[MITgcm-devel] experiment to check conservation in seaice

[Jean-Michel Campin]

Martin,

With the set-up you gave me, I run with ifort and 2 mpi proc,
and when I run the check_conserve script, I got:
> -New- salt budget: ocean
>   dSalt = 0, oceSflux = 0, SFLUX = 4.793766e-15
>   residual = -4.793766e-15 g/m^2
I don't know why you are getting a residual of ~1.e-08

Cheers,
Jean-Michel

On Fri, Nov 12, 2010 at 05:29:59PM +0100, Martin Losch wrote:
> Thanks,
> 
> this was an obvious case of a serious knot in my head.
> 
> for the salt budget: is a residual of -4.660087e-08 g/m^2 OK? I'm not sure where this comes from (as the SFLUX is 1e-15)
> 
> Martin
> 
> On Nov 12, 2010, at 5:01 PM, Jean-Michel Campin wrote:
> 
> > Hi Martin,
> > 
> > I've checked the set-up, looks OK to me.
> > (relatively simple on the ocean side since there is no
> > relaxation on SSST & SSS).
> > I've changed check_conserve.m (attached), to check over the period
> > here it's 1 year) instead of per unit of time (/s), 
> > since there was already some conversion to /year, it's simpler.
> > 
> > I am getting a good conservation for the ocean (vol,salt,heat),
> > but did not look at seaice & atmos interface.
> > 
> > Cheers,
> > Jean-Michel
> > 
> > On Tue, Nov 02, 2010 at 03:23:13PM +0100, Martin Losch wrote:
> >> Hi Jean-Michel,
> >> 
> >> I finally set up a simple experiment based on the verification/global_ocean.cs32x15 with seaice and seaice thermodynamics (no winton). I failed miserably in trying to check the salt and heat conservation even for the ocean. The only thing that i managed to get right is volume conservation in the ocean, which is almost trivial of course. The seaice model has a drift of about 4mm/y as expected from earlier experiment.
> >> 
> >> I put the configuration on faulks:~mlosch/cs32conserve.tgz (just the code and input directories).
> >> 
> >> In the input directory there's a matlab script that tries to make use of the diagnostics output in netcdf (check_conserve.m). Please teach me how to close the heat and salt budgets.
> >> 
> >> Martin
> >> 
> >> 
> >> _______________________________________________
> >> MITgcm-devel mailing list
> >> MITgcm-devel at mitgcm.org
> >> http://mitgcm.org/mailman/listinfo/mitgcm-devel
> > <check_conserve.m>_______________________________________________
> > MITgcm-devel mailing list
> > MITgcm-devel at mitgcm.org
> > http://mitgcm.org/mailman/listinfo/mitgcm-devel
> 
> 
> _______________________________________________
> MITgcm-devel mailing list
> MITgcm-devel at mitgcm.org
> http://mitgcm.org/mailman/listinfo/mitgcm-devel