NetCDF: Attribute not found when use_init_interp=.true.

[amans]

What version of the code are you using?
alpha-ctsm5.2.mksrf.23_ctsm5.1.dev171-1-g95dfcff30

Have you made any changes to files in the source tree?
Some code changes for irrigation partitioning described in GitHub - amansnama/USGS_irrigation_clm: Implementation of USGS irrigation ratio in CLM5

Please attach any log files showing error messages or other useful information.
Attached two lnd.log files with use_init_interp true and false.

Describe your problem or question:

Case: /glade/work/amans/cases_0.05/006_IHistClm51BgcCrop_MRB_0.05_MOSART_0.125_USGS_irrig_ratio_rainfed_irrig_params_v9
This is a rerun of a previously running case: /glade/work/amans/cases_0.05/005_IHistClm51BgcCrop_MRB_0.05_MOSART_0.125_USGS_irrig_ratio_rainfed_irrig_params
The new case has changes in the paramfile.

Case run fails when use_init_interp=.true. with the lnd.log failing at:
Attempting to initialize the land model .....

Attempting to initialize run control settings .....
Read in clm_inparm namelist from: lnd_in
NetCDF: Attribute not found
ENDRUN:
ERROR: Unknown error submitted to shr_abort_abort.

When I retry with use_init_interp=.false. (as suggested by previous troubleshooting thread in the forum), the run fails later at

Reading restart dataset
check_dim_size ERROR: mismatch of input dimension 540172
with expected value 563040 for variable landunit
Did you mean to set use_init_interp = .true. in user_nl_clm?

It looks like my initial condition file needs use_init_interp=.true. The older case was running fine with the same initial file and use_init_interp=.true.
However, I am struggling to figure out what the actual issue is with the new case. I will be testing the case with the older paramfile. In the meantime, I would welcome any inputs to troubleshoot this issue.

 

[amans]

Update: Testing with the older working paramfile causes the same NetCDF: Attribute not found error.

 

[slevis]

You mentioned that an older case ran fine. Did the older case have the code changes mentioned in your post?
- If not, then the source of the problem is in your code changes.
- If yes, then I would go back and make sure that it still works. Then try to isolate what's different between the new case and the old case. I like doing that with the diff command on the whole case directories. Usually the most important differences will be in env_run.xml and user_nl_clm, but you may spot an important difference somewhere else.

 

[oleson]

Sometimes there can be a problem associated with an older interpolated file in your case run directory:

/glade/derecho/scratch/amans/006_IHistClm51BgcCrop_MRB_0.05_MOSART_0.125_USGS_irrig_ratio_rainfed_irrig_params_v9/run/init_generated_files/finidat_interp_dest.nc

In addition to @slevis suggestions, you could remove that directory and that file and try your use_init_interp=.true. run again

rm -rf init_generated_files/

 

[amans]

Thank you Sam and Keith! The issue was the older interp files in the case run dir. The new case is running fine as of now after removing the init_generated_files/.

I have made a separate thread for another issue bugging me from a long time.