[MITgcm-support] diagnostics package

[jeff polton]

After a conversation with Baylor I learnt the following in response to 
this thread:

1) There is no list of formulae, as such, for the diagnostic package 
output, since the routines spit out the diagnostics on the variables 
you give it which can be called from anywhere in the code.
So you have to grep to find where "call write_diagnostics" appears and 
see how the relevant variable is defined.

2) If you are using the flux limiting schemes and if you are bothered 
about closing budgets and if you think there might be steep gradients 
in your fields then it might be better to use velocities backed out of 
fluxes rather than the uvel, vvel dumps because of the flux 
corrections.

I hope I haven't misrepresented what Baylor said too much.

Jeff

On Apr 12, 2005, at 1:41 PM, jeff polton wrote:

> Jean-Michel,
>
> How did you know the exact formulaic definitions of ADVy_TH and other 
> variables? I can't find anything other than their string titles (in 
> available_diagnostics.log).
>
> Thanks,
>
> Jeff
>
> On Apr 7, 2005, at 3:10 PM, Jean-Michel Campin wrote:
>
>> Hi Jeff,
>>
>> You are right and there are few verification exemples that are
>> using pkg/diagnostics. The one you can look at is:
>> ideal_2D_oce/input/data.diagnostics
>> with:
>>>  frequency(4) = 20,
>>>   fields(1,4) = 'ADVy_TH ','ADVr_TH ',   <- advective fluxes of Temp
>>>                 'DIFy_TH ','DFrE_TH ',   <- diffusive fluxes of Temp.
>>>                 'DFrI_TH ',              <- diffusive fluxes of Temp.
>>> #                'ADVx_SLT',
>>>   filename(4) = 'flxDiag',
>>
>> that will write time-average fluxes (as computed by the advection 
>> scheme)
>> to the file: flxDiag.
>> Note that:
>> 1) those outputs are already multiplied by the grid-cell cross area
>>  (e.g.: ADVy_TH = vT * dx*dz)
>> 2) you will need to add the zonal advective flux: 'ADVx_TH '
>>   (not here since it's a 2D lat x depth set-up, with no salinity)
>> 3) frequency is specified as an iteration number.
>>
>> Jean-Michel
>>
>> On Thu, Apr 07, 2005 at 10:39:13AM -0700, jeff polton wrote:
>>>
>>> I am using a flux limited advection scheme and want to dump various
>>> time-averaged terms such as uT, vT, wT. I am under the impression 
>>> that
>>> the timeave package doesn't do the calculations correctly for these
>>> advection schemes but that the diagnostics package would.
>>>
>>> Are there any verification examples using the diagnostics package, or
>>> does anyone have a copy of a similar data.diagnostics that I can
>>> modify.
>>>
>>> Thank you,
>>>
>>> Jeff Polton
>>>
>> _______________________________________________
>> 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
>