[MITgcm-support] nTimeStep
Jody Klymak
jklymak at uvic.ca
Fri Jun 9 12:22:59 EDT 2017
You still haven’t told us what the CFL criteria is in STDOUT. You may
need to make `monitorFreq` lower to get the monitor output. Best to set
this to the same value as deltaT until you find the problem. Its also a
good idea to save every model timestep so you can see where in the
domain the model is blowing up. That is often a good clue.
Good luck. Jody
On 9 Jun 2017, at 8:38, smaeilpour wrote:
> My data file is as:
>
> viscAz=1.E-3,
> viscAh=1.E-2,
> no_slip_sides=.FALSE.,
> no_slip_bottom=.TRUE.,
> diffKhT=1.E-2
> diffKzT=1.E-3,
> f0=0.00002,
> beta=2.4228E-11,
> eosType='LINEAR',
> tAlpha=2.E-4,
> sBeta =0.E-4,
> gravity=9.81,
> implicitFreeSurface=.TRUE.,
> exactConserv=.TRUE.
> nonHydrostatic=.FALSE.,
> hFacMin=0.2,
> implicSurfPress=0.5,
> implicDiv2DFlow=0.5,
> nonlinFreeSurf=3,
> hFacInf=0.2,
> hFacSup=1.8,
> saltStepping=.TRUE.,
>
> But in STDOUT I have:
>
> (PID.TID 0000.0001) hFacMin = /* minimum partial cell factor (hFac)
> */
> (PID.TID 0000.0001) 2.000000000000000E-01
> (PID.TID 0000.0001) ;
> (PID.TID 0000.0001) hFacMinDr = /* minimum partial cell thickness ( m)
> */
> (PID.TID 0000.0001) 0.000000000000000E+00
>
> after a while of run:
>
> EXTERNAL_FIELDS_LOAD, 1850 : iP,iLd,i0,i1= 12 1 12 1 ;
> Wght= 0.4964313272 0.5035686728
> cg2d: Sum(rhs),rhsMax = NaN Infinity
> EXTERNAL_FIELDS_LOAD, 1851 : iP,iLd,i0,i1= 12 1 12 1 ;
> Wght= 0.4964293981 0.5035706019
> cg2d: Sum(rhs),rhsMax = NaN 0.00000000000000E+00
>
>
> Cheers,
> Saeid
>
>
>
>
>
>
>
> --------------------------------------------
> On Fri, 6/9/17, Jody Klymak <jklymak at uvic.ca> wrote:
>
> Subject: Re: [MITgcm-support] nTimeStep
> To: mitgcm-support at mitgcm.org
> Date: Friday, June 9, 2017, 11:09 AM
>
>
> What did STDOUT say? What is
> your value of hFac min?
> Cheers, Jody
>
> On 9 Jun 2017, at 8:02, smaeilpour
> wrote:
> Thank you Jody,
>
> I restarted with a deltaT=5, that is small as following:
>
>
>
> startTime=0,
>
> endTime=86400,
>
> deltaT=5,
>
>
>
> but Model after a while (about 1500 of nTimeStep) becomes
> unstable, what solution else do you offer except that? and
> what CFL criteria else should i consider for it?
>
>
>
> Thank you again,
>
> Saeid
>
>
>
>
>
>
>
> --------------------------------------------
>
> On Wed, 6/7/17, Jody Klymak <jklymak at uvic.ca>
> wrote:
>
>
>
> Subject: Re: [MITgcm-support] nTimeStep
>
> To: mitgcm-support at mitgcm.org
>
> Date: Wednesday, June 7, 2017, 12:26 PM
>
>
>
>
>
>
>
> 9 times out of 10 you will get this if deltaT is too
>
> large. Look for advcfl_*_max in
>
> STDOUT.0000. They should be less than one, and
>
> preferably less than 0.5. A couple of things to be aware
> of
>
>
>
>
>
> setting hFacMin to be too
>
> small value can mean your vertical cell is very small,
> and
>
> the CFL
>
> condition becomes impossible to meet. hFacMin
>
> is the fraction of dR that a cell is allowed to
>
> be near the bottom, so dR=10 m and
>
> hFacMin=0.1 means the cell can be as thin as 1
>
> m. If your deltaT=50 then W
>
> can’t exceed 2 cm/s in that cell. If the topography
> is
>
> steep that is not a gigantic value for W near
>
> topography.you may have strong start-up
>
> transients in your solution that make the model behave
> more
>
> violently at the start than after it has been allowed to
>
> evolve. So check STDOUT.0000 and if the CFL
>
> criteria starts to drop again, you may be able to
> restart
>
> with a smaller deltaT.
>
>
>
>
>
> Good luck, Jody
>
>
>
> On 7 Jun 2017, at 8:35,
>
> smaeilpour wrote:
>
>
>
>
>
> Hi everyone,
>
>
>
> I am trying to run Internal wave, my bathymetry file is
>
> 2160*876 and precision 30s. I'd like to run it in 38
>
> layer and by 3500m depth. The Model with nTimeStep=100
> would
>
> be run well but whenever I increase it for more times I
>
> encounter with the following error:
>
>
>
>
>
>
>
> ABNORMAL END: S/R MON_SOLUTION, stops due to EXTREME
>
> Pot.Temp
>
>
>
>
>
>
>
> My error in STDER.0000 is:
>
>
>
> (PID.TID 0000.0001) SOLUTION IS HEADING OUT OF BOUNDS:
>
> tMin,tMax=-1.918+101 1.617+101
>
>
>
> (PID.TID 0000.0001) exceeds allowed range
>
> (monSolutionMaxRange= 1.000E+03)
>
>
>
> (PID.TID 0000.0001) MON_SOLUTION: STOPPING CALCULATION
> at
>
> Iter= 20
>
>
>
> (PID.TID 0000.0001) ***ERROR***S/R ALL_PROC_DIE: ending
> the
>
> run
>
>
>
>
>
>
>
> Of course, once I used startTime and endTime instead of
>
> nTimeStep but it didn't matter.
>
>
>
>
>
>
>
> My data file is as following:
>
>
>
> nIter0=0,
>
>
>
> nTimeSteps=8640,
>
>
>
> deltaT=50,
>
>
>
> dumpFreq=21600,
>
>
>
> pChkptFreq=0,
>
>
>
> chkptFreq=0,
>
>
>
> monitorFreq=2500.,
>
>
>
> periodicExternalForcing=.TRUE.,
>
>
>
> externForcingPeriod=2592000.,
>
>
>
> externForcingCycle=31104000.,
>
>
>
> tauThetaClimRelax=31104000.,
>
>
>
> abEps=0.02,
>
>
>
> &
>
>
>
>
>
>
>
> Is it possible that my domain is big for internal wave
>
> analysis that I get that error? if not so, how can I fix
>
> this problem? how can I run model for 4 years and more
> with
>
> this situation?
>
>
>
>
>
>
>
> Thank you in advance for any help,
>
>
>
> Saeid
>
>
>
>
>
>
>
> _______________________________________________
>
>
>
> 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
>
>
>
>
>
> _______________________________________________
>
> 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
>
>
> _______________________________________________
> MITgcm-support mailing list
> MITgcm-support at mitgcm.org
> http://mitgcm.org/mailman/listinfo/mitgcm-support
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mitgcm.org/pipermail/mitgcm-support/attachments/20170609/d5f1b308/attachment.htm>
More information about the MITgcm-support
mailing list