Our research group has developed a movable nested-grid ocean model for the purposes of examining the details of the local ocean response to mesoscale atmospheric phenomena in the context of the broader, basin-scale circulation.
Publications:
Ginis, I. , R.A. Richardson, and L.M. Rothstein, 1998: Design of a multiply - nested primitive equation ocean model. Mon. Wea. Rev., 126, 1054-1079.
Richardson, R.A., I. Ginis , and L.M. Rothstein, 1998: A numerical investigation of the local ocean response to westerly wind burst forcing in the western equatorial Pacific, J. Phys. Oceanogr. 29, 1334-1352.
Rowley C.D., and I. Ginis, 1999: Implementing a mesh movement algorithm into a multiply-nested grid primitive equation ocean model and its application to air-sea interaction studies. Mon. Wea. Rev. 127, 1879-1896.
The model is based on a primitive equation, multi-layered formulation. The mixed layer is treated as a homogeneous, turbulent boundary layer that exchanges momentum and heat with the atmosphere at its surface and with the thermocline by entrainment at its base. The stratified thermocline below is divided into an arbitrary number of numerical layers according to a sigma coordinate system, in a manner similar to the Gent & Cane sigma model. The vertical diffusion processes in the model include three major mechanisms of vertical turbulent mixing in the upper ocean, i.e., wind stirring, shear instability, and convective overturning. We have incorporated realistic coastline geometry, bottom topography and climatological surface forcing.
The nested-mesh system we employ belongs to a class of two-way nesting systems in which two neighboring mesh areas interact dynamically with each other. We use a nesting technique closely related to that employed in the GFDL hurricane model. The number of grids and the ratio of their resolutions can be freely adjusted by the user.
Our developmental efforts on the nested-grid model have included a variety of experiments ranging from highly idealized to quite realistic. In the idealized category, we have examined the propagation of Rossby and Kelvin waves along the equator, generated by the evolution of an initially imposed perturbation in the ocean thermocline. We have also observed the propagation of a dipolar vortex pair, consisting of two counter-rotating eddies. The former experiment principally explores the efficiency of momentum transport across the grid interfaces, while the latter examines the mass transport. Some sample results are shown below (click on images to enlarge).
|
|
|
|
|
|
In the category of more realistic experiments, we have examined the ocean response to a westerly wind burst and a tropical cyclone. In both of these experiments, the full Pacific basin was spun-up using annual mean wind forcing and surface heat flux for 3 years and 10 days prior to the imposition of the wind burst or cyclone. We compare the ocean response using a single grid at 1/2 degree resolution and a triply nested grid with resolutions of 1/2, 1/4 and 1/8 degree.
|
|
|
|
|
|
Our testing of the mesh movement algorithm has included both idealized and realistic simulations. Here we present some results of this testing. Animations are probably the best way to demonstrate the ability of a movable mesh model to efficiently perform fine-resolution numerical simulations. Please note that some of these animation files are quite large.
Rich Signell of the USGS has put together a page containing information about making and playing FLC/FLI format movies, such as we use here. For viewing these movies on UNIX systems, we have had good luck using the XAnim package. On a Windows PC, we have used the Autodesk, Inc., AAPlay package (waaplay.zip, 154298 bytes) with some success.
dipole2.fli (560466 bytes)
This animation compares single-mesh and moving -mesh simulations of a moving dipolar vortex modon in the 1.5 layer version of the model. The upper panel shows the zonal velocity from the solution using a single fine mesh of 1/30 degree resolution, and the lower panel shows the solution from a moving nested mesh configuration using grids of 1/30 (inner mesh) and 2/15 (outer mesh) degrees. The grid ratio of 1:6 used in this simulation is quite large. It is possible to use this large ratio here because of the compact and closed nature of the dipole eddy, which is largely isolated from the far field.
kelvin2.fli (1596182 bytes)
In this animation, we compare two solutions for the equatorial Kelvin wave response to an initial anomaly in the layer thickness of the 1.5 layer model. The upper panel shows the layer thickness anomaly using a moving triply-nested configuration, and the lower panel shows the solution using fixed triply-nested grids. The grid resolutions are 4/5, 4/15, and 2/15 degrees. The entire domain is not shown.
virgin2.fli (809926 bytes)
This movie shows the surface temperature change produced in a horizontally uniform ocean, initially at rest, when forcing, which is based on the intensity and track of Typhoon Virginia (1978), is applied. The horizontal resolutions of the nested meshes are 1, 1/3, and 1/6 degrees. The grid positions, and typhoon location and track are shown. Typhoon Virginia's track looped back on itself and the storm was largely stationary for several days while it was near peak strength. A strong cyclonic eddy formed as a result. In this simulation using the uniform ocean, this eddy stands out as a bulls-eye of large (>> 3 degC) surface temperature decrease.
We simulated the ocean response to Typhoon Roy (1988) by initializing the domain 124 degE to 70 degW, 30 degS to 30 degN using the January fields of a six-year spinup run. The fields were interpolated to a triply-nested configuration with resolutions of 1, 1/3, and 1/6 degrees. Typhoon Roy's track and intensity were used to create the typhoon forcing for the model.
royfull.fli (682712 bytes)
In this set of two animations, we show first the SST for the whole domain for the model integration period of 08 Jan to 14 Jan. The typhoon track and location, and the positions of the nested grids are shown to give an idea of the relative sizes of the meshes.
royreg2.fli (1896358 bytes)
The second animation shows a close-up view of the region 124 to 185 degW, 0 to 30 degN, with the position of the grids, the typhoon position and track, and the surface velocity vectors plotted over the SST. The SST decrease due to the storm forcing, and the strong inertial motion in the wake, are readily seen here.
Last modified: 8/31/00
