树轮记录的小冰期以来青藏高原气候变化的时空特征

SPATIAL AND TEMPORAL PATTERNS OF CLIMATE VARIATIONS OVER THE TIBETAN PLATEAU DURING THE PERIOD 1300～2010

小冰期气候为评估现代气候变化提供了最直接的背景。本文主要依据树轮资料,同时结合现代仪器观测记录,利用经验正交函数(EOF)方法探讨了青藏高原小冰期以来气候变化的时空特征。首先分析了高原的温度变化。近50年来青藏高原的温度变化基本同相位变化,没有明显的区域差异;乌兰树轮序列是青藏高原的冬半年(9～4月)温度代用指标;利用6条指示夏季或暖季温度变化的树轮序列合并形成的新序列可指示高原春季-夏季(3～8月)温度变化;依据RCS(区域曲线标准化)方法建立的昌都树轮序列,能够反映整个青藏高原的年平均温度变化;不同季节的温度重建序列均显示17世纪和18世纪20～70年代是高原小冰期气候寒冷的时期,而18世纪初,19世纪后半叶,20世纪中期的气候较温暖,且均显示20世纪末期气候的快速增暖事实。其次,从重建的亚洲区域夏季PDSI(Palmer Drought Severity Index)网格化数据集中提取42个网格点数据,分析了过去700年(1300～2005A.D.)高原的湿度变化。发现前3个特征向量代表了高原过去湿度变化的主要空间模态,与利用器测降水记录展开的EOF模态基本一致,表明主导高原干湿变化时空差异的物理过程是稳定的,不随时间而变化;近700年来高原南北部湿度变化具有明显的区域差异,最显著的差别是:自20世纪中期以来高原北部存在明显的变湿趋势,而高原南部却恰恰相反,呈现逐渐变干的趋势;近700年来高原南部的干湿变化有超前于高原北部的趋势。

The climate conditions during the Little Ice Age （LIA）provide immediate background for assessing current climate warming. In this paper ,the spatial and temporal climate variability on the Tibetan Plateau（TP） since the LIA was studied by applying the empirical orthogonal function （EOF）technique to tree-ring chronologies and instrumental meteorological data. Firstly, temperature variations on the TP were analyzed. It was found that temperature variations are consistent in different parts of the TP and there is no obvious regional difference during the recent 50 years. Spatial correlation between Wulan tree-ring index and gridded temperature data from CRU TS3. 1 during the period 1955 ~ 2003 showed that the Wulan series can be regarded as a reasonable indicator of winter half year （September-April） temperature and reflected the temperature variability for the whole of TP. A new temperature composite reconstruction was derived by combining six summer/warm temperature-sensitive tree-ring width chronologies. This new composite is well indicative of spring-summer（March-August） temperature variations over the TP. In addition, an updated well-replicated tree-ring width record covering the period A. D. 1000 -2010 was developed for Qamdo region based on RCS （regional curve standardization） method. Spatial correlation between tree ring width and observed temperature data indicates that the Qamdo RCS chronology is representative of mean annual （January- December prior to the growth year） temperature variations for the whole TP. According to this reconstruction, regional warm conditions occurred the 14th century,the early 18th century, the 19th century and mid-20th century, while cold periods were identified during the 15th century,the 17th century, 1720s -1770s,the early 20th century and 1960s - 1980s. The climate warming since 1990s is rapid but remains within the range of natural variability during the past 700 years. Comparison of temperature reconstructions in different seasons shows that common cold periods occurred during the 17th century and the 1720s ~ 1770s,whereas warm intervals were found during the early 1800s,the late 1900s and the middle 20th century. All temperature curves display the rapid warming during the late 20th century. Secondly,the moisture variability on the TP was discussed for the past 700（ A. D. 1500 -2005）on the basis of the extracted 42 gridded PDSI （Palmer Drought Severity Index） dataset from monsoonal Asia. The first three EOF principal components represent the spatial moisture variations over the past 700 years. The EOF analysis of instrumental summer（June-August）precipitation data was also carried out over the TP. It was found that the EOF analysis results of instrumental data are similar to those derived from the gridded PDSI data. These similarities indicate that the physical process controlling the spatial-temporal patterns of the wet/dry variability on the TP was stable and did not change over time. There are regional differences in moisture variability between the northern and southern TP over the past 700 years. A pluvial condition since the mid-20th century occurred in the northern part of the TP, while a drying trend was found on the southern TP. Furthermore, there is obvious phase shift in moisture change between the southern and northern TP. Cross-correlations showed that the wet/dry variation in the southern TP is 17 years earlier than that in the northern TP. If it is true,we can project future climate change for the northern TP based on current climate conditions on the southern TP.