中文摘要：

根据采集自青藏高原东部长江源区内3个采样点的大果圆柏（Sabina tibetica Kom.）树轮样本建立了3个标准化年表,并将3个年表合成为一个总年表RC以代表整个长江源.通过与气候要素间相关分析和响应分析发现,总年表对整个长江源区4～6月平均最高气温响应敏感,负相关性显著,相关系数为-0.57.在此基础上重建了长江源地区1470～2009年的4～6月平均最高气温序列,重建方程稳定可靠,方差解释量32.7％,调整自由度后为31.3％.通过低频变换分析发现,重建序列在过去540年来经历了12次较冷时期和11次暖期,与周边其他气温重建序列以及历史记载比较之后,重建序列的可靠性得到了进一步的证实.空间相关分析发现重建序列对于整个青藏高原地区4～6月平均最高气温变化特征都具有很好的代表性,通过多窗谱和小波分析发现重建序列存在显著2.3a、4.3a、25a以及准48a周期.

英文摘要：

Three Rivers source region located in the hinterland of the Qinghai-Tibet Plateau. It＇s not only a source region of China＇s major water resources, but also the world＇s highest wetland nature reserve. It＇s a scarcely populated area and less affected by human activities, so lots of ancient juniper and spruce trees distributed in this region. Studies have shown that these trees are suitable for the reconstruction of climate factors because the radial growth of trees were sensitive to climate response. In this paper, three tree ring width standardized chronologies were created by Sabina tibetica Kom. which collected from three sampling sites of Yangtze River source region, one of the Three Rivers source. The sampling sites were XRS （33°44′N, 96°14′E; 3908-4050m a.s.l.） YG （33°47′N, 96°11′E; 4355-4455m a.s.l.） and BG （33°46′N, 96°24′E; 3933°4233m a.s.l.） respectively. All of them located in south slope, with the slope 300-45° and the dominant tree species was Sabina tabetic Kom. According to the international conventional process, 162 root tree cores were collected. Because of the three sampling points spaced closer, three standardized chronologies could be merged as RC to represent the entire Yangtze River source region. The climate factors include mean temperature, mean maximum temperature, mean minimum temperature, monthly precipitation, sunshine hours and relative humidity collected from five meteorological stations in the Yangtze River source region. Calculate the arithmetic average of each climate factor in the 5 stations in order to represent the whole Yangtze River source region. Correlation analysis between RC and climate factors founded that mean maximum temperature had a significant negative correlation with RC in April, May and June. Then combine the three months in a certain way and the results showed that there was a significant negative correlation between RC and mean maximum temperature in April-June. The correlation coefficient was -0.57. On this basis, th