【作者】 于卫东；

【导师】 袁业立；

【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， 2004， 博士

【摘要】 气候变化作为一个全球性的环境问题受到广泛重视,成为海洋和大气科学研究的重点之一。海洋、大气作为气候系统的两个最重要的组成分量,它们在不同的时间和空间尺度上存在着相互作用。一般来讲,可以把海洋-大气的相互作用大致分为小尺度和大尺度两类。在小尺度上,海-气相互作用过程主要影响着两者之间的通量交换过程;在大尺度上,海-气相互作用过程决定了气候系统的低频变化,最显著的表现为以ENSO为代表的热带年际气候变化。本文的工作主要针对上述两种过程开展。第一方面,本文试图通过对小尺度海-气相互作用过程的显式刻画来改进耦合气候模式中的通量计算方案。目前利用海-气-冰-陆耦合数值模式来模拟气候变化过程以及人类活动对于气候的影响已成为气候变化研究的一个重要而活跃的领域。但是,耦合气候模式的气候模拟试验揭示了一个目前模式中普遍存在的问题:当耦合模式系统中的各个分量联立在一起,而且每个模式分量不再使用给定的边界强迫条件时,所模拟气候的统计状态将会逐渐偏离初始状态(也通常偏离观测到的气候状态),这种现象被称为气候漂移(Climate Drift)。这种气候漂移通常认为来自两种可能:第一,各个组成模式的初始条件之间存在不平衡;第二,各个组成模式间的通量交换不相容。为了克服这种模式缺陷,在实际的气候模拟中往往采用所谓的通量调整(Flux Correction)。目前的大气-海洋耦合模式中,气-海之间的动量、热量、水汽交换过程被简单地采用某种经验参数化方案(如Bulk formula)解决,而且各种通量的交换系数采用一个普遍适用的常数,这种处理方法值得置疑。过去二十年间,许多海上观测试验揭示了这样一个事实:海―气之间的动量通量强烈地依赖于海面状态,海浪这种小尺度过程对于交换过程具有重要意义,进而可能对大尺度现象具有影响。基于此Hasselmann(1991)提出了建立大气-海浪-大洋耦合模式的设想,本文第一部分工作实际上是发展这种大气-海浪-大洋耦合模式的一种努力,利用法国国家科研中心动力气象实验室(LMD/CNRS)所发展的LMD大气环流模式和德国MPI发展的WAM第三代海浪模式,实现了大气-海浪的耦合,探索了利用海浪模式改进传统的通量计算对于气候模拟的影响,结果表明耦合模式能够得到更好的气候模拟效果。同时,在耦合模式中发现了另外一种非常有价值的现<WP=6>象,即大气环流对于由海面粗糙度引起的机械强迫过程(mechanical forcing)呈现非常显著的响应。对比耦合模式和非耦合模式的气候模拟差异时发现,大气环流对这种机械强迫的响应在热带和中高纬度呈现不同的特征,在热带大气表现为斜压响应,而在中高纬度表现为相当正压(equivalent baratropical)响应。这种响应特性与大气对海表面温度强迫的响应特征类似,它在气候系统中的重要性值得进一步探讨。第二方面,与ENSO相关的热带大尺度海-气相互作用已被公认是影响全球气候年际变化的主要过程之一,ENSO的研究也被看作是上个世纪中海洋科学一个里程碑式的进展,但是关于ENSO的一些基本问题仍然有待于进一步认识。本文利用海洋、大气观测和再分析资料,把大气环流分解为无旋分量和无辐散分量两部分,阐明了它们在太平洋的EL Ni?o/La Ni?a和印度洋的偶极子过程中的主要变化特征,特别是讨论了其中的海-气正负反馈过程,以及太平洋和印度洋的海-气相互作用事件之间的区别与联系。更多还原

【Abstract】 Climate variability receives wide attention from the scientific community and the administrative organizations. Now it becomes one of the key research foci within the scope of climate science. Ocean and atmosphere, being the most important two components of the climate system, interact on different spatial and temporal scales. In general, this interaction can be classified into two classes, namely the small-scale interaction and large-scale one. For the small-scale case, the interaction strongly influences the air-sea flux processes. For the large-scale case, the interaction determines the low frequency variability of the climate system, of which the tropical inter-annual variability associated with ENSO is the most representative. The present work focuses on these two aspects.In the first case, the present work tries to provide one solution for the difficulties in determination of the air-sea flux in the coupled climate models, which is realized through explicitly description of the small-scale air-sea interaction process. The coupled atmosphere-ocean-ice-land model has now been one of the most important tools for climate simulation and assessment. However, it still suffers from some defects. The most significant one is the so-called climate drift. When the components of a coupled climate system model are joined together and freed from the constraints of fixed surface boundary conditions, the statistical state of the simulated climate diverges from the initial condition (and generally from the observed climate). Attention has been focused on two major sources of climate drift: lack of equilibrium in the initial condition of the component models (primarily the ocean) and incompatibilities in the fluxes across component interfaces. In the present generation of the coupled climate model systems, the air-sea flux is usually parameterized with the aid of the bulk formula and the transfer coefficient is a unified constant regardless of the real sea state, which is over-simplified and deserves questionable. However, many field experiments have revealed that the air-sea flux is strongly dependent on the sea state. Therefore the ocean surface wave, being a small-scale air-sea interaction process, may play a role in the climate <WP=8>system through its accumulative effects of the flux modulation. Thus Hasselmann (1991) conceived the establishment of the coupled atmosphere-wave-ocean model. This idea is hereby tackled in the present paper. As a first step, one AGCM (LMD) is coupled with one ocean surface wave model (WAM cycle 4) to test its feasibility. It is demonstrated that the coupled model, which takes explicit consideration of the sea state on the flux computation, is capable to give better climate simulation. In addition, the present paper also discovers an interesting fact that the oceanic mechanical forcing through the variation of the sea surface roughness state can influence the atmospheric general circulation in a significant way. Comparing the coupled model results with those from the AGCM alone reveals that the atmospheric response differs in the tropics and mid-to-high latitude region. In the tropical region, the atmosphere shows a baroclinic response while it shows an equivalent barotropic response pattern in the mid-to-high latitude region. This kind of atmospheric response resembles closely that from the SST forcing in the so-called AMIP experiments. Its importance for the climate system deserves further investigation.In the second case, the focus is given on the tropical large-scale ocean-atmosphere interaction associated with ENSO, which is considered as one of the most mechanisms of the global inter-annual climate variability. Progress on ENSO knowledge represents one of the milestone works for oceanography in last century. However, some basic aspects are still open to exploration. The present paper tries to illustrate the basic oceanic and atmospheric variability during the Pacific EL Ni?o/La Ni?a and Indian dipole cases based on the observation and re-analysis data. The atmospheric circulation is d更多还原