海洋环流与波动重点实验室知识库

热带太平洋海温是全球气候系统的关键组成要素，其气候态分布和年际变率深刻影响全球气候演变，具有重要的气候意义。以厄尔尼诺和南方涛动（El Niño and Southern Oscillation，ENSO）现象为代表的热带太平洋海洋物理状态（包括海温）的时空分布，对海洋生物地球化学（Biogeochemistry，BGC）过程具有显著影响。例如ENSO相位转换影响着热带东太平洋上层海洋营养盐浓度，从而调制海洋叶绿素（Chlorophyll，CHL）浓度分布。另一方面，CHL浓度影响太阳辐射在上层海洋中的垂向穿透率，从而调制短波辐射吸收量及其直接加热率的垂向分布，造成局地海温垂向分布变化，并通过海洋热动力过程引起海盆尺度海温变化。CHL反馈作用体现了海洋BGC过程通过影响海温分布从而参与全球气候演变的重要意义。

当前气候模式对海洋生物地球化学过程的表征方式差异，是热带太平洋海温模拟偏差的重要原因。例如，观测揭示热带太平洋次表层水体中广泛存在深层叶绿素极大值现象（Deep Chlorophyll Maximum，DCM），即次表层CHL浓度峰值高于海表CHL浓度。而当前模式大多仅考虑海表CHL分布对太阳辐射分布的影响，较少讨论次表层CHL垂向分布的贡献，即DCM效应。所以在短波辐射穿透方案中引入DCM效应，可能造成CHL对海温反馈作用的模拟差异；但目前缺少对DCM效应的定量研究。此外，在热带西太平洋等寡营养盐海域，海洋固氮率（Nitrogen Fixation Rate，NFR）较高，固氮过程为上层海洋生态系统提供30%至50%的氮盐输入，影响初级生产力与CHL分布。所以该区域NFR变化也可能通过调制CHL分布而对海温分布产生间接反馈作用；当前对此问题研究不足。

为深入探讨热带太平洋DCM分布和NFR变化对海温的反馈作用，本文采用大气-海洋物理-BGC混合型耦合模式（Hybrid Coupled Model of Atmosphere, Ocean Physics and ocean Biogeochemistry，HCM-AOPB）和一维的海洋湍流-BGC耦合模式（Generic Ocean Turbulence Model-Tracers of Phytoplankton with Allometric Zooplankton，GOTM-TOPAZ）进行模拟研究。其中，HCM-AOPB能够有效模拟热带太平洋DCM分布，并可显式调节海气耦合强度；GOTM-TOPAZ则可较好模拟海洋固氮过程；两者均采用了基于CHL垂向分布的太阳辐射穿透方案。本文主要讨论了三个科学问题：（1）给定风场强迫下DCM分布对海温的反馈作用；（2）海气自由耦合情况下，DCM反馈作用受海气耦合强度的调制影响；（3）热带西太平洋NFR变化对局地海温垂向分布的影响。主要结论如下：

（1）在给定风场强迫下，利用HCM-AOPB研究了热带太平洋DCM分布对海温气候态和ENSO振幅的影响。在气候态试验组，在给定的气候态风场强迫下，引入DCM分布后，赤道东太平洋海表温度（Sea Surface Temperature，SST）降低约0.2 ℃，次表层海温降幅约0.7 ℃℃；赤道西太平洋SST升高约0.1 ℃℃。DCM引起的直接生物加热效应（Ocean Biology-induced Heating Effect，OBH）和间接物理过程致冷作用的相对强度决定了海温变化形态。具体而言，引入DCM分布后，东部较强的DCM上缘水体吸收更多太阳辐射，加热次表层水体，减弱海水层结，加强垂向混合，使更多冷水进入混合层，导致东部SST降低。而西部DCM较弱且较深，引起的局地海温变幅小于东部，因此SST纬向梯度增大，伴随着海表高度（Sea Level，SL）和温跃层深度纬向梯度增大。由此引起赤道两侧海表西向流速加强，赤道潜流（Equatorial Undercurrent，EUC）、浅层经向翻转环流（shallow Meridional Overturning Circulation，sMOC）和赤道上升流均加强，进一步降低了东部SST；在次表层，间接致冷效应超过了直接OBH效应，导致次表层水温大幅降低；同时，在海洋动力过程的作用下，赤道西太平洋略微升温。因此，引入DCM分布，使赤道东太平洋SST降低，西部SST升高。在年际变率试验组，月平均观测风场驱动下，引入含年际变率的DCM场，ENSO振幅增大约2.5%至15%。其中主要物理过程包括垂向混合、东部温跃层抬升、sMOC和海气热交换过程。以拉尼娜事件为例，DCM较浅且较强，海洋动力环境更不稳定，导致更强的间接动力过程，如垂向混合和赤道上升流被加强，从而使SST进一步降低；同时，东部次表层海温降低导致温跃层抬升，根据ENSO的温跃层反馈作用，更多冷信号进入表层，降低SST。因此引入含年际变率的DCM分布，将加强拉尼娜事件强度。在厄尔尼诺时期，DCM较弱且背景层结更稳定，因此DCM效应对厄尔尼诺强度的影响较小，各物理过程综合表现为略微增强厄尔尼诺。综上，引入含年际变率的DCM分布将增强ENSO振幅。

（2）在海气自由耦合情况下，显式调节HCM-AOPB海气耦合强度，研究了热带太平洋DCM效应对海温气候态和ENSO振幅的影响变化。海温气候态主要变化特征是：随耦合强度增大，DCM效应导致的多年平均SST变化，逐渐由“西部变暖，东部变冷”，转变为“西部变冷，东部变暖”。赤道东太平洋SST冷暖转变对应的耦合强度阈值在1.18至1.20之间。主要物理过程包括：海表风应力强迫、海气热交换、垂向混合和海洋环流过程。其中风应力的主要特征是：低耦合强度时，赤道中太平洋东风加强；高耦合强度时，跨赤道西南风加强。前者加强背景信风，使SST降低区域西扩；后者削弱赤道北侧的东北信风，根据风-蒸发-SST反馈机制，北侧SST逐渐转为升温并扩大升温范围。此外，垂向混合过程是主导的致冷动力过程，其在东部逐渐减弱，在西部逐渐加强；海洋环流起到次要的调制作用；海气热交换对SST演变起阻尼作用。在ENSO振幅方面，耦合强度低于1.22时，DCM效应引起的ENSO振幅增量随耦合强度递增；高于1.22时，ENSO振幅的增幅回落。Bjerknes指数分析表明，DCM效应引起的温跃层反馈、纬向平流反馈和Ekman反馈增强，是ENSO振幅增大的主要原因。随耦合强度增大，DCM效应引起风应力、次表层海温等物理量年际变率增大，从而使ENSO振幅增量逐渐增大。在耦合强度高于1.22时，因西部降温幅度增大，温跃层纬向梯度增幅回落，各物理量年际变率增幅回落，导致ENSO振幅增幅回落。

（3）在热带西太平洋，利用GOTM-TOPAZ，通过人为控制NFR廓线，研究了NFR变化导致的局地CHL和海温垂向分布变化。当NFR提高时，氮盐浓度增大，表层CHL浓度增大，OBH效应增强，更少太阳辐射穿透至次表层。因此，多年平均SST升高，次表层海温降低。此外，海气热交换对SST变化起到阻尼作用，使SST变幅小于次表层；次表层热扩散率减小，起到了热量再分配的作用。在年内尺度上，NFR提高时，SST季节振荡加强，冬季更冷，夏季更热。其中热扩散率和海气热通量的季节变化起到主导作用。在NFR降低时，以上情况相反。

综合上述结论，本文较为系统地分析了热带太平洋次表层DCM分布对海温气候态和ENSO振幅的反馈作用及其物理机制，揭示了气候模式对DCM效应的表征方式差异是海温模拟偏差的重要原因，证实了模式对海气耦合强度的模拟差异可通过调制DCM效应而增大海温模拟偏差。本文也初步探究了海洋固氮过程通过海洋内部途径影响局地海温分布的物理机制，为更全面理解海洋BGC过程对海温的反馈作用提供了新的见解和研究思路。

The ocean temperature in the tropical Pacific is a pivotal element of the global climate system, with its climatological distribution and interannual variability profoundly influencing the evolution of the global climate, holding significant climatic implications. The spatiotemporal distribution of the physical state, including the ocean temperature, in the tropical Pacific, represented by the phenomenon of El Niño and Southern Oscillation (ENSO), exerts a notable impact on the ocean biogeochemistry (BGC) processes. For example, the phase transitions of ENSO affect the nutrient concentration in the upper ocean of the tropical eastern Pacific, thereby modulating the distribution of chlorophyll (CHL) concentration. Conversely, CHL concentration affects the vertical attenuation rate of solar radiation in the upper ocean, redistributing the shortwave radiation absorption and its direct heating rate, modulating the local temperature distribution, and further influencing the basin-scale temperature distribution through the ocean thermodynamic processes. The CHL feedback effect reflects the importance of marine BGC processes in participating in the global climate evolution by feedbacking on the temperature distribution.

Differences in the representation of ocean biogeochemical processes by current climate models are an important reason for the bias in temperature simulations in the tropical Pacific. For example, observations reveal the widespread occurrence of Deep Chlorophyll Maximum (DCM) in the subsurface layer of the tropical Pacific, where the peak CHL concentration in the subsurface layer is higher than that at the sea surface. While most current models only consider the influence of sea-surface CHL distribution on the solar radiation penetration process, the contribution of vertical CHL distribution, i.e. the DCM effects, is less discussed. Therefore, the introduction of DCM effects in the shortwave radiation penetration scheme may result in differences in the modeling of the CHL feedback effect on temperature, which requires a quantitative study. In addition, in oligotrophic oceans such as the tropical western Pacific Ocean, the oceanic Nitrogen Fixation Rate (NFR) is high, and the nitrogen fixation process provides 30% to 50% nutrient input to the upper marine ecosystems, which affects the primary productivity and CHL distribution. Therefore, changes in NFR in this region may also have an indirect feedback effect on SST distribution by modulating CHL distribution, which is also understudied.

To investigate the feedback effects of DCM distribution and NFR changes on temperature field in the tropical Pacific, this study uses a Hybrid Coupled Model of Atmosphere, Ocean Physics and ocean Biogeochemistry (HCM-AOPB) and a one-dimensional ocean turbulence-BGC coupled model (Generic Ocean Turbulence Model–Tracers of Phytoplankton with Allometric Zooplankton, GOTM-TOPAZ), respectively. To top it off, the HCM-AOPB can simulate the DCM distribution well in the tropical Pacific, and allows explicit adjustment of the ocean-atmosphere coupling intensity, and the GOTM-TOPAZ can represent the oceanic nitrogen fixation process well. Both of them adopt a solar radiation penetration scheme based on the vertical distribution of CHL. This study primarily addresses three scientific questions: (1) the feedback effect of DCM distribution on temperature field under prescribed wind forcing; (2) in the case of free ocean-atmosphere coupling, the DCM feedback effect is modulated by the ocean-atmosphere coupling intensity; (3) the impact of NFR changes on the local temperature vertical distribution in the western tropical Pacific. The main conclusions are as follows:

(1) Under the prescribed wind forcing, the effects of DCM distribution on the climatological temperature field and ENSO amplitude in the tropical Pacific are investigated by using HCM-AOPB. In the climatology experiment group, under the climatological wind forcing, the introduction of DCM distribution resulted in a decrease of about 0.2 ℃  in the sea surface temperature (SST) and a drop of about 0.7 ℃  in the subsurface in the eastern equatorial Pacific and an increase of about 0.1 ℃  in SST in the west. The relative strength of the DCM-induced direct ocean biology-induced heating (OBH) effect and the indirect cooling effect of physical processes determines the pattern of sea temperature change. Specifically, with introducing the DCM field, the water at the upper edge of the strong DCM in the east absorbs more solar radiation, heating the subsurface water, weakening the vertical stratification, and strengthening vertical mixing, allowing more cold water to enter the mixed layer, resulting in an SST drop in the east. In contrast, the weak and deep DCM in the west causes a smaller local temperature change, thus increasing the SST latitudinal gradient, accompanied by the increased latitudinal gradients of sea level (SL) and thermocline depth. Consequently, the resulting enhanced westerly surface current on the north and south sides of the equator, Equatorial Undercurrent (EUC), shallow meridional overturning circulation (sMOC), and equatorial upwelling further reduce the eastern SST and overtake the direct OBH effect in the subsurface layer, resulting in a significant drop of the subsurface temperature. Additionally, the western equatorial Pacific warms slightly in response to ocean dynamical processes. Eventually, the DCM field has the effect of dropping the SST in the eastern equatorial Pacific and warming the west. In the interannual variability experiment group, driven by the observed monthly average wind fields, the introduction of DCM fields with interannual variability increases the ENSO amplitude by about 2.5% to 15%. The dominant physical processes include vertical mixing, the elevation of the eastern thermocline, sMOC, and ocean-atmosphere heat exchange processes. In La Niña events, for example, the DCM is shallower and stronger, and the ocean dynamical environment is more unstable, leading to stronger indirect dynamic processes, such as the intensified vertical mixing and equatorial upwelling, thus further reducing SST. Meanwhile, the drop in the subsurface temperature causes the elevation of the eastern thermocline, and more cold signals enter the mixed layer and decrease the SST, according to the thermocline feedback of ENSO. Thus, the La Niña events are enhanced. On the contrary, during the El Niño period, the DCM is weaker and the background stratification is more stable, so the DCM effect has less influence on the intensity of El Niño, and the combined effect of physical processes slightly enhances the El Niño. In summary, the introduction of the DCM field with interannual variability increases the ENSO amplitude.

(2) In the case of free ocean-atmosphere coupling, changes in the DCM effects on the climatological temperature field and ENSO amplitude in the tropical Pacific are investigated by explicitly adjusting the ocean-atmosphere coupling intensity. The main characteristic of changes in the climatological temperature field is that as the coupling strength increases, the multi-year average SST change caused by the DCM effect gradually shifts from “warming in the west, cooling in the east” to “cooling in the west, warming in the east”. The coupling intensity thresholds corresponding to the warm-cool transition of SST in the eastern equatorial Pacific is between 1.18 and 1.20. The key physical processes include surface wind stress forcing, ocean-atmosphere heat exchange, vertical mixing, and ocean circulation. Specifically, the main feature of wind stress in the central equatorial Pacific is that under low coupling intensity, the easterlies are strengthened; under high coupling intensity, the cross-equatorial southwest wind is intensified. The former enhances the background trade winds, leading to the westward expansion of the SST reduction region. The latter weakens the northeastern trade winds on the north side of the equator, where the SST shifts to warming, and the warming region expands gradually, according to the wind-evaporation-SST feedback mechanism. In addition, vertical mixing is the dominant cooling process, which is gradually weakened in the east and strengthened in the west. The ocean circulation plays a secondary modulating role, and the ocean-atmosphere heat exchange damps the SST evolution. In terms of ENSO amplitude, the increment in the ENSO amplitude caused by the DCM effect increases with the coupling intensity until the latter reaches 1.22; and the increment falls back when the intensity is higher than 1.22. The Bjerknes index analysis shows that the enhancements of the thermocline feedback, zonal advection feedback, and Ekman feedback due to the DCM effect are the dominant mechanisms for the amplified ENSO. Specifically, as the coupling strength increases, the DCM effect induces larger interannual variabilities of physical variables, such as wind stress and subsurface temperature, thus gradually increasing the increment of ENSO amplitude. When the coupling strength is above 1.22, the increments of interannual variabilities and ENSO amplitude fall back, because the temperature drops more in the west and the increment in the latitudinal gradient of thermocline depth falls back.

(3) In the western tropical Pacific, the changes in the local vertical distributions of CHL and temperature due to the variation of NFR are investigated by adjusting the prescribed NFR profiles in GOTM-TOPAZ. When the NFR is enhanced, the nitrate concentration increases, surface CHL increases, the OBH effect is enhanced, and less solar radiation penetrates into the subsurface layer. As a result, the multi-year mean SST increases and the subsurface temperature drops. In addition, the surface heat exchange damps the SST variation, which is smaller than that in the subsurface; the decreased thermal diffusivity in the subsurface redistributes the heat. On the annual scale, the seasonal oscillation of SST is enhanced with the increased NFR, with colder winters and warmer summers. The seasonal variations of the thermal diffusivity and surface heat flux play dominant roles in the seasonal temperature changes. The above processes are reversed when NFR decreases.

Combining the above conclusions, this study systematically analyzes the feedback effects of the subsurface DCM field on the climatological temperature field and ENSO amplitude in the tropical Pacific, and their physical mechanisms, revealing that the representation of DCM effect in climate models is an important reason for the biases in the temperature simulation. It also confirms that the simulated biases of ocean-atmosphere coupling intensity in climate models can further increase the SST biases by modulating the DCM effect. This study also preliminarily explores the physical mechanism of the feedback of the nitrogen fixation process on the local temperature distribution through the internal pathway in the ocean, providing new insights and research directions for a more comprehensive understanding about the feedback effects of BGC processes on the ocean temperature distribution.

目  录      

第1章 绪论       1

1.1 研究背景      1

1.1.1 热带太平洋海温时空分布对全球气候系统的重要影响       1

1.1.2 热带太平洋生物地球化学关键量的时空分布特征       4

1.1.3 海洋物理与生物地球化学过程的相互作用   11

1.1.4 海洋物理-生物地球化学耦合模式及其光学方案的发展     13

1.2 研究现状      15

1.2.1 海洋叶绿素浓度对海温的反馈作用       15

1.2.2 叶绿素反馈效应受海气耦合强度的调制影响       16

1.2.3 海洋固氮率对海洋碳循环的调制作用   18

1.3 主要研究内容与技术路线  19

第2章 数值模式、数据与分析方法       21

2.1 HCM-AOPB模式 21

2.1.1 大气风应力统计模式       21

2.1.2 海洋环流模式（OGCM）       22

2.1.3 海洋生物地球化学模式（BGC）   24

2.1.4 海洋生物光学方案   24

2.2 GOTM-TOPAZ模式    26

2.2.1 一维海洋湍流模式GOTM      26

2.2.2 海洋生物地球化学模式TOPAZ      27

2.2.3 海洋生物光学方案   29

2.3 观测与再分析数据      30

2.3.1 大气-海洋物理场数据     30

2.3.2 海洋生物地球化学数据   31

2.4 分析方法      32

2.4.1 一元线性回归与相关性分析   32

2.4.2 快速傅里叶变换分析       32

2.4.3 混合层热收支分析   33

2.4.4 BJ指数分析       33

第3章 深层叶绿素极大值效应对海温气候态和年际变率的影响       35

3.1 模式验证与试验设计  36

3.1.1 模式验证   36

3.1.2 试验设计   37

3.2 DCM气候态分布对海温的影响 38

3.3 DCM气候效应的物理机制 40

3.3.1 DCM对太阳辐射穿透过程的直接影响  41

3.3.2 DCM引起的间接物理过程      42

3.3.3 DCM效应的季节循环特征      47

3.3.4 DCM效应与表层CHL效应的对比 48

3.4 DCM年际变率对ENSO的影响  52

3.4.1 DCM年际变率效应对ENSO振幅的响应与反馈   52

3.4.2 DCM年际变率效应的物理机制      56

3.5 小结与讨论  63

第4章 深层叶绿素极大值效应受海气耦合强度的调制作用       66

4.1 试验设计      66

4.2 DCM效应对海温气候态分布的影响受海气耦合强度的调制 68

4.2.1 DCM气候态分布对耦合强度的响应      68

4.2.2 DCM效应对海温气候态分布的影响      68

4.2.3 低海气耦合强度下的DCM效应物理过程    75

4.2.4 中等海气耦合强度下的DCM效应物理过程 80

4.2.5 高海气耦合强度下的DCM效应物理过程    82

4.2.6 DCM效应受耦合强度调制作用的物理机制  85

4.3 DCM效应对ENSO的影响受海气耦合强度的调制  90

4.3.1 DCM年际变率对耦合强度的响应  90

4.3.2 ENSO振幅与周期的响应 92

4.3.3 基于BJ指数的物理机制分析  96

4.4 小结与讨论  99

第5章 热带西太平洋固氮率对海温的反馈作用   103

5.1 试验设计与模式验证  104

5.1.1 试验设计   104

5.1.2 模式验证   104

5.2 海温对固氮率变化的响应  107

5.3 固氮率导致生物地球化学过程变化  108

5.3.1 营养盐和CHL分布的响应      110

5.3.2 局地海洋碳循环的响应   112

5.4 叶绿素分布导致海温变化的物理机制      116

5.4.1 对太阳辐射垂向分布的直接影响   116

5.4.2 海气热通量对SST的影响       117

5.4.3 热扩散过程对位温垂向分布的调制作用       120

5.5 小结与讨论  121

第6章 研究结论与展望   125

6.1 主要研究结论      125

6.1.1 给定风场强迫下，DCM分布对海温气候态和ENSO振幅的影响     125

6.1.2 海气自由耦合情况下，DCM效应受海气耦合强度的调制作用  126

6.1.3 热带西太平洋NFR变化对局地海温分布的反馈作用  128

6.2 论文的创新点      129

6.3 尚存不足与未来工作展望  129

参考文献     131

附录 图片附录   145

致谢     151

作者简历及攻读学位期间发表的学术论文与其他相关学术成果 153