[MITgcm-support] advection scheme

[Julie Deshayes]

thank you christopher for your help.
the nonlinear flux limiter scheme (33) indeed helps enhancing EKE, both 
max and mean.
I am still playing with different numbers of viscAh, viscA4, diffKhT and 
diffK4T to see which gives most "realistic" results.
of course the eddy generation mechanism is what matters in the end, and 
I still have a lot of work to do on that !

Julie

Christopher L. Wolfe wrote:
> 
> Julie,
> 
> I think you could get away with reducing both your horizontal viscosity 
> and your horizontal diffusivity. I'm running a model with 5 km 
> resolution and I have Ah = 10 and Kh = 1e-5. I also have a 
> hyperviscosity of A4 = 5e8 to prevent too much energy from accumulating 
> at the grid scale. Everything is quite stable with a time step of 500 s. 
> The smallness of my horizontal Laplacian viscosity is limited by the 
> need to resolve the Munk layers on the western boundary. Since you have 
> beta = 0, you may be able to push the horizontal viscosity even lower.
> 
> I'm also using a nonlinear flux limiter (scheme 33) with good results. 
> I'm not sure what effect it has on the eddies; I use it primarily to 
> avoid the generation of spurious temperature extrema.
> 
> What kind of surface forcing are you using? That can have a great impact 
> on how many eddies form.
> 
> Christopher
> 
> -----------------------------------------------------------
> Dr. Christopher L. Wolfe                   858-534-4560
> Physical Oceanography Research Division    OAR 357
> Scripps Institution of Oceanography, UCSD  clwolfe at ucsd.edu
> -----------------------------------------------------------
> 
> 
> 
> On Aug 27, 2007, at 12:48 PM, Julie Deshayes wrote:
> 
>> hello,
>> I am running a configuration of the model at high latitude (f=1e-4 
>> /s). my domain is 700km x 700km and the resolution is 10 km. I do not 
>> have as many eddies as I expected, so I am running some tests to 
>> decrease the values of viscAh and diffKhT that I am using now (see 
>> data file at the end of my email). I hope that this will give me more 
>> eddies. Besides, I am thinking of changing the advection scheme 
>> (currently centered 4th order) to a non-linear flux limited scheme. 
>> Would you recommend me to do that ? or, considering my configuration, 
>> what advection scheme would you recommend me to use ?
>> thank you in advance for your help,
>> Julie
>>
>>
>>
>>
>> # Model parameters
>> #
>> # Continuous equation parameters
>>  &PARM01
>>  readBinaryPrec=64,
>>  useSingleCpuIO=.TRUE.,
>>  tRef=  5.35, 5.20, 5.05, 4.9, 4.75, 4.6, 4.45, 4.3, 4.15, 4.0,
>>  sRef=10*10.,
>>  rigidLid=.FALSE.,
>>  implicitFreeSurface=.TRUE.,
>>  hFacMin=0.2,
>>  no_slip_sides=.TRUE.,
>>  no_slip_bottom=.TRUE.,
>>  bottomDragLinear=1.e-6,
>>  eosType='LINEAR',
>>  tAlpha=2.E-4,
>>  sBeta =0.,
>>  saltStepping=.FALSE.,
>>  tempAdvScheme=4,
>>  saltAdvScheme=4,
>>  diffKhT=120.,
>>  diffKzT=1.E-5,
>>  f0=1.0e-4,
>>  beta=0.E-11,
>> # useCDscheme=.FALSE.,
>>  useNHMTerms=.FALSE.,
>>  implicitDiffusion=.TRUE.,
>>  ivdc_kappa=1000.0,
>>  viscAz=2.E-4,
>>  viscAh=200.,
>>  &
>> #
>> # Elliptic solver parameters
>>  &PARM02
>>  cg2dMaxIters=1000,
>>  cg2dTargetResidual=1.E-11,
>>  &
>> #
>> # Time stepping parameters
>>  &PARM03
>> startTime=622080000.,
>>  endTime=933120000.,
>>  deltaT=1500.0,
>>  abEps=0.1,
>>  pChkptFreq=0.0,
>>  chkptFreq=311040000.,
>>  dumpFreq=311040000.,
>> # taveFreq=86400.,
>>  cAdjFreq=0,
>>  &
>>
>> _______________________________________________
>> 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
>