近百年来气候突变与极端事件的检测与归因的初步研究

【作者】 张文；

【导师】 封国林；

【作者基本信息】 扬州大学 ， 理论物理， 2007， 硕士

【摘要】 本文首先基于标准化后的高分辨率气候代用资料,应用高阶矩分析方法检测近2000年来气候极端异常演变特征;同时结合滤波方法进行具有物理背景的层次分离,进而研究了各时间层次气候极端异常变化信息及其贡献。结果表明:(1)在100年以上的时间层次上,可能存在千年左右的气候变化振荡周期,而且20世纪是近2000年来气候极端异常现象最为活跃的时段,可能对应于气候极端异常现象活跃期。(2)对于20-60年这一时间层次,公元300-1100年间气候极端异常现象比较明显,而公元1100-1980年间相对比较缓和;该层次对20世纪的气候异常没有显著贡献。世纪以上和20-60年时间层次均揭示出在近2000年的气候变化中,公元1100年前后可能是一个气候极端异常现象演变的关键转折时期。(3)在年际尺度上(小于20年),北京石花洞石笋微层厚度时间序列中发生气候极端异常现象的年份与出现E1 Nio事件和La Nia事件的年份有非常好的对应关系(本文仅讨论公元1960-1980年)。(4)高阶矩分析方法对于检测气候极端异常分布及演变规律有较好的应用前景。由于气候突变往往和极端气候事件联系在一起,因此本文的第二部分工作是将青藏高原古里雅冰芯冰川累积量时间序列和各自然因素资料(地球自传速度、太阳黑子活动以及ENSO信号)等四种时间序列进行具有物理背景的分层,并在此基础上应用滑动t-检验对滤波前后的资料进行再分析,旨在探讨自然因素对近百年来气候突变的贡献。检测和分析结果表明,青藏高原古里雅冰芯冰川累积量序列具有7-9年的准周期演变特征,对于20世纪60年代初期和70年代末80年代初的气候突变,是由10a以下和10a以上两个层次共同作用的结果,换而言之,近百年来这两个年代际的突变是全局性的。ENSO和太阳黑子的年际变化是近百年来气候突变的重要激励信号。最后,我们运用信息传递方法初步判断了ENSO信号和太阳黑子活动对近百年来气候突变的贡献大小,结果发现ENSO的年际变化的贡献较大。在上述工作的基础上,硕士论文的第三部分工作是:首先应用传统的检测突变的方法对青藏高原古里雅冰芯序列(冰川累积量)进行分析,旨在探索得到各检测方法之间的联系与区别;其次将检测出的极端事件与检测到的突变进行分析,揭示两者之间的联系与区别。研究表明:(1)对于突变的检测来说,突变年份将会随着所取的子序列长度的不同而发生漂移,并且所选择的子序列的长度较长者检测到的突变的年份相对较少些,反之则较多些。但对于公认的比较明显的突变,两种情况下都有反映。(2)滑动t-检验与Yamamoto法检测得到的突变的年份是一致的。这可能是由于虽然两种方法判别的标准不同,但他们考察的角度是一致的,即都是从考察两子序列的均值差异是否显著来判别突变的。Crammer’s法检测到的突变的年份只是滑动t-检验或Yamamoto法检测得到的突变年份中的一部分,而Mann-Kendall法只能检测到突变开始的年份。(3)应用同一种检测突变的方法分别对古里雅冰芯冰川累积量序列以及古里雅冰芯冰川累积量的距平序列进行分析,所得到的结果几乎相同。可能的原因是趋势的影响对各传统的检测突变的方法来说可以被忽略。(4)极端事件一般发生在两次突变事件之间。更多还原

【Abstract】 Based on the normalized high-resolution climatic proxy data, high-order moment method is used to detect the climate extreme anomalies in recent 2000 years in this paper first. Meanwhile, combining with filtering method to abstract the dominant components with physical backgrounds from the normalized data, then, analyze the information and its contribution of each component of the climate extreme events. The results show that: (1) On the timescale of more than 100 years, climatic oscillation with the period of about 1000 years may exits; besides, the 20th century is the most active period of climate extreme anomaly phenomenon in recent 2000 years, it may correspond to an active period of climate extreme anomaly phenomenon. (2) On 20-60-year timescale, climate extreme anomaly phenomenon in the period of A.D. 300-1100a is relatively distinctive and that in time span of A.D.1100-1980a is comparatively mitigative. Possibly, this level makes no considerable contribution to the climatic anomaly of 20th century. This level and the century scale both reflect that for climatic change in recent 2000 years, the year around A.D. 1100a maybe the key period of climate transition. (3) On the interannual timescale (less than 20years), the years of occurrences of climatic extreme anomalies reflected in Beijing Shihua cave stalagmite proxy records has good correspondence with that of the E1 Ni?o or La Ni?a events (in this paper, only the period of 1960-1980a considered). (4) high-order moment method has a good prospect in detecting the climatic extreme events.As abrupt climate changed and extreme events are often linked together. The second parts of this paper is that: based on physical backgrounds, the four time series of the Gulia (Qinghai-Xizang plateau) ice core (GIC)δ18O, and natural factors, i.e. the rotation rate of earth, sunspots, and ENSO signals are decomposed into two hierarchies of more and less than10-year, and then the running t test is used to reanalyze the data before and after filtering on purpose to investigate the contribution of natural factors to the abrupt climate change in the last one hundred years. The detection and analysis results show that the GICδ18O possesses a quasi-period of 7-9 years, and the abrupt climate changes in the early 1960s and from the end of the 1970s to the beginning of the 1980s arise from the joint effect of the two hierarchies, in other words, the two interdecadal abrupt changes of climate in the last one hundred years are global. The interannual variations of ENSO and sunspots are the important triggering signals for the abrupt climate changes in the last one hundred years. At last, the method of information entropy is employed to estimate the contributions of ENSO signals and sunspots activities to the abrupt climate changes, and it is found that the contribution of the interannual variation of ENSO signals is larger.Based on the above work, the third parts of this paper is that: First, we use the traditional detected abrupt events methods to analysis the series of the Guliya ice core (glacial accumulation) and get the relation and the difference between the detected methods. Second, do research of the relation and the difference between the years of the extreme events and the abrupt years. The results show that (1) as to the abrupt detected, the abrupt years detected will drift with the difference length of the sequence. The length of the sequence is longer and the numbers of the abrupt years will become less, other will become more. But in both cases, the obviously of the abrupt years all been reflected. (2) The abrupt years detected by the running t-test and the Yamamoto are the same. This possibly because although the two methods of the criteria are different, the inspection of the points are the same, i.e. they all differentiate the abrupt by reviewing whether the two series average means are notable. The abrupt years detect by the Crammer’s method is the only parts of by the Yamamoto method, and the Mann-Kendall method cans only detect the beginning of the abrupt years. (3) Used the same detected abrupt method to analyze the original series of the Guliya ice core and its irregular mean, the result been got are nearly the same. The possible reason is that the trend can be ignored by used the abrupt method. (4)The extreme events possibly happened between the two abrupt events.更多还原