延河流域NDVI与主要气候因子的时空相关性研究

The temporal and spatial correlations between NDVI and main climatic factors in Yanhe watershed

基于RS技术提取延河流域2000-2004年的NDVI影像,基于GIS技术生成流域气温和降水的内插分布图,并利用线性回归分析法和空间回归分析法分别对NDVI与气温和降水的时间、空间相关关系进行研究,以期为区域水资源可持续利用和生态环境的可持续发展提供科学依据。结果表明:(1)延河流域的植被NDVI存在明显的时空差异性,时间上,多年平均最高值出现在7月份,约为0.5241,最低值出现在2月份,约为0.0873;空间上,NDVI的高值区主要分布于流域南部的石质低山丘陵沟谷区,低值区主要分布于流域的河谷阶地。(2)在置信度为0.05的水平上,NDVI与气温和降水的时间相关系数分别为0.7041和0.7172,空间相关系数分别为0.7356和0.9364,表明气温和降水对于流域植被的覆盖状况有着重要的影响,而且降水对植被的空间分布影响更为显著。(3)植被NDVI与气温和降水的空间相关性要高于时间相关性。

As an index denoting vegetation cover,NDVI is influenced by many climatic factors such as average air temperature and precipitation.In this paper,NDVI is calculated based on the NOAA/AVHRR data（National Oceanic Atmospheric Administration/ Advanced Very High Resolution Radiometer）,the spatial distribution of temperature and precipitation is simulated based on climate materials in ArcGIS9.2 platform.And then,we analyze the time and spatial correlation between NDVI and temperature or precipitation by using the methods of linear regression and spatial regression,respectively,in order to provide scientific references for sustainable use of water resources and sustainable development of ecological environment.The results show that：（1） Vegetation NDVI presents temporal and spatial differences,and specifically,the mean annual maximum value is 0.5241 occurring in July and the mean annual minimum value is 0.0873 occurring in February;the high vale area lies in the rocky hilly valley in the south watershed,and the low vale area lies in the terrace in river valley.（2） On the confidence level of 0.05,the time correlation coefficients of NDVI-temperature and NDVI-precipitation are 0.7041 and 0.7172,and the spatial correlation coefficients of NDVI-temperature and NDVI-precipitation are 0.7356 and 0.9364,which shows average air temperature and precipitation are the important impact factors for vegetation cover.（3） The spatial correlation coefficients of NDVI-temperature and NDVI-precipitation are higher than the time correlation coefficients of them.