[MITgcm-support] Sneaky NaNs

[Nicolas Wienders]

Friend mitgcmers,

We are running here a coupled configuration of MITgcm (with
cheapaml) using a 1/8º grid resolution, 25 layers, Kpp, etc...
Just for a bit of backgroud.

So far dt was set to 2400, and i have tried for the night to push it
up to 2600. Doing weekly dumps, i got overnight about 82 successful
weekly dumps, then about 50 weekly dumps of NaNs (which is not so
much a surprise considering that dt is probably to high), but.... then
it comes back to physical and meaningful values.


How can the model turn to a NaN status to come back to life afterwards?
Just wondering.

Thank you. Have a great week-end. Hot hot here.
n




--

Nicolas Wienders                   (850) 644-1987 (office)
Oceanography - OSB415                294-3542 (cell)
Florida State University                    644-2581 (fax)
Tallahassee, FL - 32306    wienders at ocean.fsu.edu


-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mitgcm.org/pipermail/mitgcm-support/attachments/20070323/65670d0e/attachment.htm>