[MITgcm-support] STDOUT

[Martin Losch]

Dear Xia,

when the model solves for hydrostatic pressure (or sea surface height), a conjugate gradient (cg) algorithms solves A x = b for x (routinte cg2d). By default the solver prints sum(b) and max(b) and that’s what you see in STDOUT.0000. You can suppress this output by setting debugLevel=-1., Usually these values fluctate a little during the integration; your example is not unusual and I don’t see how these numbers can help find a problem that you might have.
For debugging it’s a good idea to have debugLevel>0 (or even >1=default), in order to get more information on where the model starts to produce funny numbers. If you want us to help you need to provide more information about your problem.

Martin

> On 14 Dec 2015, at 04:16, 夏瑞彬 <xiarb at nuist.edu.cn> wrote:
> 
> Dear all
> 
>  I am trying to set up a regional experiment in the Prydz Bay with MIT-gcm. However, I have encountered a problem. When an experiment is being calculated, the last few linesof the  STDOUT.0000 is in the follows:
>  ...
>  cg2d: Sum(rhs),rhsMax =   5.74837247313553E-04  1.96447021948755E+00
>  cg2d: Sum(rhs),rhsMax =   5.75081665374455E-04  1.96633943695091E+00
>  cg2d: Sum(rhs),rhsMax =   5.75341936837503E-04  1.96815282242543E+00
>  cg2d: Sum(rhs),rhsMax =   5.75618703919645E-04  1.96990811715766E+00
>  cg2d: Sum(rhs),rhsMax =   5.75912579620308E-04  1.97160315426773E+00
>  cg2d: Sum(rhs),rhsMax =   5.76224177393669E-04  1.97323576861342E+00
> 
> So what does the two variables "Sum(rhs)" and "rhsMax" mean? What information can we get from their changes ? 
>  
> Thank you for your considerations!
> 
> 
> Xia
> E-mail: xiarb at nuist.edu.cn
> 
> 
>  
> 
> _______________________________________________
> MITgcm-support mailing list
> MITgcm-support at mitgcm.org
> http://mitgcm.org/mailman/listinfo/mitgcm-support