[MITgcm-support] constant surface temperature boundary condition

[Martin Losch]

Hi,
Ryan already said everything, but it makes sense to have a look at the restoring term in the surface forcing terms. Assume that you turn off any other surface forcing (heat flux), then you have
dT/dt =  - 1/\tau ( T - T_{res} ),
when you discretize this in time with a simple Euler forward scheme (that's not how it is done in the model, but for illustration it's good enough) you get: 
T(n+1) = T(n) - \delta{t}/\tau ( T(n) - T_{res} )
so you can immediately see that when the restoring time scale \tau equal the time step \delta{t} you'll have 
T(n+1) = T_{res},
ie. you practically set your new temperature T(n+1) to the restoring temperature T_{res}.

Martin

On Aug 19, 2013, at 2:56 PM, Ryan Abernathey <ryan.abernathey at gmail.com> wrote:

> Ronald,
> 
> One way you can approximate a constant temperature bc is to use surface temperature relaxation with a very fast relaxation time. To do this, you specify a relaxation time scale in your data file in the PARM03 section, e.g.
>  tauThetaClimRelax=  5184000., # time in seconds = two month relaxation timescale
> To approximate a constant temperature, you would use a value close to your actual timestep (i.e. much smaller than two months). This would force the temperature back to the climatology at every timestep.
> Then you just have to specify an input file in the PARM05 section. e.g.
>   thetaClimFile=  'lev_sst.bin',
> I assume you know how to create these binary files if you have already run the tutorials.
> 
> Perhaps someone else knows a better way to do this, but I can't think of one that doesn't involve hacking the model code a bit.
> 
> -Ryan
> 
> 
> 
> On Mon, Aug 19, 2013 at 5:30 AM, zsucxr at gmail.com <zsucxr at gmail.com> wrote:
> Dear All,
>  
> I am learning the verification/tutorial_deep_convection case. Now I want to test a similar case with constant temperature (20 degree, for example) applied at the sea surface. As the tutorial only demonstrates setting surface heat flux, I wonder if it is possible to realize constant temperature boundary condition in MITgcm.
>  
> Any suggestion will be appreciated.
>  
> --
> Ronald Cen
>  
> 
> _______________________________________________
> MITgcm-support mailing list
> MITgcm-support at mitgcm.org
> http://mitgcm.org/mailman/listinfo/mitgcm-support
> 
> 
> _______________________________________________
> MITgcm-support mailing list
> MITgcm-support at mitgcm.org
> http://mitgcm.org/mailman/listinfo/mitgcm-support