海洋环流与波动重点实验室知识库

本文利用 1850~2005 年英国气象局 Hadley 中心的海表面温度数据集 HadISST 资料、美国国家大气海洋局 NOAA 的大气资料以及 15 个 CMIP5 模 式模拟输出数据，使用奇异向量值分解、合成分析与回归分析等方法研究了 南太平洋副热带偶极子(South Pacific Subtropical Dipole，SPSD)在历史观测 中和模式模拟中的时空分布特征并进行了对比分析。后面进一步讨论分析了 SPSD 与赤道厄尔尼诺(El Niño and South Oscillation , ENSO)之间可能的联 系。主要结论如下: 
(1)观测中对南太平洋区域扣除 ENSO 后进行 SVD 分解得出的第一主 要模态即为 SPSD。在海表面温度(SST)上表现为东北区域与西南区域反位相 的海表面温度分布，在海表面气压上表现为一个范围较大的负气压中心。 SPSD 发展周期为南半球春季(9~11 月)开始发展，夏季(12~2 月)成 熟，秋季(3~5 月)开始衰退。 
(2)将 15 个 CMIP5 模式模拟数据进行同样处理，得出 SPSD 的海表面 气压与海表面温度模态。结果显示，结果表明:其中 10 种模式可以模拟出完 整的 SPSD 生成发展过程，且 SPSD 的主要区域与观测较为接近，但其余 5 种模式在模拟强度、位置与观测有较大出入。 
(3)所有 CMIP5 模式在模拟 SPSD 生成阶段时比观测􏴞前一个月出现 偶极模态,1/3 的模式海表面温度偶极异常可以追溯到 6 个月之前;潜热通量与 海表面温度的时空分布显示, 潜热通量是影响偶极模态生成发展的主要因 素。模态的变化主要受大气环流的调制,在模态发展最强时部分模式的正极上 方有正潜热通量异常,即海洋向大气传递热量。分析显示模式模拟海气耦合过 程中的 SST 模拟强度较观测偏强, 气压方面与观测较为接近。 
 

This paper uses the HadISST dataset of the Met Office Hadley Center from 1850 to 2005, the atmospheric dataset of NOAA(National Oceanic and Atmospheric Administration), and the model output data of 15 CMIP5(Coupled Model Intercomparison Project Phase 3) models, with SVD(singular vector decomposition), synthetic analysis and regression analysis. The temporal and spatial characteristics of the SPSD(South Pacific Subtropical Dipole) in historical observations and model simulations were studied and compared. Then it discusses the possible links between SPSD and ENSO(El Niño and South Oscillation). The main conclusions are as follows: 
(1) The spatial distribution of the first two main modes of the equatorial ENSO and the interpretation of the variance ratio are obtained by the EOF(Empirical Orthogonal Function) method, and then returned to the South Pacific region for deduction. The first major mode resulting from SVD decomposition of the region is SPSD. SST is reflected in the northeast region of the southwest region of the anti- phase of sea surface temperatures, showing a wider range of negative pressure centers on the sea surface pressure. The corresponding time development cycle starts in the southern hemisphere spring (September to November), matures in summer (December to February), and declines in autumn (March to May). 
(2) The analog pattern data CMIP5 15 performs the same processing, successfully simulated SPSD PSL(Pressure of Sea Level) and SST mode. The results show that ten models can simulate the complete SPSD generation process, and the main areas of SPSD are close to the observations, but the other five models have large differences in intensity, position and observation. 
(3) All models exhibit dipole mode one month before compare to the observation of the SPSD generation phase, and 1/3 models SST dipole anomaly can be traced back to six months; The spatial and temporal distribution of latent heat flux and SST show that latent heat flux is the main factor affecting the development of dipole mode. SPSD’s change is mainly modulated by the atmospheric circulation. When SPSD develops strongest, there is a positive submerged latent heat flux abnormality above the positive pole of some modes, that is, the ocean transfers heat to the atmosphere. The analysis shows that the model’s SST simulation intensity during simulating air-sea coupling process is stronger than the observation, and the PSL is close to the observation.