Gallery / Simulation with regional domain using SCALE-DG (This page is under construction)
Baroclinic wave
This page shows a simulation result of idealized numerical experiment of baroclinic wave using the discontinuous Galerkin method (DGM). An example of the setting files is given in rootdir/model/atmosphere3d/test/case/baroclinic_wave/.
1. Description of dynamical core
Please see XX.
2. Experimental setup
The experiment setup is based on Ullrich et al. (2015). We consider a three-dimensional channel domain defined as $D = {(x,y,z)| 0 ≤ x ≤ L_x,0 ≤ y ≤ L_y,0 ≤ z ≤ z_T }$, where $L_x = 40000$ km, $L_y = 6000$ km, and $z_T = 30$ km. For the meridional, top, and bottom boundaries, no-flux boundary condition is applied. The zonal boundaries are periodic. As for rotation effect, we assume the beta-plane approximation.
The basic state is a steady-state geostrophically balanced flow, and the analytic expressions are derived by Ullrich et al. (2015). To trigger the baroclinic instability, we add a perturbation into the basic state of zonal flow.
In next section, we show results in the case of using the following model parameters
- Spatial resolution: NeX=80, NeY=12, NeZ=12, p=7 (LGL nodes)
- Explicit diffusion with 4th-order differential operator with the decay coefficient 4x10^15[ m4/s]
- Element-wise 16th-order exponential filter is applied for all modes and the factor of filter strength, $\alpha$, is 1.0; for example, the highest mode is dumped by $\exp{(-\alpha)}$.
- $\Delta t=30$ [sec] (with HEVI method and temporal scheme of ARK324 )
3. Result
4. Reference
-
Hesthaven, J. S., and T. Warburton, 2007: Nodal discontinuous Galerkin methods: algorithms, analysis, and applications, Springer Science & Business Media
-
Ullrich, P. A., K. A. Reed, and C. Jablonowski. “Analytical initial conditions and an analysis of baroclinic instability waves in f‐and β‐plane 3D channel models.” Quarterly Journal of the Royal Meteorological Society 141.693 (2015): 2972-2988.