[MITgcm-support] diagnostics parameter

[Martin Losch]

Hi Jody,

sorry for scewing up my question (as usual). I meant 134days (or 9648 timesteps) since the pickup at 131544*1200 = 1827 days, so 1961-1827=134.

If you have a look at the old message (follow the link), the timePhase is much larger than frequency in his example and you can indeed make the model start producing output later during the run (I have done it following Jean-Michel's example). The question remains, why doesn't it work now? And why does the diag_stats stuff behave even differently (it should do the same).

Martin

On Sep 10, 2013, at 6:46 PM, Jody Klymak wrote:

> Hi Martin,
> 
> Before Jean-Michel explains it properly:
> 
> On Sep 10, 2013, at  9:28 AM, Martin Losch <Martin.Losch at awi.de> wrote:
> 
>> Hi there,
>> 
>> this is a problem that I keep running into, and since the search function on mitgcm.org does not help very much (as usual), I have to ask this question again:
>> 
>> For debugging reasons, I want the model to start from a pickup at nIter0 = 131544. My timestep is deltaT = 1200. I want the model to write diagnostics at each timestep after timestep 141192, i.e. after the model has reached the time 141192*1200sec = 169430400sec=134days. I thought that I could do it like this in data.diagnostics (<http://mitgcm.org/pipermail/mitgcm-support/2010-August/006799.html>):
>> 
>>  filename(1) = 'myfile',
>>  frequency(1) = 1.,
>>  timePhase(1) = 169430400.,
>>  fields(1,1) = 'ETAN    ', ….
> 
> I think you want frequency(1)=1200.  
> I'm pretty sure that timePhase should be less than 1200. i.e. it says where in the 1200s the averaging occurs.  
> 
> C     frequency :: Frequency (in s) of Output (ouput every "frequency" second)
> C     timePhase :: phase (in s) within the "frequency" period to write output
> 
> I am very confused by your math.  141192*1200sec = 169430400sec=1961 days, not 134, and none of your numbers below make sense ;-)
> 
> I don't know that there is a way to start writing diagnostics after a certain time w/o restarting to model.
> 
> Cheers,   Jody
> 
> 
>> because it says in the standard data.diagnostics files in the verification experiments:
>> #  timePhase(n)     : write at time = timePhase + multiple of |frequency|
>> 
>> Obviously I got it wrong because now the model writes output at timesteps 132408 (after 12 days), 134568 (42days), 136800 (73days), etc.
>> Where did I go wrong, and how can I do it correctly?
>> 
>> Analogously in the diag_stats name list, I specified stats_freq=1 and stats_phase=169430400 and got output every time step from the very beginning (so already at time step 131544). Why are the two parameter sets behaving so differently?
>> 
>> Martin
>> 
>> 
>> 
>> 
>> 
>> _______________________________________________
>> MITgcm-support mailing list
>> MITgcm-support at mitgcm.org
>> http://mitgcm.org/mailman/listinfo/mitgcm-support
> 
> --
> Jody Klymak    
> http://web.uvic.ca/~jklymak/
> 
> 
> 
> 
> 
> _______________________________________________
> MITgcm-support mailing list
> MITgcm-support at mitgcm.org
> http://mitgcm.org/mailman/listinfo/mitgcm-support