[MITgcm-support] restart simulations using pickup files

[Martin Losch]

HI Kaveh,

as you found out yourself, the model gets confused about the forcing. I think that this is so, because you set baseTime and StartTime=0 which is in conflict with nIter0=7200. 
Assuming you don’t change the time step between pickups, all you should change is nIter0.

Martin

> On 3. Dec 2018, at 14:34, kaveh Purkiani <kavehpurkiani at googlemail.com> wrote:
> 
> Hi,
> I am trying to run a simulation (with a total length of 2 days) using a restart file after the first day. The model is forced laterally by U, V, Temp and Salt.
> The only modifications I did are renaming the pickup files and changing the nIter0 to the proper number to be able to start the simulation of the second day. The model works well with no error message but, the results are surprising to me.
> In the attached figure, two simulations for v component velocity for the same location are shown (red: with restart and black: normal run without restarting). On the first day, they are identical and they should continue the same but, the red line (run using restart files) starts deviating right after applying the pickup files. 
> From comparing the simulations it seems to me that model reads the pickup file properly, but the forcing is different.
> Similar differences are seen in other properties e.g., Temp, Sal and U component. 
> 
> I've done different tests to figure out where does this error come from but no success.
> Here is the PARAM3 in the data and everything else remains exactly similar for these two simulations!
> 
>  &PARM03
>  nIter0=7200, ----> 86400/12
>  baseTime=0.,
>  startTime=0.,
>  endTime=86400,
>  deltaT=12,
>  abEps=0.1,
>  momDissip_In_AB=.FALSE.,
>  pChkptFreq=0,
>  chkptFreq=86400,
>  dumpFreq=7200,
>  monitorSelect=2,
>  monitorFreq=100.,
>  periodicExternalForcing=.TRUE.,
>  externForcingPeriod=7200.,
>  externForcingCycle =2505600.,
>  &
> 
> Any help is really appreciated.
> Kaveh
> 
> 
> 
> 
> <Restart.pdf>_______________________________________________
> MITgcm-support mailing list
> MITgcm-support at mitgcm.org
> http://mailman.mitgcm.org/mailman/listinfo/mitgcm-support