2000—2013年新疆植被覆盖变化多尺度遥感分析  

Multi-scale Remote Sensing Analysis of Vegetation Cover Changes from 2000 to 2013 in Xinjiang

作　　者：隋悦[1,2] 鹿琳琳[1] 张熙[1] 李庆亭[1] 杨锋杰[2] 

机构地区：[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]山东科技大学地球科学与工程学院,山东青岛266590

出　　处：《气象与环境科学》2016年第3期1-8,共8页Meteorological and Environmental Sciences

基　　金：国家自然科学基金项目"基于微波散射计的全球植被物候探测与应用研究"(41471369);国家自然科学基金重大项目课题"月基对地观测传感器优化与观测方法研究"(41590853)资助

摘　　要：基于中等分辨率成像光谱仪(MODIS)8 d 500 m分辨率地表反射率数据生成归一化植被指数(NDVI)时间序列,利用线性回归、转折点检测和Mann-Kendall趋势分析方法,分析2000—2013年新疆地区植被覆盖时空变化格局,并结合Landsat数据分析典型区域植被变化。研究结果表明:近14 a来,新疆地区植被覆盖整体呈波动型上升的趋势,植被改善区面积占全区的34.02%,恶劣区的占3.20%,其中2000—2003年植被明显增长、2003—2009年的波动下降及2009—2013年的逐渐回升,植被增长显著的区域主要分布在准噶尔盆地南部和塔里木盆地北部绿洲;大多数植被类型NDVI呈增长趋势,其中增长率最高的是作物、开放灌丛和混交林,6种主要植被类型常绿针叶林、混交林、开放灌丛、多树草原、草原、作物呈现出相似的NDVI变化趋势;在植被覆盖变化显著的4个典型区,NDVI变化受土地覆盖类型变化的影响,荒漠、草地被开垦成农田导致NDVI增加,城市建成区扩张导致NDVI降低。Based on the Normalized Difference Vegetation Index （NDVI） time series generated from the Moderate Resolution Imaging Spectroradiometer （MODIS） surface albedo data of 8 days 500 m reso-lution, we used linear regression, turning point detection and Mann-Kendall trend analysis to analyze the spatial-temporal pattern of vegetation cover changes from 2000 to 2013 in Xinjiang, and combined the Landsat data to analyze vegetation cover changes in typical areas. The results showed that： during the past 14 years, the vegetation cover in Xinjiang showed a general fluctuant increasing trend. The vegeta- tion improved areas accounted for 34.02% of the whole areas, and deteriorated areas accounted for 3.20%. An obvious increasing trend from 2000 to 2003, a fluctuant decreasing trend from 2003 to 2009 and a gradual increasing trend from 2009 to 2013 were detected. The areas where vegetation cover exhibi- ted an obvious increasing trend were mainly distributed in southern Junggar basin and oasis in northern Tarim basin. The annual mean NDVI increased for most vegetation types, and the increasing rates of croplands, open shrublands and mixed forests were the highest. The six dominant vegetation types, namely evergreen coniferous forest, mixed forest, open shrubland, woody grassland, grassland and crop- land showed similar changing trends of annual mean NDVI. In the four typical areas where the vegetation cover changed obviously, the variation of NDVI was impacted by land cover changes. Transforming from desert and grassland to croplands caused an increasing trend of NDVI. Urban expansion led to a decrea-sing trend of NDVI.

关 键 词：MODIS 归一化植被指数 时间序列 植被覆盖变化 

分 类 号：TP79[自动化与计算机技术—检测技术与自动化装置]