海洋环流与波动重点实验室知识库

太平洋赤道潜流（Equatorial Undercurrent，EUC）是位于赤道次表层的一支强劲的东向流，对赤道太平洋纬向的质量和热量输送以及盐交换产生直接影响；并且与海气相互作用信号厄尔尼诺-南方涛动之间联系密切，对我国的天气和气候变化乃至全球海气相互作用过程均具有重要的影响。过去研究认为EUC的维持和变异由纬向压强梯度力所主导，也与Rossby波相关的大尺度环流异常相关，为从赤道波动的角度来研究EUC的变异奠定了基础。然而，影响EUC变异的风应力结构、位置和时期及其潜在的物理机制并不清晰。本文基于伴随敏感性方法研究了在不同的超前月份下能够使得EUC发生最大变异的风应力扰动结构，以及具有这种敏感性结构的风扰动驱动下EUC变异的物理机制。本文的主要研究内容与结果如下：

（1）利用德国海洋环流和气候估算系统（German contribution to the Estimation of the Circulation and Climate of the Ocean，GECCO）下的海洋环流模式对赤道流系进行了数值模拟。结果表明：GECCO系统能够准确刻画出EUC气候态基本结构特征和季节变异特征，与海洋再分析数据和现场观测资料保持一致，为伴随敏感性的计算提供了可接受的海洋背景态；此外，数值实验表明GECCO系统中EUC流量与风应力扰动之间具有较强的线性关系，为伴随敏感性的准确计算提供了基础。

（2）利用GECCO系统下的伴随模式得到了气候态强迫下7月份165°E断面的EUC流量对纬向风应力和经向风应力在超前11个月内的伴随敏感性结构。结果表明：敏感性结构主要分布于热带太平洋（15°S至15°N之间），呈现出关于赤道高度对称的“漏斗”状结构，且对纬向风应力的敏感性更强；不同月份的敏感性分布类似于Rossby波信号的传播过程，随着时间的发展向西移动；最大的敏感性强度发生在超前1–3个月内。

（3）在不同超前月份进行具有敏感性结构的风速扰动实验。结果表明：敏感风可以使EUC在目标月份显著增强，其中，超前4到11个月的非局地风扰动会导致海盆尺度的EUC增强，超前小于4个月的局地风扰动仅能够导致目标断面附近的EUC增强；通过分析海表面高度异常的发展过程，发现敏感风扰动所激发的赤道Kelvin波、赤道Rossby波及其反射波相互叠加是导致EUC变异的主要原因。此外，以超前6个月的具有敏感性结构的敏感风扰动为例，验证了其与实际风异常具有部分的一致性，可以用来解释EUC的季节变异和年际变异，验证了具有伴随敏感性结构的风扰动的合理性。

（4）不同断面EUC的伴随敏感性结果及其导致EUC变异的机制。通过设定目标断面分别为EUC的其他特定断面和连续经度断面，发现对应的敏感风的结构具有一定的差异性，而敏感风导致EUC变异的动力机制仍为赤道波动；对比高分辨率TOOSSE系统下的伴随敏感风结果，也得到类似结果，验证了利用GECCO研究敏感风影响EUC变异的适用性。

综上所述，本文基于伴随敏感性方法得到了在超前不同月份中最能够有效影响EUC变异的风应力扰动结构，并探讨了风驱动EUC变异的赤道波动机制，进一步补充了传统的纬向压强梯度力机制。同时，研究发现敏感风扰动与实际风异常具有一定的相似性，为探寻影响EUC变异的实际风异常提供了一个标尺；并且基于伴随敏感性的超前特性，为预测EUC的变异提供了重要理论支撑。

The Pacific Equatorial Undercurrent (EUC) is a strong subsurface eastward current, which has a direct impact on mass and heat transport and salt exchange in the equatorial Pacific Ocean. EUC also exerts a closed relationship with the air-sea interaction signal of El Niño and Southern Oscillation, which has an important impact on China's weather and climate change even with the global air-sea interaction process. It is generally believed that the maintenance and variation of EUC are dominated by zonal pressure gradient force in the previous study. Meanwhile, a large-scale circulation associated with a Rossby wave can also lead to EUC variation, which lays the foundation for studying the variation of EUC from the perspective of equatorial waves. However, it is still unclear about the structure, location, and period of influential winds that affect EUC variation, as well as the underlying mechanism of wind-driven EUC variation. Based on the adjoint sensitivity method, this paper obtains the structure of the wind that can make the EUC change most efficiently at different leading months, and the underlying mechanism. This study has drawn the following main results and conclusions:

(1) Numerical simulation of the equatorial current system is carried out using the ocean circulation model under the German contribution to the Estimation of the Circulation and Climate of the Ocean (GECCO) framework. The results show that the GECCO system can grasp the basic structure and seasonal variation of EUC in the climatology state compared with the oceanic reanalysis and in-suit observations, which provides an acceptable ocean background state for the calculation of adjoint sensitivity; in addition, numerical experiments show that there is a strong linear relationship between EUC transport and wind stress perturbation in the GECCO system, which provides a basis for the computation of adjoint sensitivity.

(2) Using the adjoint model under the GECCO framework, the structures of adjoint sensitivity of the EUC transport at the 165°E section in July of the second year to the zonal and meridional winds forcing in the previous 11 months are obtained. The results show that the adjoint sensitivity is mainly distributed in the tropical Pacific Ocean (between 15°S and 15°N), exhibiting a “funnel-like” structure which is highly symmetrical about the equator, and the EUC transport is more sensitive to zonal wind stress than meridional wind stress; the evolutions of the adjoint sensitivity are similar to that of the Rossby wave signal, which moves westward over the developed time periods. The strongest adjoint sensitivity occurs approximately 1–3 months ahead of the target month.

(3) Wind velocity perturbation experiments with sensitivity-like structures are conducted in different leading times. The results show that sensitive winds can indeed make the EUC enhance obviously in the target month, and the non-local wind perturbations, which occur 4 to 11 months prior of the target month, can indeed lead to the enhancement of EUC at the basin scale; in contrast, local wind perturbations (with the lead time of no earlier than 4 months) can only lead to the enhancement of EUC near the target section. Accordingly, by analyzing the development of sea surface height anomalies, it is found that the equatorial Kelvin waves, the off-equatorial Rossby waves, and the reflected waves superimpose onto each other, excited by sensitive wind perturbations, can cause the variation of EUC eventually. Meanwhile, taking the sensitive wind perturbation in the month –6 as an example, it is verified that it is partially consistent with the real wind anomaly, which can be used to explain the seasonal and interannual variability of EUC, verifying the rationality of adjoint sensitivity-like wind.

 (4) The adjoint sensitivity results of the EUC at various sections. By setting the target section as other specific sections and continuous longitude sections of the EUC respectively, it is found that there are some differences between the structures of the corresponding sensitive wind, however, the dynamic processes of wind-driven EUC variations can also be explained by equatorial waves. In addition, it is compared the structure of adjoint sensitivity of EUC transport to zonal wind stress calculated by GECCO and TOOSSE with high resolution, which also indicates the reliability of the conclusions obtained in this study.

In summary, based on the adjoint sensitivity method, the present work obtains the spatial patterns, and locations of critical wind perturbations within different leading-months that most effectively affects EUC variation in different leading months, and provide the equatorial waves mechanism of wind-driven EUC variation, which can be regarded as complementary to the canonical zonal pressure gradient force mechanism for basin-scale EUC variation. Meanwhile, it is found that the distribution of real wind anomaly is partly similar to the sensitivity-like wind, providing a benchmark for searching for the potentially influential winds that may be responsible for the EUC variations and ENSO prediction.

第1章　绪论... 1

1.1　研究背景与研究意义... 1

1.2　研究现状与研究进展... 7

1.2.1　EUC的生成与维持机制... 7

1.2.2　EUC的变异机制——纬向压强驱动过程... 9

1.2.3　EUC的变异机制——环流驱动过程... 10

1.3　研究内容和章节安排    11