ROMS<kshedstrom version>blowing up problems with the "specific heat for seawater" "

[yangwu1215]

I'm recently studying the kshedstrom version <https://github.com/kshedstrom/roms/tree/master>, and trying to apply it on a freshwater lake.

The model works with the Cp=3985Joules/kg/degC, which denotes the specific heat for seawater in Subrouting <Modules/mod_scalars.F>. However, if I increase Cp to 4187.6Joules/kg/degC and keep the other options unchanged, the model blows up with the "DIAG Speed Trouble" after 686932 timestep.

Should I reduce the current timestep <15s for intermal mode, 3s for external mode> to solve the problem ?

Following is the running log.

Hope to get some help with this problem. Great thanks!

[kate]

You have gone for many steps before it blows up. You might as well try a shorter step and see if that helps. You also know that it blew up on the bottom: point 28, 7, 1. What do you have for bottom drag options? Are you limiting the bottom drag?

[yangwu1215]

Thanks, Kate!

(1) The model works if I reduce the timestep, 10s for the internal mode and 2s for the external mode.

(2) I activate the "UV_LOGDRAG" to use the logarithmic bottom stress in <set_vbc.F>. The maximum and minimum value for drag coefficient <Cdb> are 0.5 and 0.000001, just the default values in <mod_scalars.F>. Do you mean that I should modify the Cdb_max and Cdb_min?

[kate]

I meant you should #define LIMIT_BSTRESS, though I haven't tried it with LOGDRAG.