Can somebody please clarify few questions regarding settings in input file.
How to specify boundary conditions.. please have a look at the attached figure. I have grid that has west side land. right side open to ocean. I guessed that north and south are periodic.
Therefore i defined CLO , PER, CLO , PER in input file. Please look at the attached input file ocean_upwelling.in
1. Which three letter codes do i need to use as per my grid?
If the image is of bathymetry, it is not periodic north-south, so using PER is not appropriate for those boundaries. I usually use RadNud, but that requires values from outside around the boundary. You can get values from SODA or from HYCOM and interpolate to your grid. Or you can set it analytically to match your initial conditions.
I will attach figure of my grid and my input file.
Also, I kept NTIMES == 200
DT == 180
NDTFAST == 20
2. Is there any equation to calculate Ntimes and DT? or can can i just keep the number of time steps i want?
The maximum stable DT is set by features of your grid including grid spacing and maximum velocity, ocean depth, etc. I use trial and error to find something that works. From there, NTIMES follows from the length of the desired run. If DT=100 and you want to run for a day, NTIMES=864.
3.NDTFAST = Number of barotropic time-steps between each baroclinic timestep as per definition.. what is this exactly mean? I have tentatively kept NDTFAST == 20. does this affect my output?
I usually use something like 20 and don't worry about it.
4. i kept TCLINE == 50.0d0 ! critical depth (m) where as my grid file has maximum depth of 500metrs. and minimum depth of 10meters. Is this correct value for TCLINE?
Could be. What is your vertical density structure?
Excuse my ignorance. I am not a scientist. just a developer. but i was able to generate grid for my region using RNT toolbox. and also generated input files required to run the model. But i have so many unconnected dots that i am trying hard to connect. Please help..
What is your ultimate goal here? For much of this, you need to develop a "try it and see" attitude instead of asking for each little detail. You also need to develop a toolbox so you can actually see what happened from the model outputs.