中文摘要：

本文利用欧洲中期天气预报中心（ECMWF）逐日再分析资料（ERA-40）,以500 hPa位势高度滤波方差为代表,对1957年12月～2001年11月期间44年528个月北太平洋区域（30°N～60°N,120°E～120°W）月平均风暴轴的多中心数目和最强中心位置进行了客观统计,在此基础上,对北太平洋区域进行了分区,通过对比各区域风暴轴的时间演变和结构变化,重点揭示了北太平洋东部地区风暴轴的时空演变特征.主要结论如下：（1）逐月来看,北太平洋风暴轴“多中心”现象普遍存在,概率高达94.7%,最典型的分布呈2～3个中心分布;从季节上来看,春季是“多中心”现象最容易出现的季节,秋季和冬季相对较少,而4个及以上的“多中心”现象则更容易出现在夏季.（2）若把最强中心出现在160°W以东地区的北太平洋风暴轴定义为东部型风暴轴,那么从月份上来看,7月相对最容易出现东部型风暴轴,1月和2月最难;从季节上来看,夏季相对最容易出现东部型风暴轴,冬季最难;总的来看,出现东部型风暴轴的频数大约占总频数的三分之一.（3）从垂直结构上看,在北太平洋160°W以东地区,风暴轴的强度可以最强,但与斜压性密切联系的涡动向极和向上热量通量的最大值却并不是最强.（4）经验正交函数分解（EOF）分析的结果表明,在不同季节、不同区域以及是否单独考虑东部型风暴轴的情况下,风暴轴的变化虽然表现出了一定的差异,但都反映出在北太平洋东部区域风暴轴的变化特征有其独特特点,如在该区域风暴轴的主要变化模态并不一定时时与其他区域的主要变化模态一一对应.北太平洋东部区域风暴轴变化的原因和机制值得进一步深入探讨.

英文摘要：

On the basis of daily reanalysis data from the European Center for Medium-Range Weather Forecasting （ERA-40）, the filtered 500-hPa geopotential height variance between 2.5 and 6 days is first used for a statistical analysis of the number of maxima centers and the position of the strongest center for 528 monthly mean storm tracks over the North Pacific ocean （30°N-60°N, 120°E-120°W） from December 1957 to November 2001. Then, after a comparison with other subregions, the temporal and spatial variation in the eastern part of the North Pacific storm track is revealed. The results are summarized as follows： （1） As viewed from month to month, a polycentric distribution, most typically with two or three centers, is common, with a probability as high as 94.7% over the entire Pacific storm track. This distribution appeared most often in spring and least often in autumn and winter, whereas a polycentric distribution with four or more centers appeared most often in summer. （2） The eastern pattern storm track, which is defined as having the strongest center located east of 160°W, generally with a 33% probability, appeared most often in summer, especially in July, and least often in winter, especially in January and February. （3） In terms of the vertical structure, the track intensity can be strongest over the eastern part of the North Pacific storm track, but not the baroclinic transport of energy poleward and upward by transient eddies. （4） The empirical orthogonal function （EOF） result shows that some differences do exist under different conditions when seasonal, regional, and track type effects are considered, but a few unique features still appear in the variation in the eastern part of the North Pacific storm track, e.g., the major EOF modes are not always in a one-to-one correspondence with their counterparts over the entire North Pacific storm track. These results suggest that the mechanism of the variation in the eastern Pacific storm track is worthy of further investigation.