【作者】 张世轩；

【导师】 封国林；

【作者基本信息】 兰州大学 ， 气象学， 2013， 硕士

【摘要】 自然天气季节的划分是天气气候学研究的主要内容之一,对农业生产和中长期天气和气候预测具有十分重要的意义。当前气候背景下,全球增暖使得气候系统的一些固有属性发生了变化,由此也将对季节产生影响。为了研究季节对全球增暖的响应,迫切需要重新审视原有的季节划分指标,提出适用于当前气候背景的客观和定量化方法,以此揭示当前季节变化所呈现出的规律和特征,为现有的农业生产以及天气、气候预测提供指导。本文基于这一问题,首先从多气象要素的角度描述季节变化,结合相似理论和突变检测理论,研发了适用于季节识别的新方法。通过与传统划分方法进行对比,发现利用多要素比单一的温度要素能更全面贴切地描述气候状态的变化,由此得到的划分结果更贴近大气环流形势和大气活动中心等的季节变化；而从突变角度考虑季节划分,更能反映出季节转换时大气系统的调整和气候状态的转折突变,更好地反映了大气系统在季节内保持相对平稳,在季节间发生较大的调整的事实,而且所划分结果能够突显出局地气候系统及地形等差异对季节变化的影响。其次,利用中国1961-2008年752个站点逐日温度、气压、相对湿度和降水量资料,采用非线性相似检测方法对中国四季进行了划分,研究了近50年全球显著增暖情形下中国四季的时空演变特征。结果表明：(1)气候增暖背景下中国大部分地区冬季持续时间明显缩短,缩短幅度达到10天以上,尤其在20世纪80年代中后期发生转折后缩短趋势突然增强,平均0.29天/年；夏季持续时间自1961年以来增加2.8天,秋季持续时间自1961年以来增加4.7天,且它们在20世纪80年代中后期也各发生一次转折变化,春季持续时间总体48年增加2.6天变化相对较小,但在20世纪90年代后期发生转折变化后呈现出明显增加趋势；(2)与持续天数的变化相对应,中国春季和夏季的起始时间呈现出提前的趋势,而秋季和冬季则呈现出推后的趋势,尤其是气候显著增暖的20世纪80年代中后期以后变化更为明显；(3)不同季节的变化存在一定的差异,中国大部分地区冬季持续天数和起始时间变化最为明显,春、夏、秋三季相对较弱；同一季节不同区域响应也不同,就全国而言北方比南方响应明显,而以黑龙江为代表的东北地区、以新疆为代表的西北地区以及以海南为代表华南等地变化最为明显。最后,利用1961-2010年NCEP/NCAR提供的逐日平均再分析资料,对中国东部地区前冬季节来临时间进行划分,并探讨中国东部地区平均前冬季节来临时间偏早/偏晚与中国东部夏季降水之间的关系,结果发现：(1)中国东部前冬来临时间与中国东部夏季降水存在两个正相关带和两个负相关带,且与其夏季雨带的位置大致吻合。正相关带分别位于华南、西南以及黄河以北的39°N～42N一带、东北中南部等地区,负相关带分别位于黄淮、江淮以及长江中下游一带和42°N以北的大部分地区。(2)统计特征表明：中国东部地区前冬季节来临时间偏早的年份,Ⅱ类雨型发生频率大于其气候发生频率,Ⅰ类和Ⅲ类雨型发生频率则比气候频率偏小；前冬季节来临偏晚的年份,Ⅰ类雨型发生的频率偏大,而Ⅱ类和Ⅲ类雨型发生频率偏小,典型年份的合成分析结果与雨带类型的统计结果相一致。反之,从雨型与前冬来临时间的对应关系统计分析发现,Ⅰ类雨型年份,中国东部地区前冬季节来临时间显著偏晚年份占优,而Ⅱ类雨型年份,前冬季节来临显著偏早年份占优,Ⅲ类雨型年份对应关系不显著。总体而言,季节来临早晚与雨带分布表现出一定的对应关。(3)前冬来临早晚对应了不同的前期和同期海温、水汽和环流型,受海—气相互作用的影响,造成降水空间分布的不同。更多还原

【Abstract】 Season division, which has significant sense to agricultural production and the medium-and long-term weather and climate prediction, is one of the most important issues in the study of synoptic climatology. Under the background of global warming, the raising of air temperature has changed the inherent properties of climate system, which will subsequently affect the natural seasons. Thus, it is necessary to investigate the variations of four seasons under global wanning in order to offer the guidance to agricultural production as well as weather and climate prediction. Base on the above understandings, this paper performs three works. First, we propose two new methods, which named similarity method and non-linear similarity method, to distinguish the four seasons in the current climate. Compared to the existing methods, these two methods use multi-factors to divide season, which makes them more objective, and the division results of the seasons are close to the changes of actual atmospheric circulations and atmospheric activity centers. Moreover, the division results of the new methods could also reveal the influences of the differences of local climate systems and topography on the seasons. Second, based on the752stations’daily observational data of temperature, pressure, relative humidity and precipitation in China from1961to2008, and the non-linear similarity method, we further study the spatial-temporal characteristics of four seasons in China under global warming. It is found that the length of summer has extended and the length of winter has shortened in the majority of the country. In particular, the length of summer, autumn, spring has increased2.8,4.7,2.6days respectively since the1960s. However, the length of winter has shortened by5.6days since the abrupt climate change in mid-1980s. These changes in the lengths of seasons are linked to the changes in the starting dates of seasons. Averaged across the whole country, the starting date of autumn and winter has been delayed while spring and summer in China has started earlier; this phenomenon is especially obvious since1980s. On the other hand, the changes exhibit apparent seasonal and regional differences under the background of global warming. The most significant changes occurred in winter, with summer, autumn and spring to be less. They are more significant in the north than the south where the trend of local changes in Heilongjiang, Xinjiang provinces and south China are the most significant. Third, the relationship between the onset date of preceding winter and the following summer precipitation in eastern China is also investigated by using NCEP/NCAR reanalysis datasets. Results indicate that there are two positive correlation bands and two negative correlation bands. The four bands are corresponding with the main rainbelts in eastern China in summer. The positive ones are located in South China, Southwest China, the region north of the Yellow River with latitudes between39°N and42°N, and central and southern Northeast China. The negative ones are in Huanghuai, Jianghuai, the middle and lower reaches of the Yangtze River, and most regions north of latitude42°N. The statistical results show that the frequency of rainfall pattern Ⅱ is larger than that of the climate state, those for both rainfall patterns I and III are less than that of the climate state when earlier preceding winters come. While for later winters, the pattern I is larger than that of the climate state, those for both rainfall patterns Ⅱ and Ⅲ are much smaller than that of the climate state. The rainbelt outcomes from the statistics and composition analysis in the typical years are consistent with each other. The statistical results show that the later preceding winters are prime when it is pattern I, earlier winters for pattern Ⅱ, and there is not significant relationship between the winter starting dates and pattern Ⅲ. In a word, the patterns of precipitation are closely related to the starting dates of the seasons. The different starting dates of winters are corresponding to different air-sea interactions which result in different patterns of precipitation.更多还原