Scheduled Downtime
On Tuesday 24 October 2023 @ 5pm MT the forums will be in read only mode in preparation for the downtime. On Wednesday 25 October 2023 @ 5am MT, this website will be down for maintenance and expected to return online later in the morning.
Normal Operations
The forums are back online with normal operations. If you notice any issues or errors related to the forums, please reach out to help@ucar.edu

Specify CAM initial files in a multi-instance CESM run

Hi there, I want to run multi-instance CESM (for the test run, 2 instances) with different atmosphere initial conditions. So I first perform a hybrid run to obtain CAM initial files at different times (exp1). Then I carry out another hybrid run with 2 instances (exp2) based on the CAM initial and restart files of exp1.For exp2, the related variables in env_run.xml are set as below:
  • RUN_TYPE = hybrid
  • RUN_REFCASE = exp1
  • RUN_REFDATE = 0001-01-03 (exp1 was run for 2 days from 0001-01-01)
  • RUN_REFTOD = 00000
  • GET_REFCASE = FALSE
  • CONTINUE_RUN = FALSE
After ./build, I copy all the files in the exp1 restart directory to the $RUNDIR of exp2, and I am sure that two different CAM initial files (exp1.cam_0001.i.0001-01-03-00000.nc & exp1.cam_0002.i.0001-01-03-00000.nc) with required names are provided. But after I submit exp2, I get many lines of the below error in cesm.log:
  • ERROR: GETFIL: FAILED to get e1.cam.i.0001-01-03-00000.nc
But in cam.buildnml.csh, I notice that below lines are included:
  • if ($RUN_TYPE == 'hybrid') then
  •   if (-e ${RUN_REFCASE}.cam${inst_string}.i.${RUN_REFDATE}-${RUN_REFTOD}.nc) then
  •     set ncdata = "ncdata='${RUN_REFCASE}.cam${inst_string}.i.${RUN_REFDATE}-${RUN_REFTOD}.nc'"
  •   else
  •     set ncdata = "ncdata='${RUN_REFCASE}.cam.i.${RUN_REFDATE}-${RUN_REFTOD}.nc'"
  •     if ($inst_string != "") echo "WARNING: $ncdata is being used"
  •   endif
  • endif
I do not understand why the model cannot find the 2 initial files I provided, and still try to find the default initial file, which leads to the above error. After I specify the 2 provided initial files in 2 user_nl_000* files with ncdata, the model no longer get the above error, but I get many lines of below error in the cesm.log:
  • ERROR: aborting in ice-pio_ropen with invalid file
Besides, if I only run 1 instance the same way as above to specify the CAM initial file from a previous run, the model can complete successfully. Can you help me with the above 2 questions which only found in the multi-instance scenario? (1. why the model cannot recognize the 2 CAM initial files I provided before specified with ncdata, 2. why I get the ice-pio_ropen error which not happened in the single instance run) Your help will be much appreciated. Thanks!
 

xyli@geo_uio_no

New Member
Thanks a lot!
Using CESM2.1.1, copying all the restart files to
inputdata/cesm2_init/${REFCASE}/0024-01-01
works. Of course, one needs to make sure that
the ice files are indeed ouputted.

But I got the following error? Did you see this?
forrtl: severe (24): end-of-file during read, unit 97, file /cluster/work/users/xiangyuli/cesm/BW-control-190626-hybrid/run/./BW-control-190626.pop.ro.0025-10-01-00000
 
Top