为了研究气候变化对西北地区不同类型植被的影响,利用NASA GIMMS 1982~2003年逐月归一化植被指数(Normalized difference vegetation index, NDVI)数据集和西北地区138个气象站点同期的气温和降水资料,计算了各站22年月平均气温和降水与NDVI的相关系数。同时,选西北地区森林、草原、绿洲和雨养农业4类有代表性的植被类型为研究区,对各类植被NDVI与气温和降水的相关关系进行分析。研究结果表明:除无植被的戈壁沙漠地区外,西北地区NDVI与气温和降水均有较好的相关性。除祁连山中部地区外,西北地区NDVI与气温的相关系数大于降水。天山、阿尔泰山和秦岭的NDVI与气温相关系数最高,而青海东北部NDVI与降水相关系数最高。西北地区各种类型植被对气候变化反映敏感。其敏感度因植被类型不同和同类植被所处的地理位置不同而有差异;纬度较高的新疆林区与温度相关性最高,高寒草甸次之。在植被生长最旺盛的夏季(6~8月),22年来西北地区各林区的NDVI均呈下降趋势。其中西北东部林区下降趋势显著,与这些地区的降水减少和气温增加有关。草原区植被以上升趋势为主,高寒草甸和盐生草甸上升趋势最为显著,气温升高是这些地区植被生长加速的原因之一。西北绿洲是NDVI增加极为显著的地区,以新疆绿洲NDVI上升趋势最大。气候变暖是近年绿洲NDVI增加的主要驱动力之一,绿洲面积扩大、种植结构调整和种植品种变化等人为因素对绿洲NDVI增加的作用不可忽视,这种作用在新疆表现的尤为突出。雨养农业区NDVI年际间波动较大,各区域间变化不太一致。NDVI的波动与降水变化有很好的正相关,与气温变化有很好的负相关,近年来西北东部气温升高和降水的减少是雨养农业区NDVI下降的原因,农业措施的实施也改变了植被生长对气候条件的依赖性。

Aims We sought to understand the impacts of climate change on vegetation in Northwest China and the relationship between normalized difference vegetation index (NDVI) and climate elements.
Methods Correlation analyses were done using the GIMMS NDVI data and monthly mean temperature and precipitation data from January 1982 to December 2003. We selected different regions in Northwest China, representing major types of vegetation, such as forest, grassland, oasis, and rain-fed cropland, for detailed study.
Important findings We found strong correlations between NDVI and temperature/precipitation, except for the Gobi and other desert areas. Correlation coefficients of NDVI and temperature are higher than them of NDVI and precipitation in almost all regions, particularly for the Hexi Corridor and most of the Xinjiang area. During the vegetation growth period, temperature has greater effect on the various types of vegetation than precipitation. The forests in the higher latitude of Xinjiang area are most sensitive to temperature. This sensitivity reduces in sequence from forests to oases, grasslands and unirrigated croplands. Grasslands are most sensitive to precipitation. The sensitivity to precipitation decreases from grasslands to forests, unirrigated croplands, and oases. In summer, the NDVI of forest decreased during the last 22 years, especially forest in the eastern portion of the northwest. This was related to decreases of precipitation and increases of temperature in these areas. The NDVI in most grassland increased. The trends were significant for high cold meadow and halophytic meadow. Climate warming is the main reason for grass growth speeding up. For oases, the NDVI increases were the most significant. The trends were the highest in Xinjiang oasis. Climate warming was one of the factors driving increases in NDVI. The impact of human activities, such as oasis expanding, crop structure change and crop varieties on the NDVI variation cannot be ignored. The NDVI interannual change was high and varied among the unirrigated croplands. NDVI had a strong positive correlation with precipitation and a negative correlation with temperature. The temperature increase and the precipitation decrease caused the NDVI decrease in these areas.