[MITgcm-support] Some questions about 3D float (flt example)

Martin Losch Martin.Losch at awi.de
Tue May 16 07:33:29 EDT 2017


Dear Yi Liu,

to me the code looks like “flt_noise” scales the dimensionless random number (between -.5 and .5) and then the “noise" is scaled by the amplitude of u1/v1. So for flt_noise=1, the “random noise” has an amplitude of 0.5*u1/v1 and you can scale it down/up by decreasing/increasing the dimensionless factor flt_noise.

Martin

> On 16. May 2017, at 11:56, liuyi315 <liuyi315 at mails.ucas.ac.cn> wrote:
> 
> Dear all users,
> 
> I am trying to do the 3D float experiments using MITgcm, but I have encountered a problem. I have noticed that we need set up the "flt_noise" in the data.flt and its default value is zero. I have found the code containing "flt_noise" in pkg/flt/flt_runga4.F:
>              
>                 u1 = u1 + u1*(PORT_RAND(seed)-0.5)*flt_noise
>                 v1 = v1 + v1*(PORT_RAND(seed)-0.5)*flt_noise
> 
> I am not sure the unit of flt_noise, so I don't know what value I should set. Could you give me some advice? All the suggestions are welcome. Thanks very much!
> 
> Best,
> 
> Yi Liu
> 
> 
> 
> 
> 发自网易邮箱大师
> 
> _______________________________________________
> MITgcm-support mailing list
> MITgcm-support at mitgcm.org
> http://mitgcm.org/mailman/listinfo/mitgcm-support




More information about the MITgcm-support mailing list