[MITgcm-support] obcs problems

[Martin Losch]

Hi Aviv,

the error message means, that the model does find the file, but that it tries to read a record that's not there (basically the file is "too short"). There are several potential reasons for this:
1. you created a real*4 file, but tell the model hat you are reading real*8 files, then the files is "too short" by a factor of two, and everything you read is garbage
2. you have specified time varying boundary values and you simply do not have enough records (time slices) in your file.
(3. there is bug, not unlikely in this package)

The format for a east/west boundary is ny*nz*nt and for an north/south boundary nx*nz*nt, where nt is the number of time records required. The 2 in exp4 is just used for testing not really necessary for the example. The model reads only the first record, because we tell it that the forcing is constant in time (periodicExternalForcing=.false.).

Please send your "data", "data.obcs" (and "data.exf" if you are using the exf package), if you need more help.

Martin

----- Original Message -----
From: aviv solodoch <avivsolo at gmail.com>
Date: Friday, February 27, 2009 22:10
Subject: [MITgcm-support] obcs problems
To: mitgcm-support at mitgcm.org

> Hi all

> I have some problems using the open boundary conditions package.
> The error I get when running is:
> forrtl: severe (36): attempt to access non-existent record, unit 9, file /run/OBmerid_Zero.bin

> The mentioned file OBmerid_Zero.bin, does exist in the specified directory.
> There could be some problem with its definition.
> I created this file as input for the following fields specified in data.obcs:
> OBNuFile='OBmerid_Zero.bin',
> OBSuFile='OBmerid_Zero.bin',
> OBNvFile='OBmerid_Zero.bin',
> OBNuFile='OBmerid_Zero.bin',

> What would be the correct format for such a file?
> In the example exp4 such files had size Nx*Nr*2 to my best understanding. The 2 factor I can't understand, and also it doesn't seem to change the error I get.
> Any other ideas as to things that could have went wrong?
 
> Thanks
> Aviv
> _______________________________________________
> MITgcm-support mailing list
> MITgcm-support at mitgcm.org
> http://mitgcm.org/mailman/listinfo/mitgcm-support

Martin Losch
Alfred Wegener Institute 
Postfach 120161, 27515 Bremerhaven, Germany; 
Tel./Fax: ++49(0471)4831-1872/1797




-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mitgcm.org/pipermail/mitgcm-support/attachments/20090227/0d7d2347/attachment.htm>