Under the background of global warming, recently meteorological disasters and their evocable socio-economic losses have risen, since the extreme weather and climate events occurred frequently. For disaster prevention and reduction, it has guiding significance to deeply understand the characteristics of extreme air temperature and precipitation events which are closely related to the social life. This paper calculated the probability density values of extreme daily air temperature and precipitation through their percentile thresholds, and, based on the probability distribution of extreme values, the hierarchical clustering method was used to incorporate stations. Then the number of regions for the extreme meteorological elements in China was determined according to the stability of variation, and the serial numbers for the various regions were determined by their most probable daily temperature means and precipitation means. The temporal variation and characteristics of the probability distribution of the representative stations that were chosen according to their maximum similarity coefficient to the stations in the same region, were analysed. The analyses show that the spatial distributions of extreme minimum air temperature (except for summer), extreme maximum air temperature in winter, as well as daily precipitation in winter and spring are similar to the distributions of their annual climatic average values, which shows fair relationship with latitude and altitude, that is, latitude and altitude have the greatest influence on the spatial distribution pattern of extrema in spring and winter, respectively, whereas both have great influence in other two seasons; and the spatial difference of the extreme high/low values is mostly determined by altitude/latitude. It is also shown that extreme air temperature events more likely occur in spring and autumn; serious extreme high- and low-temperature events more likely occur in spring and winter over the northern part of China, and serious extreme high-temperature and precipitation events more likely occur in summer over the southeast of China. Compared between the extreme values of the two different climate states, in recent 30 years daily temperature has obviously risen in winter. And the first and second greatest increase in daily precipitation occurs in autumn and summer. The increase in the extremeness and probabilities of low-temperature events in four seasons (especially in winter) are obvious. And the probabilities of extreme maximum temperature in autumn and winter and extreme precipitation in winter are all increased. The increase in the extremeness of extreme precipitation events in four seasons occurs as well. At the end of this paper, some suggestions of disaster prevention and reduction for the regions where serious extreme temperature and precipitation events more likely occurred were provided.%在全球增暖背景下,当前极端天气气候事件频发,由此引发的气象灾害及其所带来的社会经济损失日益增加.深入了解与社会生活密切相关的气温和降水极值的特征,对开展防灾减灾工作有指导意义.文中用百分位阈值求算变量极值的概率密度值,从极值概率角度用系统聚类分析法合并站点,根据方差稳定性特征和变量最概然值的均值确定划分区域个数和对区域编号.从各区挑选与同区站点相似系数最大的代表站,分析其极值概率分布时空变化特征,以此研究中国极端气温和降水的空间分布随时间变化的情况.分析可知气温极小值在除夏季外的其他三季,极大值在冬季,日总降水量在冬、春季,其极值空间分布特征与它们气候年均值的分布相似;极值分区形式与地形和纬度的相关性较大:春、冬季的纬度、海拔高度分别对分区影响较大;其他两季两者都有较大影响;极大、极小值分区分别受海拔高度、纬度影响较大.由气温、降水的概率密度曲线、方差及偏度系数可知春、秋季易发生极端高、低温事件;中国北方地区春、冬季分别易发生极端高、低温事件,夏季东南部地区易发生较严重的极端高温和降水事件.对比1960-1990年和1978-2008年这两个气候态下的极值特征,发现近30年冬季增温最明显;秋季和夏季降水明显增加.全年极端低温的极端性和发生概率都有所增加,尤其是冬季;秋、冬季的极端高温有所增加.极端降水的极端值和发生概率分别在四季和冬季有所增加.最后在易频发严重极端高、低温及降水事件的季节和地区给出了防灾减灾建议.
展开▼
