Impact of physically relevant and numerically induced diapycnal mixing and meso-scale dissipation on meridional mass and tracer transports in the Southern Ocean

南大洋经向翻转环流（MOC）由两支组成，绕极上层深层水（UCDW）向极地方向流动，向上穿过南极绕极流（ACC），在ACC向极地一侧上升至表层，然后作为近表层流向赤道方向回流。已知该上肢在很大程度上是绝热的。与此形成鲜明对比的是，由下绕极深水（LCDW）形成的MOC的下肢在靠近南极洲的地方上升，在表面附近大幅冷却，并在高度湍流混合下对流下降，形成南极底层水（AABW）。虽然绝热的上肢是由一个向北的埃克曼运输和一个相反的中尺度涡动运输的不平衡，下肢的动力学是更加复杂的，由于额外的意义的diapycnal混合。因此，一个很好的量化这些动态需要一个正确的表示小尺度（diapycnal）和中尺度（横向）涡混合。 另一方面，人们早就知道，特别是在SO中，由于平流项的离散化，海洋环流模型中的大数值混合和耗散模糊了纵摇混合的表示，从而极大地限制了海洋模型的可预测性。因此，本项目的目的是使用并进一步开发一种新的分析工具，用于数值混合和耗散，以量化有效的混合和耗散，由显式参数化和数值诱导值的总和给出。通过估算实际的有效混合和耗散率，利用我们在数值数学和小尺度湍流参数化以及大尺度高分辨率建模方面的综合专业知识，我们能够第一次评估SO的实际模式对径向混合和数值耗散的敏感性，并了解中尺度（横向）和小尺度之间的相互作用（横辐）涡旋混合在MOC的下肢上。

The Meridional Overturning Circulation (MOC) in the Southern Ocean (SO) is composed of two limbs, the Upper Circumpolar Deep Water (UCDW) flows polewards and upwards across the Antarctic Circumpolar Current (ACC), upwells to the surface at the poleward flank of the ACC and then returns equatorwards as a near surface current. This upper limb is known to be largely adiabatic. In contrast to that, the lower limb of the MOC formed by the Lower Circumpolar Deep Water (LCDW) upwells closer to Antarctica, cools down substantially near the surface and convects down at high turbulent mixing to form the Antarctic Bottom Water (AABW). While the adiabatic upper limb is driven by an imbalance of a northward Ekman transport and an opposing meso-scale eddy transport, the dynamics of the lower limb is more complex due to the additional significance of diapycnal mixing. Thus, a good quantification of these dynamics requires a correct representation of both small-scale (diapycnal) and meso-scale (lateral) eddy mixing. On the other hand, it has long been known that in particular in the SO, large numerical mixing and dissipation in ocean circulation models due to the discretisation of the advection terms obscures the representation of diapycnal mixing and thus strongly limits the predictability of ocean models. It is therefore the aim of this project to use and to further develop a novel analysis tools for numerical mixing and dissipation to quantify effective mixing and dissipation, given by the sum of explicitly parameterised and numerically induced values. By estimating realistic effective mixing and dissipation rates, using our combined expertise of numerical mathematics and small-scale turbulence parameterisation and large-scale high-resolution modelling, we are able to the first time to assess the sensitivity of realistic models of the SO to diapycnal mixing and numerical dissipation and to understand the interplay between meso-scale (lateral) and small-scale (diapycnal) eddy mixing on the lower limb of the MOC.