【作者】 王宏娜；

【导师】 何宜军； 陈锦年；

【作者基本信息】 中国科学院研究生院（海洋研究所） ， 物理海洋学， 2009， 博士

【摘要】 随着全球变暖和各种极端天气事件发生频率的增加,气候变化已成为全世界关注的焦点问题,ENSO事件是年际尺度上气候变化的最强信号,它的影响波及到世界各地,所以,对ENSO事件进行深入研究,弄清其发生机制,具有重要的科学意义。虽然,目前对ENSO事件的研究已经取得了巨大成果,但仍不能对其发生发展的全过程进行准确的预测,因此,仍需要对其进行深入研究,本文正是从热带太平洋次表层温度场和流场的变化着手,对ENSO循环过程中的相关机理进行研究。通过对TOGA/TAO、SODA和NCEP等海洋大气实测和再分析资料的分析,研究了热带太平洋次表层海温和流场的变化特征,分析了它们的变化机理,进而深入探讨了它们与ENSO事件的相互关系,并利用全球海洋数值模式的敏感性试验,探讨了大气风场的变化对海洋次表层要素的影响,主要得到以下结论:1.上世纪70年代末的气候突变之后,用28℃定义西太平洋暖池(WPWP),已不能合理的描述WPWP的基本特征。我们通过对比分析提出,用28.5℃来定义WPWP更合理,这一定义即可以充分反映WPWP突变前的特征,又能够合理的反映WPWP突变后的特征;对新定义的WPWP区域不同深度的海温距平的分析表明,次表层(148m)海温距平的变化趋势和变化幅度与表层和深层的变化差异较大,次表层海温的变化幅度最大并且年代际变化趋势与上下层正好相反;进一步的研究表明,WPWP次表层海温的年代际变化与PDO的变化有一定关系。2.利用热带太平洋次表层海温的变化特征,定义了表征ENSO事件的新指数——赤道太平洋温跃层海温振荡指数(EPOI),与其它ENSO指数相比,EPOI将东、西太平洋次表层海温的变化信息都包括在内,能够较全面的反映出ENSO事件的变化特征,特别是EPOI可以较好的反映出ENSO循环的年代际变化特征。另外,EPOI的变化比ONI的变化超前2个月,更有利于对ENSO事件的提前预报。由于EPOI主要反映海洋次表层的变化特征,因此能够更好的满足对ENSO机理研究的需要。3.赤道潜流距平场的EOF分析表明,其前两个模态的方差贡献较大,第一模态方差贡献为30.75%,主要反映东太平洋赤道潜流的变化特征;第二模态方差贡献为16.18%,主要反映中太平洋赤道潜流的变化特征。赤道潜流前两个模态与ENSO指数的变化有很好的相关关系,其中东太平洋潜流在滞后ENSO指数1个月时,二者达到最大负相关(r=-0.74)。中太平洋赤道潜流的变化对“东部型”和“中部型”El Ni?o事件的形成有一定影响。“东部型”El Ni?o事件发生前,中太平洋赤道潜流异常增强,次表层异常海温信号随着潜流中心迅速向东移动到达东太平洋,使得“东部型”El Ni?o事件爆发;而“中部型”El Ni?o事件发生前,中太平洋赤道潜流则异常减弱,西太平洋异常海温信号不能迅速向东传播,而是在中西太平洋堆积并向上扩展,使得异常海温首先在中太平洋出现,“中部型”El Ni?o事件爆发。4. ENSO循环过程中,异常冷(暖)信号之所以在8oN-10oN附近向西传播的原因较多,其中温跃层深度在8oN-10oN的特殊分布特征对其有一定贡献,具体表现为在北半球8oN-10oN正好是温跃层深度较浅的区域,该区域的温跃层相当于从东到西的一个海下“山脊”,使得来自南北两侧的异常信号都很难穿过这一区域,而只能沿该纬度向西传播;而南半球的温跃层对来自赤道地区的异常信号没有阻挡作用,使其可以直接传播到高纬度地区。ENSO信号的强度在传播过程中发生了明显的变化,主要是ENSO事件爆发后,4-5年的周期信号并没有传到东太平洋10oN附近,在从东到西的传播过程中,4-5年的周期信号有所增加,但增加的幅度较小。在西太平洋有来自南、北半球中高纬度异常信号的补充,从而使得ENSO循环得以维持。5.数值模拟表明,不同区域风应力的变化对海洋的影响各不相同。赤道地区风应力对海洋的影响主要通过纬向分量的变化来产生作用,它的变化主要对赤道次表层海温的变化产生影响,并且东西太平洋呈现反位相变化趋势;而对流场的影响则是上下层反位相变化。北太平洋副热带地区风应力的变化对海洋的影响与赤道有明显不同,对温度场的影响表现为东西太平洋次表层海温的变化一致,而对流速的影响则是东西太平洋反位相变化。更多还原

【Abstract】 Due to the global warming and the increase of extreme weather phenomena, the climate variability has already become a central issue which was attended by the people all over the world. ENSO cycle was the strongest signal of climate variability in the interannual time scale and it could affect the weather around the world. Therefore, the study about ENSO cycle has important significance. Although the ENSO event’s research had already made great progress, the process of ENSO event still couldn’t be accurately forecasted. Therefore, our study was to analyses the variability of subsurface temperature and current in tropical Pacific Ocean and then deep research the mechanism of ENSO cycle.Based on the TOGA/TAO, SODA and NCEP datasets, the variability of ocean temperature and current was studied on the subsurface in tropical Pacific Ocean. The relationship between these factors and ENSO events was researched. Using the ocean model, the wind stress impacting on ocean subsurface factors was analyzed. The mainly results are as following:1. After the climate shift which occurred at the end of 1970s, the definition of western Pacific warm pool (WPWP) by SST≥28℃couldn’t describe its basic character exactly. According to the analyses, we propounded that using the SST≥28.5℃defined the WPWP. The new definition could well reflect the character of WPWP not only before the climate shift but after that. The study about temperature in different depth of WPWP indicated that the anomalous temperature variability on subsurface (148m) was different from that on its upper and under layer. The ocean temperature on subsurface had the largest variation and the reverse trend of interdecadal variability to upper and under layers temperature. The interdecadal variability of subsurface ocean temperature in WPWP had relationship to the variability of PDO.2. An index to the equatorial Pacific sea temperature oscillation (EPOI) was defined to describe the ENSO phenomenon according to the ocean temperature on thermocline surface in Pacific Ocean. Comparing with the older ENSO indexes, the EPOI contained the information of ocean temperature on thermocline in eastern and western Pacific Ocean. Therefore, it could reflect the character of ENSO events in full especially in the interdecadal variability. Furthermore, the EPOI varied earlier than the ONI about two months. We could ahead forecast the ENSO event occurring using the EPOI. The EPOI could satisfy the research on ENSO cycle.3. We analyzed the anomalous equatorial undercurrent in Pacific Ocean by the method of Empirical orthogonal function (EOF). The results indicated that the previous two modes of equatorial undercurrent had the bigger variance contribution. The first mode’s variance contribution was 30.75% and it mainly reflected the variability of eastern Pacific equatorial undercurrent. The second mode’s variance contribution was 16.18% and it mainly reflected the variability of central Pacific equatorial undercurrent. These two main modes have obvious correlations with ENSO events. There had the biggest negative correlation between the first mode of undercurrent and ENSO when the first mode of undercurrent lagged the ENSO (r=-0.74). This indicated that eastern Pacific equatorial undercurrent weakened (strengthened) after the onset of warm (cold) ENSO events. The variation of central Pacific equatorial undercurrent had influence on the different pattern El Ni?o events. Before the onset of eastern pattern El Ni?o events, the central Pacific equatorial undercurrent singularly increased and then the signal of anomalous temperature quickly propagated to the eastern Pacific. Therefore the El Ni?o event was eastern pattern. Before the onset of central pattern El Ni?o events, the central Pacific equatorial undercurrent singularly decreased and then the signal of anomalous temperature accumulated on the central Pacific. Therefore the El Ni?o event was central pattern.4. In ENSO cycle, the depth of thermocline in 8oN-10oN Pacific was one of reasons that anomalous temperature signal transport to western Pacific along 8oN-10oN. Because the thermocline was very shallow in this area, the thermocline was like a hill across the eastern and western Pacific under the sea surface. Then the anomalous signal couldn’t across this area from south (or north) to north (or south), so the signal in eastern Pacific could only transport to western Pacific along this area. On the contrary, the thermocline in the southern hemisphere couldn’t block the signal’s transportation from equator to the high latitude. There had been a marked change of signal intensity in the ENSO cycle. After the onset of ENSO events, the signal of 4-5 years hadn’t reached 10oN. When the ENSO signal transported from eastern to western Pacific along 8oN-10oN, the signal intensity of 4-5 years increased. But the increased signal too weak to maintain the ENSO cycle. The ENSO cycle could maintained because there had other signal supplement from the middle and high latitude. 5. Modeling test indicated that the variability of wind stress in different area had different effect on the ocean factors. The wind stress in equatorial Pacific affected the ocean mainly through its zonal component. Its effect on the variation of subsurface ocean temperature was different in the eastern Pacific and western Pacific. But for the currents the difference was between the surface and subsurface. The variability of wind stress affecting the ocean in the northern hemisphere subtropical area was obviously different from in the equatorial area. The effect of subtropical wind stress on the subsurface ocean temperature was identical from eastern to western Pacific. But for the currents there had the contrary effects between eastern and western Pacific.更多还原