[MITgcm-support] query ?
Martin Losch
mlosch at awi-bremerhaven.de
Thu Jun 16 03:05:19 EDT 2005
Hi there,
I repeat: the plotting routines are really only an example and at most
a starting point for your own set of plotting routines.
deltaTclock/deltaTtracer:
I don't think that having different values for deltaTtracer and
deltaTclock is a good idea. deltaTclock is the "real" time step in the
model: based on deltaTclock, forcing fields are loaded and output is
written, etc. Having a different deltaTtracer means that you use a
different timestep for your tracer equations than for your forcing
fields, something you probably do not want. [In the case of deltaTmom
and deltaTtracer this is the desired behavior (for spin-up integrations
only), in order to speed-up the equilibrations times for for temparture
and salinity distribution in the deep ocean. Have a look at the
relevant papers (Bryan 1984, JPO, and more recently Danabasoglu 2004,
Ocean Modelling, etc.) to understand how that works.]
I personally would make sure that for asynchronous time stepping
deltaTclock=deltaTtracer and for synchonous timestepping
deltaTclock=deltaT (by not setting deltaTclock in data!!).
Also, as written in the data file of global_ocean.90x40x15, I recommend
setting deltaTfreesurf=deltaTtracer, if you use an model version >
checkpoint45.
Hope that helps,
Martin
On Jun 15, 2005, at 1:07 PM, Amitabh Mitra wrote:
>
>
> Dear Sir,
>
> I have run the global_ocean.90x40x15 experiment for
>
> one year with a resolution of 1x1.
>
> In the 'data' file I am putting the value for
>
> deltaTtracer=800. and deltaTClock=172800.
>
> While plotting the output for the last iteration (180)
>
> I am changing the line no. 4 in mit_timesteps.m as
>
> deltat = mit_getparm('D:\MITrun\data','deltaTClock');
>
> i.e. putting 'deltaTClock' in place of 'deltaTtracer',
>
> (to get proper headings in the plot).
>
> Is this change logical and valid ? Will the vast difference
>
> between the two timings have any effect on the model output ?
>
> Thanking You,
>
> With regards,
>
> Amitabh Mitra
>
>
>
>
>
> <inbox.gif>_______________________________________________
> MITgcm-support mailing list
> MITgcm-support at mitgcm.org
> http://mitgcm.org/mailman/listinfo/mitgcm-support
More information about the MITgcm-support
mailing list