[MITgcm-support] Buoyancy anomaly

[Martin Losch]

Lee,

1. I am not sure, if you are looking at the appropriate diagnostics.  
the density anomaly for a linear equation of state is
rho-rhoConst = rhoNil*tAlpha*(theta(i,j,k,bi,bj)-tRef(k))
Your stratification is the vertical gradient of this.  I believe the  
diagnostic for  this is:
       diagName  = 'DRHODR  '
       diagTitle = 'Stratification: d.Sigma/dr (kg/m3/r_unit)'
       diagUnits = 'kg/m^4          '

2. If you use non-hydrostatic dynamics with a coarse resolution (what  
is coarse? O(1 km) in the horizontal? the vertical resolution is  
usually not that crucial for this) without any vertical mixing  
scheme, you may not have any vertical mixing at all. The simplest  
choice of vertical mixing scheme is to turn on convective adjustment  
( for example, cAdjFreq = -1., ), or use implicit diffusion or KPP.  
That should remove any static instabilities (in fact the convective  
adjustment implementation in the MITgcm is quite slow in removing  
static instabilities). If, on the other hand, you want to resolve  
vertical convection with non-hydrostatic dynamics, you need to  
resolve the convective scales (I use a resolution < 100m for that at  
least).

3. A remark to Christopher's statement:
> (the fact that MITgcm uses relaxation instead of true surface flux  
> complicates things, but doesn't change the basic result.)
In fact, with the MITgcm you can use EITHER relaxation, OR a true  
surface flux, OR both. To turn on relaxation you can set (in data)  
tauThetaClimRelax (PARM03) and  thetaClimFile (PARM05), to use a true  
flux set surfQFile (PARM05).

4. PLEASE do send large files to this list (you last email was  
4.8MB). There are better formats than eps (jpeg, png, ...), or you  
can send links to the files, I you cannot convert them (ImageMagick's  
convert command is very hand for converting between graphics formats).

Martin

On 22 Mar 2007, at 07:07, Yi HanSoo wrote:

> Hi Laure and Christopher
>
> Thanks for your replies.
> In my modeling of lake dynamics,
> I use the non-hydrostatic through the all experiments
> with rather coarse vertical resolution,
> and as you mentioned, the model is forced by momentum flux, heat  
> flux and pressure at the surface with their hourly variations  
> showing the surface warming and cooling clearly from the results.
> As shown from the plots, the depth (30m) is the almost maximum  
> value in the lake,
> and the negative stratification is getting larger and stronger.
> I am not clear with what you said
> so could you point me a reference about it?
>
> Lee
> -- 
> Lee HanSoo
>
> Ph.D student
> Research Center for Fluvial and Coastal Disasters,
> Disaster Prevention Research Institute, Kyoto University,
> Gokasho, Uji-si, Kyoto, Japan
> 611-0011
>
> _______________________________________________
> MITgcm-support mailing list
> MITgcm-support at mitgcm.org
> http://mitgcm.org/mailman/listinfo/mitgcm-support