The impact of mixing parameterisation and bathymetry filtering on the simulated hydrography along steep continental shelf regions in terrain following ocean models
The impact of mixing parameterisation and bathymetry filtering on the simulated hydrography along steep continental shelf regions in terrain following ocean models
复制标题
混合参数化和测深过滤对海洋模型地形中陡峭大陆架区域模拟水文学的影响
DOI:
--
复制
发表时间:
2006
期刊:
影响因子:
--
通讯作者:
M. Lange
中科院分区:
文献类型:
--
作者:
M. Thoma;K. Grosfeld;M. Lange
General vertical coordinates in ocean models can be parameterised in three different ways: The most straightforward and mathematical easiest method are geopotential surfaces, where vertical coordinates are parallel to one another and perpendicular to the earth’s radius. One example for this type of ocean models is the Modular Ocean Model (MOM) (e.g., Bryan & Cox, 1968; Cox, 1984; Haidvogel & Beckmann, 1999; Griffies et al., 2003). One disadvantage of this z−coordinate type models is that several nodes (and therefore computational costs) are wasted if processes in the deep ocean and on shallow shelves are investigated. This type of ocean models is also not able to simulate overflow process adequately without additional empirical bottom boundary layer mixing schemes (e.g., Beckmann & Döscher, 1997; Ezer & Mellor, 2004). The perhaps most natural vertical coordinate system is an isopycnal one, where the surfaces are orientated along constant densities. No spurious diapycnal mixing occurs in this type of model, but the vertical coordinates are adaptive and must be recalculated each timestep. In addition, a special treatment of surface and bottom boundary layers is necessary, leading to additional complexity and computational costs. The most common isopycnic ocean model is the Miami Isopycnic Model (MICOM) (e.g., Bleck, 1978; Haidvogel & Beckmann, 1999). The third type of ocean models uses terrain following coordinates. These σ−coordinate models are most convenient to represent the bottom topography. So far they are mainly used in high resolution regional models where the pressure gradient error, associated with this type of models (e.g., Haney, 1991) is probably of minor importance (Ezer et al., 2002). Models using terrain following coordinates are the Princton Ocean Model (POM) (Blumberg & Mellor, 1987), the s-coordinate primitive equation model (SPEM) (Haidvogel et al., 1991) and Rombax (Thoma et al., 2005b).