Regarding_mixing_settings_in_roms_model

[tara]

Dear ROMS Users'
Can anyone suggest that whether the TS_SMAGORINSKY, UV_SMAGORINSKY both are using only TS_DIFF4 as well as UV_VIS4 only or it needs both harmonic as well as the biharmonic diffusion and viscosity coefficients. I am getting a problem of abnormal cooling in coastal regions, that is covering the whole study domain with respect to time. For, this issue I got a clue from forum that my diff2 coefficient is 0 (zero)so, I should turn off the TS_DIFF2 cpp option in order to save the cycle.

So, I turned off the both
TS_DIFF2 (off)
UV_VIS2 (off)
and I kept on only BIHARMONIC COEFFICIENTS
TS_DIFF4 (on)
UV_VIS4 (on)

But, model has blown up at 7th time step itself by giving
Minimum horizontal diffusion coefficient = 2.01368521E+10 m4/s
Maximum horizontal diffusion coefficient = 2.22459675E+10 m4/s

Minimum horizontal viscosity coefficient = 2.01368521E+10 m4/s
Maximum horizontal viscosity coefficient = 2.22459675E+10 m4/s
(BLOWN_UP at 7th time step only DIFF4,VIS4)

Can anyone tell whether the use of smagorinsky with only TS_DIFF2,UV_VIS2 options is logically and physically correct or not?
OR/

smagorinsky always needs a biharmonic coefficients of both diffusion and viscosity or only biharmonic viscosity coeff.

[kate]

Well, you know, the one true way to answer this is to look at the code. A search on SMAGOR points to ROMS/Nonlinear/hmixing.F, where one can find:

Code: Select all

#  ifdef TS_SMAGORINSKY
!
!  Smagorinsky diffusion at RHO-points.
!
#   ifdef TS_DIF2
            diff3d_r(i,j,k)=Hdiffusion(i,j)+                            &
     &                      SmagorCoef*omn(i,j)*DefRate
#   elif defined TS_DIF4
            diff3d_r(i,j,k)=Hdiffusion(i,j)+                            &
     &                      PecletCoef*(omn(i,j)**2)*DefRate

What I said in that other thread is wrong and you should be turning off the xxx4 ones instead.

[tara]

Thanks Kate,

I will try this suggestion and get back to u hopefully with good news.


With Regards--
Tara