Pinzon J E, Brown M E, et al.An extended AVHRR 8-km NDVI data set compatible with MODIS and SPOT vegetation NDVI data[J]. International Journal of Remote Sensing, 2005,26(20):4485-5598.Daily daytime Advanced Very High Resolution Radiometer (AVHRR) 4‐km global area coverage data have been processed to produce a Normalized Difference Vegetation Index (NDVI) 8‐km equal‐area dataset from July 1981 through December 2004 for all continents except Antarctica. New features of this dataset include bimonthly composites, NOAA‐9 descending node data from August 1994 to January 1995, volcanic stratospheric aerosol correction for 1982–1984 and 1991–1993, NDVI normalization using empirical mode decomposition/reconstruction to minimize varying solar zenith angle effects introduced by orbital drift, inclusion of data from NOAA‐16 for 2000–2003 and NOAA‐17 for 2003–2004, and a similar dynamic range with the MODIS NDVI. Two NDVI compositing intervals have been produced: a bimonthly global dataset and a 10‐day Africa‐only dataset. Post‐processing review corrected the majority of dropped scan lines, navigation errors, data drop outs, edge‐of‐orbit composite discontinuities, and other artefacts in the composite NDVI data. All data are available from the University of Maryland Global Land Cover Facility ( http://glcf.umiacs.umd.edu/data/gimms/ ).

DOI

Options
Outlines

/

〈 〉