【作者】 李芳；

【导师】 蒋志荣；

【作者基本信息】 甘肃农业大学 ， 水土保持与荒漠化防治， 2011， 硕士

【摘要】 本文基于MODIS-NDVI数据,运用植被重心法、一元线性回归法以及R/S分析法对张掖地区2000~2008年的204幅遥感图像进行植被覆盖动态分析,并运用相关分析法分析近9年植被NDVI与气温、降水、日照时数和蒸发量的相关性和滞后性。研究结果表明:(1)张掖市植被AMNDVI和ANNDVI的年际变化表现为增加的趋势,AMNDVI>0.1的植被覆盖2003年和2007年为界呈不稳定上升–稳定极小上升–不稳定下降的变化;年内变化表现出很强的季节性,4月NDVI急剧上升,6、7月达到最大值,9月急剧下降,四季的变化表现为夏季和秋季的增加和冬季和春季的减小。(2)2000~2008年植被重心变化表现为:以2002年和2006年为界呈不稳定–稳定–不稳定变化。2000~2003年植被重心向东移动,说明东部植被增加或西部植被减少;2006~2008年向北移动,说明北部植被增加或南部植被减少。年内变化得出:以5月和11月为界的不稳定–稳定–不稳定变化。1~3月向东移动,3~5月向西北变化,5~11月相对稳定,但表现为顺时针变化趋势,11~翌年1月向西南回归。这是温度和降水共同作用的结果。(3)2000~2008年张掖市NDVI的AMNDVI空间变化趋势:整体改善,局部退化。退化面积占研究区总面积的19.7%,分布在南部祁连山地区。生长季的年际变化表现为以植被改善为主,明显改善区主要分布在山丹、民乐和张掖盆地,主要土地类型为旱地、人工草地、改良草地等;退化区主要是张掖盆地和祁连山地;基本不变区域为北部的戈壁荒漠;生长季的各月份植被覆盖变化表现出很大的差异性:5月以张掖盆地为代表的植被明显退化;6月、7月是以南部祁连山地为代表的植被中度退化;8月、9月表现为较5月张掖盆地退化有所改善,较6月、7月祁连山地的退化有所改善。生长季不同土地利用类型的slope值的排序为:草地(改良草地和人工草地)>耕地>城镇及工矿用地>果园>林地>未利用土地,表明人工植被增长速度高于天然植被。(4)生长季平均值的Hurst指数值以弱持续性和反持续性为主。持续性较强分布在张掖盆地,反持续性较强则零星分布在东南部祁连山地。生长季各月的Hurst指数分布差异较大,由5月、6月份的以持续性趋势为主到7月份的持续性减弱,反持续性增强,再到8月、9月份的反持续性趋势减弱,持续性增强。各土地类型的Hurst值排序为:其他未利用土地<建设用地<草地<耕地<未利用土地<林地<水体。(5)张掖市气象因子的年际变化:气温、降水呈增长趋势;日照时数呈双“V”型变化;蒸发量表现为6县区都呈下降变化。张掖市各气象因子的年内变化以7月份最为显著,表现为平均气温的显著降低、降水量的极显著增加、日照时数和蒸发量的显著减少。(6)NDVI与气象因子的年际相关性分析得出:针对简单相关系数的排序为:平均气温温>蒸发量>降水量>日照时数;在偏相关分析中偏相关系数的排序为:蒸发量>平均气温>降水量,说明在剔除气温和降水对NDVI的影响外,NDVI对蒸发量的响应最明显的。(7)NDVI与张掖各县区四季的年内相关性分析得出:张掖市大多数县区的NDVI与气温的响应主要在冬季,甘州区则以春季最明显,但随着同期气温的降低伴随着NDVI的降低;降水主要表现为夏季相关性的显著或极显著,以张掖和临泽为最高正相关;日照时数与NDVI的相关性较差,只有民乐夏季的相关性达到显著相关。NDVI与蒸发量的相关性主要以夏季和冬季显著为主,其中夏季主要表现为中度负相关,冬季表现为中度正相关。(8)NDVI对于气象因子的滞后性研究表明:NDVI对气温和降水的滞后期为0个月,即NDVI与气温和降水的相关系数以同期为最高,除民乐县外。日照时数其滞后期表现为1~2月;蒸发量的滞后期表现为东部地区滞后两个月,西部地区滞后一个月。更多还原

【Abstract】 In this paper, the temporal and spatial variation of NDVI of vegetation in Zhang-ye region during 2000~2008 is analyzed by using the gravity centre of vegetation, linear regression and R/S analysis based on the MODIS-NDVI data and it also analyzed the correlation and hysteretic nature between vegetation NDVI and climate factors of temperature, precipitation, sunshine hours and evaporation. The results showed that:(1).Annual variation of AMNDVI and ANNDVI in Zhang-ye region presented the trend of increasing. The vegetation cover of AMNDVI>0.1 increases unstable from 2000 to 2003, increases stable a little from 2003 to 2007 and decreases unstable from 2007 to 2008. The intra-annual variation exhibits stronger seasonal, that is, NDVI increased rapidly in April, NDVI reached the maximum value in June or July and decreased in September. Seasonal variation indicated that vegetation NDVI increased in summer and autumn and decreased in winter and spring.(2).The gravity centre of vegetation moving Eastward during 2000~2003indicated that vegetation increased in East or decreased in West; the gravity centre of vegetation moved Northward during 2006~2008, which revealed the vegetation increased in North or decreased in South. The intra-annual variations indicated that:the gravity centre of vegetation was unstable from January to May and from November to January, but stable from May to November. The gravity centre of vegetation moved Eastern from January to March and moved Northwestern from March to May, but it presented stable from May to November and the trend of clockwise changing, and it moved Southwestern from November to January. This is because the interaction of the temperature and precipitation(3).The spatial variation of AMNDVI in Zhang-ye region from 2000 to 2008 indicated: improve integrally, part-degenerate. The degradation region of vegetation, which took up 20% of the whole researching area, distributed in Qilian Mountain in South. The annual variation of growing season is mainly improvement. The increasing region of vegetation, distributes in Shan-dan, Min-le and Zhang-ye basin. The main land types were dry land and artificial grassland; artificial vegetation increased faster than natural vegetation. Degradation regions sporadically distributed over Zhang-ye basin and Qilian mountain due to much human activity. The region of almost invariant is gobi desert. The Slope of each month in growing season was different. The Slope showed significant degradation in Zhang-ye basin in May, middle-degraded in the South of Qilian Mountain in June and July and in in August and September the Slope of vegetation is better than the ones in May in Zhang-ye basin and in June or July in the South of Qilian Mountain. The order of Slope value of different land types in growing season was: grassland (dry land and artificial grassland) > cultivated land > Construction land> forest > unused land, and it indicated that artificial vegetation growing faster than that of natural vegetation.(4).The Hurst index of growing season of NDVI focused on 0.4~0.6, which dominated by weak sustainability and anti-sustainability, but the index of each month in growing season was different. The Hurst index showed the decreasing trend of persistence and increasing reverse- persistence from May to July, while the persistence was increased and reverse- persistence was decreased in August and September. The order of Hurst value of different land types was: Other unused land< Construction land<grass<cultivated land<unused land<forest.(5).The annual variation of climate factors in Zhang-ye region showed: temperature and precipitation presented the trend of increasing; the changes of sunshine hours manifested as double“V”curve; the change of evaporation decreased from 2000 to 2008 in Zhang-ye region. The intra-annual variation of climate factors showed that the changes of climate factors are most significant in July. It presented that the temperature, sunshine hours and evaporation decreased significantly, and precipitation increased most significantly.(6).The annual correlation coefficient between NDVI and climate factors showed that: the order of Pearson correlation coefficient is average temperature>evaporation> precipitation>sunshine hours. The order of partial correlation coefficient is evaporation> average temperature>precipitation. It showed that when rejecting the influence of average temperature and precipitation to NDVI, The response of NDVI to evaporation is most significant. (7).The correlation between NDVI and climate factors in different county or region of Zhang-ye showed that: the response of NDVI to temperature occurred in winter in most regions, but the one of Gan-zhou appeared in spring. The positive correlation between NDVI and precipitation appeared in summer which is most significant, particular maximum in Gan-zhou and Lin-ze. The correlation between sunshine hours and NDVI is weaker, not including Min-le. The correlation between NDVI and evaporation appeared in summer which is middle negative correlation and in winter which is middle positive correlation.(8).The hysteretic nature of NDVI to climate factors showed that: NDVI to temperature and precipitation have not lag phase, meaning the correlation of synchronization is maximum, not including Min-le. The lag phase of sunshine hours is one or two months. The lag phase of evaporation presents two months in eastern and one month in western.更多还原