Submit case failed in Derecho

[Shuai Li]

I am using the CESM2.1.4 version on Derecho, and the specific situation is:
I replaced the default fsurdat data with my own modified data and attempted to simulation. It went smoothly until the ./case.build, but failed in the final step ./case.submit.
The specific details of the case are: --res f09_f09_mg17 --compset F2000climo
I only modified the STOP_ N = 2 and STOP_ OPTION = nyears, as well as hist_nhtfrq = 0, -24 hist_mfilt = 1, 365 in user_nl_clm, and mfilt = 1, 365 nhtfrq = 0, -24 in user_nl_cam.

The received error is: run command is mpiexec --label -n 512 /glade/derecho/scratch/dyang/F2000climo1984test2/bld/cesm.exe >> cesm.log.$LID 2>&1
ERROR: RUN FAIL: Command 'mpiexec --label -n 512 /glade/derecho/scratch/dyang/F2000climo1984test2/bld/cesm.exe >> cesm.log.$LID 2>&1 ' failed
See log file for details: /glade/derecho/scratch/dyang/F2000climo1984test2/run/cesm.log.2682862.desched1.231221-231848
I have pasted the cesm.log file below.
I want to know how to solve this problem

 

[Shuai Li]

Because the original cesm.log file was too large, I copied some of the content.

 

[katec]

Hi there, so the error in this file is:
ERROR: \
ERROR in /glade/u/home/dyang/cesm2.1.4/components/clm/src/main/ncdio_pio.F90.in\
at line 368\
dec0266.hsn.de.hpc.ucar.edu 365: check_dim ERROR: mismatch of input dimension 50862 with expected value \
52315 for variable landunit\
Did you mean to set use_init_interp = .true. in user_nl_clm?\
(Setting use_init_interp = .true. is needed when doing a\
transient run using an initial conditions file from a non-transient run,\
or a non-transient run using an initial conditions file from a transient run,\
or when running a resolution or configuration that differs from the initial conditions.)\

It looks like your modified fsurfdata dimensions do not match what the model expects. I'm going to move this post into the CTSM forum as well, since this is a land specific issue.

 

[slevis]

@Shuai Li
the error message that the model generated, as posted above by @katec, tells you how you may correct the problem:

Did you mean to set use_init_interp = .true. in user_nl_clm?
(Setting use_init_interp = .true. is needed when doing a
transient run using an initial conditions file from a non-transient run,
or a non-transient run using an initial conditions file from a transient run,
or when running a resolution or configuration that differs from the initial conditions.)

 

[Shuai Li]

Hi there, so the error in this file is:
ERROR: \
ERROR in /glade/u/home/dyang/cesm2.1.4/components/clm/src/main/ncdio_pio.F90.in\
at line 368\
dec0266.hsn.de.hpc.ucar.edu 365: check_dim ERROR: mismatch of input dimension 50862 with expected value \
52315 for variable landunit\
Did you mean to set use_init_interp = .true. in user_nl_clm?\
(Setting use_init_interp = .true. is needed when doing a\
transient run using an initial conditions file from a non-transient run,\
or a non-transient run using an initial conditions file from a transient run,\
or when running a resolution or configuration that differs from the initial conditions.)\

It looks like your modified fsurfdata dimensions do not match what the model expects. I'm going to move this post into the CTSM forum as well, since this is a land specific issue.

Thanks for your help in this matter.
I found content consistent with your answer in the previous post.
After setting use_init_interp = .ture., the case can run smoothly.
But I still have some questions.
I changed some variables in fsurfdata, such as PCT_ GLACIER, PCT_ NAT_ PFT and other variables related to land cover and vegetation functional types.
I used NCL scripts to replace the variables mentioned above in my specific study domain.
But in previous cases, there was no issue of dimension mismatch under the same operation process, and there is no need to set use_ init_ interp=. ture. to ensure the successful operation of the case.
So I want to know if this parameter (use_init_interp = .ture.) will interpolate the initial condition (which seems to me to be f.e20.FHIST. f09_f09. esm2. 1.001_v2. clm2. r2000-01-01-00000. nc) to the same dimension as the surface data I replaced, or will it interpolate the surface data I replaced to the same dimension as the initial condition?
And whether it (use_init_interp = .ture.) will have a negative impact on the model results (like compared to if the dimension matches and there is no need to open this parameter).

use_init_interp = .true.

 

[Shuai Li]

@Shuai Li
the error message that the model generated, as posted above by @katec, tells you how you may correct the problem:

Did you mean to set use_init_interp = .true. in user_nl_clm?
(Setting use_init_interp = .true. is needed when doing a
transient run using an initial conditions file from a non-transient run,
or a non-transient run using an initial conditions file from a transient run,
or when running a resolution or configuration that differs from the initial conditions.)

Thank you, sir.
This is indeed a solution to the problem, but I still have some doubts about it (as I mentioned in my response to Katec).

 

[slevis]

So I want to know if this parameter (use_init_interp = .ture.) will interpolate the initial condition (which seems to me to be f.e20.FHIST. f09_f09. esm2. 1.001_v2. clm2. r2000-01-01-00000. nc) to the same dimension as the surface data I replaced, or will it interpolate the surface data I replaced to the same dimension as the initial condition?
And whether it (use_init_interp = .ture.) will have a negative impact on the model results (like compared to if the dimension matches and there is no need to open this parameter).

The model runs on the fsurdat dimensions and interpolates the finidat file to those dimensions. The greater the spatial differences between the fsurdat and the finidat files, the greater potential effect on your simulation. It is impossible for us to predict that. You can measure the effect by generating a new finidat file by spinning up the model with the new fsurdat file.

 

[Shuai Li]

Thanks for your timely response.
We will conduct the experiment as soon as possible and continue to provide feedback.