文献类型：期刊文献

中文题名：基于TM卫星影像数据的北京市植被变化及其原因分析

英文题名：Driving factor analysis on the vegetation changes derived from the Landsat TM images in Beijing

作者：贾宝全[1,2,3]

第一作者：贾宝全

机构：[1]中国林业科学研究院林业研究所;[2]国家林业局林木培育重点实验室;[3]国家林业局城市林业研究中心

年份：2013

卷号：33

期号：5

起止页码：1654-1666

中文期刊名：生态学报

外文期刊名：Acta Ecologica Sinica

收录：CSTPCD;;Scopus;北大核心:【北大核心2011】；CSCD:【CSCD2013_2014】；

基金：国家科技部"十二五"科技支撑项目(2011BAD38B03)

语种：中文

中文关键词：植被盖度;NDVI差值指数;数量分析;Landsat;TM卫星影像;北京

外文关键词：vegetation coverage; NDVI; difference vegetation value; quantitative analysis; landsat TM image; Beijing

分类号：S757.2

摘要：植被覆盖变化是全球变化研究的重要内容之一,由于NDVI与植被的分布密度呈线性相关,是指示大尺度植被覆盖的良好指标,因此在宏观植被盖度的估算中常被应用。利用1987年9月26日和2009年9月22日的Landsat TM卫星影像,以NDVI为桥梁,分别计算了北京市域的植被盖度和大于0.1的NDVI差值指数,北京市域与不同生态区域两个尺度对其植被变化情况进行了量化分析,结果表明,2009年与1987年相比,北京市极低覆盖度、中覆盖度和高覆盖度植被的面积均有所减少,其所占全市土地面积的比例从1987年到2009年分别降低了5.15%和0.54%和0.03%;而低覆盖度和极高覆盖度植被的面积比例则分别增加了5.71%和0.01%。大于0.1的植被差值指数统计结果显示,全市域植被质量以改善为主,全市植被发生改善变化的土地面积共919302.3 hm2,其中发生轻微改善的比例为28.31%,中度改善的为41.33%,极度改善的面积为30.36%;全市植被发生退化变化的面积326931.12 hm2,其中发生中度退化、轻微退化和极度退化的面积分别占到了退化变化土地面积的41.98%、43.20%和14.82%。从不同区域的植被差值指数看,植被发生退化变化最明显的区域为燕山山区北部、五环以内和五至六环间区域,这几个区域退化变化的植被面积占相应区域的面积比例分别达到了30.25%、58.17%和47.38%,而且均以严重退化与中度退化为主,两者合计的面积比例分别为15.79%、44.72%和34.19%。而发生退化变化面积比例最小的区域为太行山区和延庆盆地,其退化面积占该区域植被面积的比例分别为13.35%和17.02%,且退化程度均以轻微退化和中度退化为主,其面积比例介于5%—8%之间。从植被变化的驱动力看,目前还看不出北京这种植被变化结果与气候变化之间的直接关联。北京市植被变化的驱动力主要还是人为因素。这包括了区域性的大环境绿化生态工程建设(包括山区与平原区),城市绿化市政工程建设、平原区农业结构调整、新农村生态环境建设,以及由于降水而导致的山区河岸带变化等。其中河流水面变化对河岸带植被变化的影响范围在多年平均水面线外0—150 m范围内,0—100 m范围为受影响较大的区域。
The normalized difference vegetation index （NVDI）, linearly correlated with density distribution of vegetation, is widely used for investigating macro-scale vegetation coverage and thus vegetation changes at short and long time scales. We calculated the vegetation coverage and NDVI greater than 0.1 by the Landsat TM satellite images on Sept. 26, 1987 and Sept. 22, 2009 for Beijing, China. The driving forces for vegetation changes were also analyzed including climate, geomorphology, human activities. In addition, the buffer analysis tool in the ARCGIS was used to derive the vegetation changes along the rivers as well as around the reservoirs and lakes. The quantitative analysis indicated that the land area of extremely low coverage, medium coverage and high coverage decreased by 5.15% , 0.54% and 0.03% respectively, while the area of vegetation of low coverage and extremely high coverage increased by 5.71% and 0.01%, respectively from 1987 to 2009 for the whole region of Beijing. The statistical results in NDVI greater than 0.1 showed that the vegetation quality of the whole city had basically improved and improved area reached 919302.3 ha. The ratio of slightly, moderately and extremely improved area was 28.31% , 41.33% and 30.36%, respectively. In addition, the degraded area was identified as 326931.12 ha. Considering NDVIs for different sub-regions of the city, the most noticeable vegetation degradation took place within the 5th ring and between the 5th ring and the 6th ring and the percentage of the area of degraded vegetation was 58.17% and 47.38%, respectively for these two 2 regions. Contrarily, the regions with minimum percentage of vegetation degradation included Taihang Mountain and Yanqing Basin and the vegetation degradation area was 13.35% and 17.02%, respectively. Furthermore, the degradation for the Taihang Mountain and Yangqing Basin was basically attributable to slightly moderate degradation. There was no significant correlation between vegetation changes and climate changes both for whole Beijing and its sub-regions. However, human activities were found to be the major driving forces for vegetation changes in Beijing including regional green ecological restoration projects for both mountains and plains）, urban landscape and greening projects, agricuhural structure amendment for plain regions, and eco-environment restoration efforts for rural regions. In addition, water bodies had a very important effect on the vegetation changes along the river and around the reservoirs and lakes. A massive construction of reservoirs in the mountain region resulted in the no flow phenomena for the rivers in plain regions, thus, water body effects on the vegetation occurred in the mountain area. NDVI and vegetation coverage changes were buffered by 0--50m, 50--100m, 100--150m, 150--200m and 200--250m. We found that the buffer distance can be divided into two spatial levels for NDVI calculation. NDVIs buffered by 0--100m were enhanced over 0.1, but the NDVIs buffered by 100--250m increased by 0. 0084--0. 0089 from 1987 to 2009. We also found that the 0-- 250m buffer ranges also can be divided into two levels for vegetation coverage analysis. Compared with 1987, the ratio of vegetation coverage in 2009 was reduced for the buffer of 0--200m and increased for the buffer of 200--250m. Therefore, the changes of river water body caused by precipitation can influence riparian vegetation within 150 m from the average water level, especially within 100 m.