[MITgcm-devel] RE: mon_solution & eedie

[Alistair Adcroft]

A) Change the name if it makes you happy.

B) Make it more verbose (mention temperature) if you have the time.

C+D) I understood the EEDIE to be the "clean" way to stop the model. The
difficulty in
     removing it is that the ocean only model won't die cleanly - so leave
it in.
     I agree that not calling it when coupled is fine but how does that help
- the model
     will just sit there and spin with out ending.

A.
--
Dr Alistair Adcroft            http://www.mit.edu/~adcroft
MIT Climate Modeling Initiative        tel: (617) 253-5938
EAPS 54-1523,  77 Massachusetts Ave,  Cambridge,  MA,  USA

-----Original Message-----
From: Jean-Michel Campin [mailto:jmc at ocean.mit.edu] 
Sent: Saturday, December 13, 2003 6:31 PM
To: adcroft at mit.edu
Cc: MITgcm-devel at mitgcm.org
Subject: mon_solution & eedie


Hi Alistair,

I don't like too much this "mon_solution.F" :
a) the name does not tell that this is "checking"
 the solution. but this is more a matter of personal taste.
b) the error message that is printed:
> MON_SOLUTION: STOPPED DUE TO EXTREME VALUES OF SOLUTION
 does not tell that the temperature is wrong.
c) the eedie generate a strange message (Stephanie was 
 wandering what was going on the first time it happened).
d) the eedie prevent the coupled-model to stop (all the processor  are
waiting for something that does not come).

I fixed the point (d) simply not calling eedie when using the coupler, 
but I wander if we cannot simply remove the "call eedie" from 
mon_solution.F, if it's too difficult to print a different 
error message.

See you,

Jean-Michel

PS: Chris, I commented out the MPI_BARRIER in eedie.F but this does not
change anything. When the temperature is out of range, the eedie (called
from mon_solution) continue to wait.