[MITgcm-support] clever pickup too clever for me

[Matthew Mazloff]

Hi Patrick,

Actually it seems that if you provide no meta file at all, which is  
what I did, the model will try to read the pickup in a "new" way and  
crash.  I think you need the old meta file and also need
pickupStrictlyMatch=.FALSE.,

I think that with an an old meta file nbFields = 0 and is read in as  
how use to be:

           CALL READ_REC_3D_RL( fn, fp, Nr, uVel,  1, myIter,myThid )
           CALL READ_REC_3D_RL( fn, fp, Nr, guNm1, 2, myIter,myThid )
           CALL READ_REC_3D_RL( fn, fp, Nr, vVel,  3, myIter,myThid )
           CALL READ_REC_3D_RL( fn, fp, Nr, gvNm1, 4, myIter,myThid )
           CALL READ_REC_3D_RL( fn, fp, Nr, theta, 5, myIter,myThid )
           CALL READ_REC_3D_RL( fn, fp, Nr, gtNm1, 6, myIter,myThid )
           CALL READ_REC_3D_RL( fn, fp, Nr, salt,  7, myIter,myThid )
           CALL READ_REC_3D_RL( fn, fp, Nr, gsNm1, 8, myIter,myThid )
           CALL READ_REC_3D_RL( fn,fp,1, etaN, nj*Nr+1, myIter,myThid )

if no meta file is presented, as I did, nbfields = -1, and reads in  
pickup file in the new way (which goes uVel, vVel, theta, salt,  
guNm1, gvNm1, gtNm1, gsNm1, etaN), which isn't consistent, and this  
causes the model to crash.

So I'll make up some meta files and throw them in there...(or I could  
hardcode nbFields to 0)....and let you know if it works

-Matt

ps> of course, i don't know what's gonna happen with seaice, should  
be fun :)








On Sep 3, 2008, at 9:56 PM, Patrick Heimbach wrote:

> Hi Matt,
>
> what has changed is that you now
> also provide the .meta file, and its content is checked for  
> consistency
> (might be important for seaice where it's not clear which options/ 
> grids etc you're using).
> Previously, the .meta files weren't evaluated.
> If you have an old pickup and didn't keep the .meta file,
> then just add in file data under PARM03 following option (as you did)
>   pickupStrictlyMatch=.FALSE.,
> In that case the model warns you that the pickup might not be  
> consistent,
> but continues running (assuming you know what you're doing).
> So ideally, keep the .meta files for your new runs and provide them
> when re-starting from pickup.
>
> -p.
>
>
>
> On Sep 3, 2008, at 11:47 PM, Matthew Mazloff wrote:
>
>> Hello JMC,
>>
>> I recently updated. It has been quite awhile since last update.     
>> When trying to run the model it immediately blew up.  I strongly  
>> suspect that it is the  "clever pickup" as the model blows up in 2  
>> time steps.  I think fields are being improperly initialized.
>>
>> I don't have the pickup meta files so I use
>>  pickupStrictlyMatch=.FALSE.,
>> and I get the following warnings
>> (PID.TID 0001.0001) WARNING >> MDS_READ_META: file: pickup. 
>> 0000000001.meta , pickup.0000000001.001.001.meta
>> (PID.TID 0001.0001) WARNING >> MDS_READ_META: Files DO not exist
>> (PID.TID 0001.0001) WARNING >> READ_PICKUP:  no field-list found
>> (PID.TID 0001.0001) WARNING >>  try to read pickup as currently  
>> written
>> (PID.TID 0001.0001) WARNING >> MDS_READ_META: file: pickup_seaice. 
>> 0000000001.meta , pickup_seaice.0000000001.001.001.meta
>> (PID.TID 0001.0001) WARNING >> MDS_READ_META: Files DO not exist
>> (PID.TID 0001.0001) WARNING >> SEAICE_READ_PICKUP:  no field-list  
>> found
>> (PID.TID 0001.0001) WARNING >>  try to read pickup as currently  
>> written
>>
>> So what has changed...I don't mind restructuring pickup.data or  
>> making a meta file...but can you please let me know the best way  
>> to do this.
>>
>> Or is it possibly not the clever pickup?
>>
>> -Thanks
>> Matt
>>
>>
>>
>> _______________________________________________
>> MITgcm-support mailing list
>> MITgcm-support at mitgcm.org
>> http://mitgcm.org/mailman/listinfo/mitgcm-support
>
> ---
> Patrick Heimbach | heimbach at mit.edu | http://www.mit.edu/~heimbach
> MIT | EAPS 54-1518 | 77 Massachusetts Ave | Cambridge MA 02139 USA
> FON +1-617-253-5259 | FAX +1-617-253-4464 | SKYPE patrick.heimbach
>
>
> _______________________________________________
> MITgcm-support mailing list
> MITgcm-support at mitgcm.org
> http://mitgcm.org/mailman/listinfo/mitgcm-support