5:15 PM - 7:15 PM
[MIS15-P06] Estimation of horizontal diffusion coefficient in the Indian Ocean sector of the Southern Ocean using HYCOM data
Keywords:horizontal eddy diffusion coefficients, Lagrangian tracking, the Indian Ocean sector of the Southern Ocean
Meridional circulation is often considered to be an important process supporting the deep ocean circulation, but in most cases, no structure indicating vertical circulation in the meridional direction is observed in the north-south cross section actually observed. In general, fluxes associated with vertical and horizontal eddy motions are considered to affect material circulation, except for direct transport by advection. However, direct observations of turbulence in the Southern Ocean by the Umitaka-Maru have not found any strong turbulent mixing regions that dominate material fluxes. Therefore, in this study, we used numerical experimental data to examine how materials spread horizontally due to eddies and how they should be evaluated.
In the evaluation of horizontal eddy diffusion coefficients, Lagrangian tracking is often used, which makes it easier to directly observe the movement of actual water masses, but it has many disadvantages, such as a short tracking period, a limited number of traceable particles, and a long measurement time interval. Due to these data shortages, it has been pointed out that the Lagrangian method tends to overestimate the diffusion coefficient compared to the Eulerian evaluation method. To clarify this problem and consider a solution, we used data from the "HYCOM Global Ocean Forecasting System (GOFS) 3.1" and investigated the relationship between the diffusion coefficient evaluation method using the Lagrangian method and material fluxes.
The target area was the Indian Ocean sector of the Southern Ocean, which is known to have active eddies. Horizontal eddy diffusion coefficients were calculated by varying the tracking time interval and the number of particles tracked, based on the displacement variance growth rate method of the flowing particle group, the method of Davis (1991), and the combination of the two methods of Zhurbas and Oh (2003). It was found that the horizontal eddy diffusion coefficient increases when the tracking time interval is extended, but the diffusion coefficient obtained is stable when the time interval is set to 24 hours or less. Similarly, when the number of tracked particles is increased, the value is stable when 66 particles or more are used per 5° × 5° grid. In this study, we also performed a Lagrangian tracking evaluation method that combines the displacement variance growth rate method and the method of Davis (1991), and compared it with the meridional temperature and salinity fluxes and eddy kinetic energy (EKE). The highest correlation with temperature and salinity fluxes was observed when the displacement variance growth rate was applied in both the upstream and downstream directions, and the highest correlation with EKE was observed when the Davis (1991) method was applied in both the upstream and downstream directions. This is thought to be because the Davis (1991) calculation formula is ultimately an expression that is mathematically almost similar to the calculation of EKE.
The results of this study show that the Lagrangian method does not deviate from the Eulerian method if there is a sufficient amount of data. At the presentation, we will also introduce the distribution of diffusion coefficients at various depths in the Indian Ocean sector of the Southern Ocean obtained from HYCOM data.
In the evaluation of horizontal eddy diffusion coefficients, Lagrangian tracking is often used, which makes it easier to directly observe the movement of actual water masses, but it has many disadvantages, such as a short tracking period, a limited number of traceable particles, and a long measurement time interval. Due to these data shortages, it has been pointed out that the Lagrangian method tends to overestimate the diffusion coefficient compared to the Eulerian evaluation method. To clarify this problem and consider a solution, we used data from the "HYCOM Global Ocean Forecasting System (GOFS) 3.1" and investigated the relationship between the diffusion coefficient evaluation method using the Lagrangian method and material fluxes.
The target area was the Indian Ocean sector of the Southern Ocean, which is known to have active eddies. Horizontal eddy diffusion coefficients were calculated by varying the tracking time interval and the number of particles tracked, based on the displacement variance growth rate method of the flowing particle group, the method of Davis (1991), and the combination of the two methods of Zhurbas and Oh (2003). It was found that the horizontal eddy diffusion coefficient increases when the tracking time interval is extended, but the diffusion coefficient obtained is stable when the time interval is set to 24 hours or less. Similarly, when the number of tracked particles is increased, the value is stable when 66 particles or more are used per 5° × 5° grid. In this study, we also performed a Lagrangian tracking evaluation method that combines the displacement variance growth rate method and the method of Davis (1991), and compared it with the meridional temperature and salinity fluxes and eddy kinetic energy (EKE). The highest correlation with temperature and salinity fluxes was observed when the displacement variance growth rate was applied in both the upstream and downstream directions, and the highest correlation with EKE was observed when the Davis (1991) method was applied in both the upstream and downstream directions. This is thought to be because the Davis (1991) calculation formula is ultimately an expression that is mathematically almost similar to the calculation of EKE.
The results of this study show that the Lagrangian method does not deviate from the Eulerian method if there is a sufficient amount of data. At the presentation, we will also introduce the distribution of diffusion coefficients at various depths in the Indian Ocean sector of the Southern Ocean obtained from HYCOM data.