Basic research on pathways of Pacific deep water with high-performance ocean circulation model.

为了阐明北太平洋深层环流的三维结构，利用具有真实海底地形的全球海洋环流模式进行了一系列数值试验。所得结果总结如下：1。通过萨摩亚通道的深/底水向北输送依赖于北太平洋深层垂直涡旋扩散系数的垂直剖面，而其通道则受底部地形的强烈控制。在夏威夷东北盆地或热带盆地，水最终向北流动。因此，我们可以在东北太平洋盆地的深层中找到最古老的水。水体在盆地内上涌，经纬向西流向西输送，并在埃克曼吸力作用下浮出海面。与热带盆地的上升水相结合，深层源的地表水以印尼通流的形式流出。年龄示踪剂实验和拉格朗日粒子实验很好地定义了这些途径。如果给出垂直涡旋扩散系数在上深层急剧下降的垂直剖面，则上深层（下温跃层）的大部分西进水沿上深层深度（2000-3000m）界定的西部边界返回太平洋南大洋（Tsujino,1998）。根据观测研究，在地转分析的基础上，这种回流是在热带太平洋东部通过一些示踪观测实现的，而不是通过模式实验提出的西边界。我们检查了最近沿WHP剖面的示踪观察，即delta14C， delta3He和其他营养物质。结果所有的示踪模式都表明没有向南流过太平洋赤道的气流。因此，我们得出结论，为了同时满足示踪观测和模式结果，“太平洋经向平面上层深层无回流”是必要条件。

A series of numerical experiments are made by use of a global ocean general circulation model with realistic bottom topography, in order to clarify 3D structure of the North Pacific deep layer circulation. Results obtained are summarized as follows1. Northward transport of deep/bottom waters through Samoan passage is dependent upon vertical profile of vertical eddy-diffusivity in the North Pacific deep layer, whereas their pathways are strongly controlled by bottom topography. Waters moving northward eventually upwell either in Hawaiian north-eastern basin or in tropical basins. Therefore, we can find the oldest water in the deep layers of the north-eastern Pacific basin. Waters upwelled in the basin, are transported westward by the zonal westward currents and emerged to the sea surface by the Ekman suction. Joined by the upwelled water in the tropical basin, surface waters with deep layer origin, flow out as Indonesian through flow. These pathways are well defined by age-tracer experiments and Lagrangean particle experiments.2. If we give vertical profile of vertical eddy-diffusion coefficient with sharp decrease in the upper deep layer, most of the westward moving waters in the upper deep (lower thermocline level), returns to the Pacific southern ocean (Tsujino,1998) along the western boundary, defined by the upper deep depth(2000-3000m).3. According to the observational studies, based on the geostrophic analysis, this return flow is realized in the eastern tropical Pacific with some tracer observation, contrary to via the western boundary as suggested by the model experiments.4. We checked recent tracer observation along WHP sections, i.e., of delta14C, delta3He, and other nutrients. It turns out that all the tracer patterns indicate no southward flow across the Pacific equator.5. Therefore, we conclude that, to satisfy both the tracer observation and the model results, "no return flow in the upper deep layer in the Pacific meridional plane" is required.