和静草地覆盖度变化遥感监测及分析

Remote sensing monitoring and analysis of grassland coverage in Hejing

选取2004年和2005年4-10月的MODIS 250 m×250 m的遥感影像,在EOS/MODIS资料接收处理系统、卫星遥感应用系统平台以及MAPGIS6.5的支持下,通过预处理、几何校正、归一化植被值指数(NDVI)提取以及最大值合成,主要对巴音郭楞蒙古自治州北部的和静县的植被盖度进行了计算,并按ND-VI值的大小将地表植被分为3类:低植被覆盖,0.1-0.3;中植被覆盖,0.3-0.6;高植被覆盖,0.6-1.0。在C++环境下计算出了不同植被覆盖的面积,并结合2004年7月的实地的样方测量,在SPSS统计软件中将2004年的理论数据与实测数据进行了相关分析,结果表明:在选取的18块样地的54个样方中,在置信度为0.01的条件下,相关系数达到了0.695。说明在新疆利用MODIS数据进行大尺度的植被动态监测较为可行、可信。

MODIS image（250 m× 250 m）from April to October in 2004 and 2005 was selected for grassland coverage calculation of Hejing in Northern Bayinbuluk, supported by EOS/MODIS system, satellite remote sensing application system and MAPGIS6.5 software through pre-processing, geometry emendation, Normalized Difference Vegetation Index （NDVI） extraction and Maximum Value Composition. The ground coverage was divided into 3 groups according to NDVI value, low coverage with 0. 1-0. 3, medium coverage with 0. 3-0. 6 and high coverage with 0. 6-1.0 NDVI. The area of different vegetation was also calculated under C^＋＋ , and correlation analysis between theoretical data and measured data in 2004 was conducted using SPSS integrating with actual measurement in July. The result showed that among 54 samples from 18 sampling sites, correlation coefficient was 0. 695 under 0.01 degree, which indicating that it was feasible and reliable to use Modis for large scal.e vegetation dynamics monitoring in Xinjiang.