[MITgcm-support] advection scheme for T/S

[Martin Losch]

True, but I do not expect a stably stratified water column to mixing 
spontaneously, only because there is a little noise (much smaller than 
the stratification) on the profile.

M.

On Sep 20, 2004, at 4:38 PM, samar khatiwala wrote:

> Martin, good to hear I am not the only one. (FYI: I use
> the nonlimited DST3 (30) rather than the limited one (33).)
>
> Also, what makes you think that the 2d order centered scheme
> is giving the "true" answer? We are comparing one FD scheme
> against another.
>
> Samar
>
> On Mon, 20 Sep 2004, Martin Losch wrote:
>
>> steady state is very different from the case with the default 
>> advection
>> scheme. I have refrained from using 33 or 77 for temperature and
>> salinity since. (o:
>>
>
>
>> Martin
>>
>> On Sep 17, 2004, at 2:36 PM, Jean-Michel Campin wrote:
>>
>>> Hi Samar,
>>>
>>> The results could be significantly different between Centered, 2nd
>>> order
>>> with AB2, and DST3.
>>> For a passive tracer, the differences will be larger at high CFL
>>> number, more diffuve with DST3 and more dispersive with 2nd.O.
>>> Now, when you allow feedbacks to the dynamics (an active tracer
>>> like temperature), you can get different results even at small CFL,
>>> but in this case, it might be mainly due to vertical advection.
>>>
>>> This is all I can say at this point.
>>>
>>> See you,
>>>
>>> Jean-Michel
>>> _______________________________________________
>>> MITgcm-support mailing list
>>> MITgcm-support at mitgcm.org
>>> http://dev.mitgcm.org/mailman/listinfo/mitgcm-support
>>
>> _______________________________________________
>> MITgcm-support mailing list
>> MITgcm-support at mitgcm.org
>> http://dev.mitgcm.org/mailman/listinfo/mitgcm-support
>>
> _______________________________________________
> MITgcm-support mailing list
> MITgcm-support at mitgcm.org
> http://dev.mitgcm.org/mailman/listinfo/mitgcm-support