I'm now running a realistic case in a coastal area with baroclinic time step of 180s and a grid resolution of 3km. During the first run, it looked like there was some strange mixing in the temperature field, so I try to take Kate's advice in the pervious post:viewtopic.php?t=5213, #define LMD_SHAPIRO and fixed the problem.
However, when I tried to simulate the area for a month, it blew up, and I found while looking at his file that near the bottom, the velocity field at the boundary seemed to be noisy, as shown in the image below:
I tried defining UV_U3ADV_SPLIT to fix these spurious mixing, but without success. Do you have any ideas about this?
Another problem I'm in puzzled about that how to get the values of visc and tnu. I have seen https://www.myroms.org/wiki/Horizontal_ ... ent_Values for some information,but I still don't fully understand how to take these values. How to understand spindown time? and What are the values of delta x and beta?
Thank you for your interesting suggestion.my .h file is shown below.
