【作者】 李茜；

【导师】 李栋梁； 魏凤英；

【作者基本信息】 南京信息工程大学 ， 气象学， 2011， 博士

【摘要】 东亚夏季风的强弱、进退活动对中国东部地区夏季降水有很大的影响,因此,研究东亚夏季风的长期变化规律是我国气候变化研究的重要课题之一。然而,由于目前获得的器测资料大多只有50-60年,限制了从长时间尺度的角度对东亚夏季风变化的研究。本文首先比较了近50年我国东部地区旱涝等级和器测降水量时空分布特征的异同。然后在分析1850-2008年我国东部旱涝等级与东亚地区海平面气压关系的基础上,利用宝贵的历史气候代用资料(旱涝等级、干湿指数等),基于海陆热力差异驱动夏季风从而影响中国东部地区夏季旱涝分布这一物理机制出发,分别重建出1470-2008年、960-2000年的东亚夏季海平面气压场。在此基础上建立了近500年、近千年来的东亚夏季风指数,并分析了它们的变化特征及其与中国东部旱涝分布的关系。另外,对重建的海平面气压与数值模拟及重建的亚洲-太平洋涛动指数(APO)进行了比较。主要结论如下：(1)旱涝等级的典型空间分布和年代际变化与近50年器测的5-9月降水量的典型空间分布和年代际变化基本一致,表明旱涝等级代用资料可以反映中国东部旱涝变化的主要特征。本文定义的1850-2008年的东亚夏季风指数的年代际变化与中国东部旱涝典型分布型的年代际变化有密切关系,但两者的相关关系并不是稳定不变的,而是存在显著的年代际位相差异。(2)基于1850-2008年我国东部旱涝分布与东亚海平面气压关键区域的显著相关关系,重建出1470-2008年东亚夏季海平面气压场。交叉检验结果表明,东亚夏季风关键区重建效果相对于其它地区要好。分析表明,1470-2008年重建的东亚夏季海平面气压场可以将高纬与中纬地区大尺度反向的气压涛动及海洋与陆地热力差异的突出模态显现出来。海陆热力差异驱动夏季风,其中青藏高原地区是主要的气压活动中心之一,在东亚夏季风的形成、进退或扩展中起重要作用。(3)重建与数值模拟的海平面气压时空分布特征的比较表明,1470-1999年FGOALS_gl数值模拟的东亚夏季海平面气压主要体现了海陆气压差异。根据重建的东亚夏季海平面气压场定义的1470-2008年东亚夏季风指数的演变具有明显的阶段性,16世纪中期到17世纪初东亚夏季风偏强,17世纪偏弱,18世纪经历了“弱-强-弱-强-弱”的变化,19世纪则是“强-弱-强-弱”的变化,20世纪是明显的“弱-强-弱”变化。而1470-1999年数值模拟的东亚夏季风指数序列与重建序列的主要差异出现在16世纪末和18世纪末,两者的变化趋势相反,其它时段的变化趋势基本一致。(4)REOF分析表明,中国东部地区旱涝空间分布大致可以分为6个区域：东北区、华北区、江淮区、华南区、西北区和西南区。它们与东亚夏季风的对应关系是：东亚夏季风偏弱时,华北偏旱、江淮偏涝、华南偏旱；东亚夏季风偏强时,华北偏涝、江淮偏旱、华南偏涝。华北区和江淮区的旱涝变化对东亚夏季风强弱变化的响应最为明显。根据我国东部六区干湿指数与我国华北区和江淮区旱涝等级具有很好的一致性及近530年东亚夏季海平面气压场与中国东部六区干湿指数具有的密切关系,重建了960-2000年的东亚夏季海平面气压格点场,同时进行了重建效果检验。结果表明,东亚夏季海平面气压重建场在东亚夏季风关键区重建的效果较好,而青藏高原以西的地区重建效果较差。(5)从近千年的时间尺度考察,东亚夏季风的强度指数主要经历过9次明显的趋势突变。其中,13世纪30年代东亚夏季风的强度指数经历了一次最显著的振动。(6)1470-1985年APO指数的变化趋势具有明显的阶段性,即1470-1630年APO指数处于偏弱时期,1631-1796年处于偏强期,1797-1920年处于偏弱期,1921-1985年处于偏强期。从60-80年的周期分量上考察,APO指数和东亚夏季海平面气压场的年代际变化趋势在1797年以前存在差异,而1797年之后两者的年代际变化趋势较为一致。更多还原

【Abstract】 The activities (strength, course and so on) of East Asian summer monsoon (EASM) have great influence on the summer precipitation over eastern China. So, it is important to study the long-term variation of EASM. Due to the observe data only have 50-60 period, the study of the long-term variation of EASM is restricted. This paper analyzed the coupling relationship between the drought/flood grades and the East Asian summer sea-level pressure (SLP) by using the history of climate proxy data (for example drought/flood grades, dry/wet series and so on) and summer average SLP. Based on this relationship, the East Asian summer SLP are reconstructed over the past 500 or 1000 years. According to the reconstructed EASM indices, the long-term evolution of EASM and its associations with drought/flood distribution in eastern China are discussed. Besides, the reconstructed SLP, simulated SLP and Asian Pacific Oscillation (APO) are compared. The mainly conclusions are as follows:(1) The typical spatial distribution and interdecadal variations of drought/flood grads in the last 159 years and precipitation in the last 50 years over eastern China are basically the same, which indicate that drought/flood grads can reflect the main features of drought/flood changes over eastern China. The interdecadal variation of drought/flood distribution patterns over eastern China had a close relationship with that of the East Asian summer monsoon defined in this study, although their correlation was not stable and had a significant difference in interdecadal phase change.(2) Based on the closely relationship between drought/flood distribution over eastern China and the key areas of the EASM (i.e. China and some part of the Northwest Pacific) during 1850-2008, the SLP is reconstructed back to 1470. The tests show that the key areas of the EASM are more credible than the other areas for the reconstructed SLP. The analysis shows that the primary patterns of the reconstructed SLP were differences between high-latitude area and middle-latitude area as well as differences between sea and land during 1470-2008. The EASM is drived by the heat difference between sea and land. As one of the mainly air-pressure activity center, the Tibet Plateau plays an important role on the activities of EASM(3) Compared the characteristic of time/spatial distribution between the reconstructed SLP and simulated, the results show that the primary patterns of the stimulated SLP are differences between sea and land during 1470-1999. The EASM defined by the reconstructed SLP shows obviously periodically. The strength of reconstructed EASM is stronger during mid-16th century and the early 17th century and subsequently weaker during 17th century. The EASM experienced 4 modulations from weaker start during 18th century and 3 changes from weaker to stronger during 19th century. During 20th century, the EASM has obvious change of the "weaker-stronger-weaker". The primary differences between the reconstructed and simulated EASM appeared at the end of 16th and 18th century, and the trend is basically consistent in the other times.(4) The active centers of six drought/flood grads are defined by using the large contours of the six REOF modes in eastern China. They are Northeast China, North China, Changjiang River-huaihe River valley, South China, Northwest China and Southwest China. The response of EASM is obviously in South China and Changjiang River-huaihe River valley, and instability in Northwest China and Southwest China. The stronger strength of EASM may result in less rainfall over South China and North China, and more rainfall over Changjiang River-huaihe River valley. The weaker strength of EASM may result in more rainfall over South China and North China, and less rainfall over Changjiang River-huaihe River valley. The six regional dry/wet series and the drought/flood grades over South China and Changjiang River-huaihe River valley are basically the same during the last 530a. Then, based on the closely relationship between SLP and dry/wet series, the grid data of Eastern Asian summer SLP is reconstructed. The results show that the key areas of the EASM are more credible than the other areas for the reconstructed SLP during the last 1000a.(5) From the millennium time scale, the trend of EASM index experienced nine times turning. The most significant vibration of the EASM strength is occurred in 1330s.(6) The evolution trend of APO index has obviously characteristics of stage, i.e. the APO is stronger than normal during 1631-1796 and 1921-1985 and it is weaker than normal during 1470-1630 and 1797-1920. On 60-80 periodic components, the interdecadal evolution trend between APO and SLP has difference before 1797 and accordance after 1797.更多还原