中文摘要：

利用1982年至2003年NASAGIMMS逐月归一化植被指数（NDVI）数据集和玛曲县气温、降水资料,对玛曲县近22年来NDVI变化和气候变化特征及其相互关系进行分析,以揭示黄河上游地区植被对全球变暖的区域响应.结果表明：（1）玛曲县植被变化在不同时段表现出较大差异,NDVI年际变化略有增加.（2）夏季是NDVI增长最快的季节,春季NDVI在20世纪90年代后期到本世纪初呈下降趋势,秋、冬季NDVI呈下降趋势.（3）返青期和NDVI值在春季达到同一水平值的时间及夏季达到峰值的时间逐年提前,说明生长季提前是该地区植被对全球变暖的主要响应表现.（4）玛曲县近22年来植被的NDVI变化在中等盖度水平（0.3～0.5）呈增加趋势,高盖度水平（≥0.7）的植被呈下降趋势,而在年代间变化水平上,气温和降水对植被生长都有影响,其中气温要比降水更显著;玛曲县降水和气温在季节尺度上对NDVI的影响不明显,除夏季气温与夏季NDVI关系密切之外,其他季节的关系均不明显.（5）按植被变化特点,将NDVI变化斜率最大值0.005定义为返青期指标,发现该地区牧草返青期的变化主要受温度条件影响,随气候变暖,返青期提前,其变化规律为4～5月份平均气温每10年升高0.6℃,返青期每10年提前3 d.

英文摘要：

The variation characteristics of vegetation and their relationship with climate changes in Maqu County of Yellow river upper reaches area from 1982 to 2003 were examined by using the monthly GIMMS Normalized Difference Vegetation Index （NDVI） data and local meteorological records in order to find out the regional response of the vegetation in upper reaches area of Yellow river to the global warming. The results indicated that there were larger variations in different stage in Maqu County of China. The annual mean of NDVI weakly increased from 1982 to 2003. The NDVI increase speed was the fastest in summer. In spring NDVI increased from 1982 to 2000 and then decreased after 2000. In autumn and winter NDVI decreased since recent 22 years. The same standard both the turning green period and the date of the spring NDVI value brought forward year by year. The date that the summer NDVI value increased to the maximum one was forward from the early 1980＇s to the early 21＇st century. The main way to responding the global warming was that the growing season of the pasture brought forward over the last 22 years. The cover degree in middle standard（0.3-0.5） is increased and that in high standard（≥0.7） is decreased in recent 22 years. There were some influences both the temperature and the precipitation on the NDVI in interdecadal scale, but it was not notably in season scale except the summer temperature was the main factor to impact on the NDVI. In all,temperature was the main factor which influenced on the vegetation growing in high height and cold area. An index was defined to reflect the turning green period using the GIMMS NDVI data~ it was the maximum value which equals 0.05 of the slope of the NDVI variation in term of the vegetation changing. The definition for the turning green period accorded with the laws of the NDVI rising; the turning green period was effected by the temperature and it brought forward with the temperature rising. The tendency of the turning green period was forward 3 d/10a with