[MITgcm-support] Climtological run

[Martin Losch]

Hi,

it takes much longer than 100 years to equilibrate an ocean model. It's more like 1,000-10,000 years, see the published literature for that. I recommend Danabasoglu (2004 in Ocean Modelling 7:323-341) as a starting point.

I recommend that you do not use monthly output for 2 and 3D fields (set dumpFreq/taveFreq=0) and only use the monitor output for monitoring the spin-up. But you should set your monitorFreq to something like 10-30 days (not 10.) so that you'll have monitor-output only every 50 time steps or so. This will already speed up your inegration a lot.

To further reduce the output  the minimum you need you can use the diagnostics package, see examples and documentation for that.

Good luck,

Martin


On Jun 17, 2010, at 2:21 PM, anuradha.modi at iccsir.org wrote:

> Dear Users,
> 
> I am new to the Ocean modelling.
> Recently I successfully installed the MIT-OGCM model on my workstation ( Interl core-2 Quad, Redhat Linux).
> After that I run the model for 100 years using climatological data, by changing the data file as follows
> # Time stepping parameters
>  &PARM03
>  startTime   = 0.,
>  endtime     = 3110400000.,
>  deltaTmom              =    1800.,
>  deltaTtracer           =   10800.,
>  deltaTClock            =   10800.,
>  tauCD                  =   36000.,
>  abEps                  = 0.1,
> #cAdjFreq               = -1,
>  cAdjFreq               =  0,
>  pChkptFreq             = 311040000.,
>  chkptFreq              = 311040000., 
>  dumpFreq               = 2592000.,
>  taveFreq               =2592000.,
>  dumpMean               =2592000.,
>  tauThetaClimRelax      =   864000.,
> #tauThetaClimRelax      =          ,
>  tauSaltClimRelax       =  2592000.,
>  monitorFreq            = 10.,
>  periodicExternalForcing=    .TRUE.,
>  externForcingPeriod    =  2592000.,
>  externForcingCycle     = 3110400
> 
> And written the output for every one month, It took almost 6 days for running. 
> After that I plotted the variables like Temperature, Salinity Timeseries and I feel that model is not stable yet (Timeseries Figure attached)
> 
> So, Now I gave a 300year run by changing 
> startTime   = 0.,
> endtime     = 9331200000.,
> 
> How to decide that the model become stable or not? ( By seeing the Timeseries figure whether we can decide model is stable or not?)
> 
> Please help me.......
> 
> Regards
> Anuradha Modi
> India
> 
> 
> <model-temp-son-mean.png><model-sal-son-mean.png>_______________________________________________
> MITgcm-support mailing list
> MITgcm-support at mitgcm.org
> http://mitgcm.org/mailman/listinfo/mitgcm-support