Model not complete error in the SCAM model run

[arti-14]

Hi,

I am simulating a case in the SCAM model.

I have set ./xmlchange DEBUG='TRUE'

The model runs for the time step given in the namelist, but then it stops with the following error:
"ERROR: Model did not complete - see /home/artij/OUTPUT/projects/scratch/SCAM_RUN_PARCEL_MODEL/run/cpl.log.211110-101135"

The cpl logfile ends with :
(prep_lnd_merge) x2l%Flgg_hflx00 = = g2x%Flgg_hflx00
(prep_lnd_merge) x2l%Flgg_hflx01 = = g2x%Flgg_hflx01
(prep_lnd_merge) x2l%Flgg_hflx02 = = g2x%Flgg_hflx02
(prep_lnd_merge) x2l%Flgg_hflx03 = = g2x%Flgg_hflx03
(prep_lnd_merge) x2l%Flgg_hflx04 = = g2x%Flgg_hflx04
(prep_lnd_merge) x2l%Flgg_hflx05 = = g2x%Flgg_hflx05
(prep_lnd_merge) x2l%Flgg_hflx06 = = g2x%Flgg_hflx06
(prep_lnd_merge) x2l%Flgg_hflx07 = = g2x%Flgg_hflx07
(prep_lnd_merge) x2l%Flgg_hflx08 = = g2x%Flgg_hflx08
(prep_lnd_merge) x2l%Flgg_hflx09 = = g2x%Flgg_hflx09
(prep_lnd_merge) x2l%Flgg_hflx10 = = g2x%Flgg_hflx10


And from the cpl.log file, I am unable to determine where the error is coming from.
Is there a way to know, what is causing this and determine the origin of the error ?

I have attached the log files generated during the run.

Thank you in advance for the help

Best Regards,
Arti Jadav

 

[jet]

Hi Arti:

It's hard to tell what the problem is from these log files. I haven't seen this error but if I were to hazard a guess I think its either a setup issue (wrong parameters in the land namelist) or you may have hit a bug in this version of the land. From the land log it looks like the land has finished 3 days and encounters an error when trying to close a file that it hasn't opened. Also from the land log it looks like the land will write a monthly history file but you are only running for 3 days. In that case, the land should only write out monthly restarts (no land history) at the end of day 3. If you have some special namelist logic to manage the writing of land history or restarts you should retry the run without those additional land namelist settings. If you are using the default land namelist created by the case then there is probably a bug somewhere in the code. You can make a test run for a month instead of 3 days and see if it finishes and everything is written out correctly.

Other than the small error at the end of the land log I don't see any errors in the other logs. My guess is that the land error is causing the model to crash after 3 days and the other logs just aren't flushed so they all have partial output.

jt