【作者】 杨萍；

【导师】 丑纪范； 封国林；

【作者基本信息】 兰州大学 ， 气象学， 2009， 博士

【摘要】 本文基于对极端气候事件研究的全面回顾,指出极端气候事件及其衍生灾害对社会和经济的影响力和破坏力越来越严重,因此关于极端气候事件的评估和规律分析是非常重要的研究领域。从极端气候事件的影响看,群发性极端气候事件的影响力更大,破坏力更强。极端气候事件的群发性有很多理解,本文主要从区域丛集性角度理解单一极端气候事件的群发性特征,将前人对于极端气候事件各气象要素的统计转化成衡量站点权重的标准,得到了强权重的站点分布具有空间群发性特征的结论,紧紧围绕“群发性”这一主题,沿着“如何提取群发性区域”这一研究主线,以“k阶最近邻距离提取算法”为技术路线,完成了提取极端气候事件群发性区域的研究目标。以该研究结果为基础,从年际、年代际和区域性三个角度具体分析了群发性极端气候的年际、年代际和区域性的特点,得到了一些有意义的结果,从而为极端气候事件的评估提供了一条新的思路和方法。主要结果和结论如下:(1)从数学物理的角度对本文的技术路线“k阶最近邻距离提取算法”进行了深入的理论研究,发现该算法的错误率受到数据点总数、距离阶数以及疏密差异比三个要素的影响,通过采用控制变量的方法,发现数据点总数影响相对最小,由于算法错误率高时会带来算法不适用的问题,因此我们对区域内随机分布的数据点进行了适用性研究,该结论是后文算法在极端气候事件中有效性检验的基础。此外,我们发现数据点的提取结果不受分布形态的影响,同时,引入权重的概念,用数据点重复出现的次数来体现数据点的权重,多次数值实验证明,通过增加大权重区域的分布密度,可以有效地将大权重区域内的点提取出来,这样的改进使该算法能够应用到极端气候事件群发性的研究中,算法的理论研究增强了算法应用的可靠性。(2)以算法理论研究的适用性研究为基础,同时结合数据点权重的思想,分析了本文的技术路线“k阶最近邻距离提取算法”在群发性极端气候事件中应用的可能性,定义了12种极端温度事件和6种极端降水事件,以群发性极端气候事件的基本运用流程为指导,以多年疏密差异比的平均值和有效率为检验指标,对各种极端气候事件进行有效性检验,结果发现,该技术路线适用于群发性极端气候事件的研究。(3)以极端温度事件、大暴雨事件、强降水事件为研究对象,以去除新疆、西藏和内蒙古三省区的中国东亚季风区域为研究范围,基于逐年极端气候事件的群发站点结果,进行群发性极端气候事件的年际研究。在年际变化中,用群发站点数体现极端气候事件的年际群发程度,发现夏季高温事件、异常偏高事件、冬季低温事件、异常偏低事件其转折期各不相同,基本介于70年代中期至80年代中期,大暴雨事件10年代际的周期性较明显,强降水事件以年际振荡为主。群发性极端气候事件的年际变化较为复杂,不同的极端气候事件没有较为统一的年际变化规律。同时对东部四个气候区域(东北、华北、华中、华南)各个区域的群发站点数进行统计,并将区域极端气候事件群发站点数年际变化曲线与相应的全国年际变化曲线进行相关性分析,发现东北地区夏季高温事件的群发性与全国的相关性较差。(4)基于技术路线所得到的各年提取结果,进行了年代际群发性特征的研究。定义了年代际群发性指数,定义了一级群发和二级群发两个年代际群发性指数的高值区作为本论文的重点研究对象,分析了极端温度和极端降水事件的年代际群发性指数高值区的年代际特征,发现群发性极端气候事件的年代际变化与前人所研究的气候变化年代际背景较为一致。例如,极端降水事件的年代际群发区域与多雨带的年代际变化区域有非常好的吻合,这说明,在年代际尺度上,多雨区域很可能是由于群发性极端降水事件占据主要贡献造成的。(5)对所研究的中国区域进行了分区讨论,从气候影响和经济发展水平两个角度进行分区,分为气候区域(东北、华北、华中、华南、西北、西南)和经济区域(京津唐、长三角、珠三角)两类,结合站点的年代际群发性指数的结果,求出上述几个区域的年代际群发性指数的平均值(区域内所有站点年代际指数的总和/区域站点总数),重点研究经济区域与所属气候区域的区域差值的变化规律。研究结果发现,人为因素的影响在京津唐区域的冬季极端温度事件中有所体现,但整体来看,三大经济发展区域与所属的气候区域之间的区域差值没有明显的规律可寻,可以认为,对于群发性极端气候事件而言,人类活动的影响是非常微小的。(6)通过大量的文献查阅和对历史上极端气候事件的概况研究,对于90年代以后的逐年极端气候事件的群发站点结果与历史概况进行了对比分析,发现极端气候事件群发性站点提取的结果图和实际发生的区域有非常好的吻合,这再次证明了本论文技术路线的可靠性,同时给极端气候事件的研究提供的一个新的思路和角度。在进行上述对比分析之后,还对群发性极端气候事件与火山喷发、太阳活动、季风指数和ENSO事件四方面的可能影响因子进行了相关讨论,发现火山喷发与夏季极端高温事件群发性的减弱有明显的相关性,太阳黑子数减弱的阶段往往容易发生较强的群发性极端气候事件,夏季风指数与东北极端降水事件的相关性较高,冬季风对华南极端温度事件的群发性有影响,极端气候事件群发性最强的几个年份中,往往都发生了中等以上的ENSO事件,可见,群发性极端气候事件与这几个影响因子有着较好的响应关系。更多还原

【Abstract】 Based on a comprehensive review for extreme climate events,it is pointed out extreme weather and climate events have received increased attention in the last few years,due to the often large loss of human life and exponentially increasing costs associated with them. Therefore,the assessment and prediction of extreme climate events is an important field.Through the influence,group-occurring extreme climate events may create more potential for catastrophic impacts.There are number of ways group-occurring events can be understanded,this paper mostly means to spatial clusted events which belong to same events.We transform the meteorological factors to the weight of stations,finding that high weight stations are centralized instead of separated,which can be called group-occurring events.On the paper,group-occurring is the topic, how to delineate the group-occurring area is the masterstroke and kth-order nearest-distance arithmetic delineating clusterd events is the primary technique.On the basis of sush things,the group-occurring area is delineated,and annual,decadaI,regional analysis have been done.Many exciting results are achieved,which provide a new approach for the assessment of extreme climate events,the major conclusions of the study may be summarized as follow:Firstly,much theoretical research has been done.We introduced the algorithm of cluster extraction which is based on kth order nearest-distance,finding that kth order,the ratio of cluster and noise’s densities and the number of data all have effect on the method.But the effect of the two formers are more important than the last one.This result is the base of the efficience test following.Furthermore,we introduced the concept of weight to extend the range of the simulated data,and gave the data different weights for delineating clusters,so that the range of the algorithm can be extended.Secondly,the possibility of algorithm’s applying on the extreme events is analysed.12 kinds of temperature extremes and 6 kinds of extreme precipitation become the paper’s objects.Based on the average ratio of cluster and noise’s densities and efficient,we test all kinds of extreme climate event,finding that the algorithm is fit for group-occurring extreme climate events,especially for heavy events.Thirdly,temparture extremes defined by 99 percent,heavy rain and strong precipitation are the dominating objects.Chines region wiping off Sinkiang,Tibet and Leymus are the research scope.Based on the annual result of group-occurring stations about extreme climate events,we make attentions to the interannual varity of the number of group-occurring stations which can repress the group-occurring degree.We find that different extreme climate events have different turning point,which between middle of 1970s and midde of 1980s.Heavey rain has a notable decadal cycle,while strong precipitation has a obvious annual surge.It is provided that interannual variety is complex on group-occurring extreme climate events.Forthly,decadal group-occurring index if defined to access interdacadal variability.The result expressed that group-occurring decadal variability is consistent of the existing research.For example,strong precitpitation is unanimous of the existing dacacal variability.Fively,we investigate the annual and decadal climate characteristics of four major regions in eastern China,e.g.northerneast,north,south and central China.We are especially interested in the identity between the annual variation of climatic regions and the country as a whole in our annual cluster anlysis and find that the conformity of high temperatures of summer in northeast China is the worst.In the analysis of decadal characteristics,we draw a comparison between the climatic regions and three major metropolitan area of China(Jing-jin-tang Economic Circle,Yangtze River Delta and Perl River Delta).Results indicate that human-induced influences is somewhat reflected in the clustered extreme events in winter of Jing-jin-tang Economic Circle,however,no significant correlations can be drawn between three metropolitan area and the climatic regions it belongs to, e.g.the influence of anthropic factor on clustered extreme events is comparatively small.The lastly,based on large amount of literature consult and existing research work of extreme climate events,we find the obtained distribution of clustered extreme events is in good accordance with the observational data through making comparative analysis between the annual result of clustered extreme events after 1990 and historical records,thus certify the reliability of technical proposal in this paper and shows a new angle and direction of extreme climate event research. Furthermore,we investigate the correlation between clustered extreme events and its possible factor of influence,e.g.volcanic explosion,solar activity,index of monsoon and ENSO,and find the extreme climate events show a good response to the factors mentioned aboved.Volcanic explosion has relations to extreme high-temperature events.When the number of macula reduces,strong extreme climate events are likely to take place.Summer monsoon has effect to northeast extreme precipitation events while winter monsoon is contact with extreme temperature events on east of south China.ENSO events are possibly to happen in the years that strong extreme climate events take place.更多还原