This appendix explains how to set up the CANDIE model for a simple test case application. The initialization will be set as an idealized case, so that it will not be necessary to interpolate temperature, salinity, or bathymetry from real data sets. This test case is to observe a Kelvin wave in a circular basin, and is a result of the research done with the DieCAST model which was written up by Beletsky et al. (1997). The circular basin is of diameter 100 km and has constant depth 100 m. The idealized temperature structure is chosen to resemble a strong thermocline that can be found in the Great Lakes in late summer: 20°C above 5 m, 5°C below 15 m, and a linear temperature decrease across the thermocline.
The grid spacing will be 1 km in both the north-south and east-west directions, with a one-grid cell ``ghost zone" around the domain, so that there will be 102 grid points in each direction. We will choose 12 grid cells in the vertical (13 interfaces including the top and bottom).
An inpulsive wind stress will be applied over 1.25 days which causes upwelling and downwelling on opposite sides of the basin. When the wind stops a new balance of forces is set up between the pressure gradient and the Coriolis force, and a Kelvin wave progresses cyclonically (counterclockwise in the Northern Hemisphere) around the basin.
First, obtain the CANDIE model code as explained in Section 2.2. The project directory you obtain CANDIE_V01 can be renamed to something more appropriate to the test case, such as IKW for ``Internal Kelvin Wave". The unix move command ``mv" will rename a directory and its subdirectories. However, the CANDIE model files do not refer directly to the project directory name, so this is not absolutely necessary.