issues with A_B1850BGC ?

[worleyph]

Should perhaps be two different issues, but maybe you have to use a spun-up initial condition to get a good run? Or is the second error the same as issue #571?

A) Just tried to run

 -res ne30_m120 -compset A_B1850BGC

on Titan and it died with

...

  Reading initial conditions from
  /lustre/atlas1/cli900/world-shared/cesm/inputdata/lnd/clm2/initdata_map/clmi.I1850CRUCLM45BGC.0241-01-01.ne30np4_g1v6_simyr1850_c140111.nc
  (GETFIL): attempting to find local file
  clmi.I1850CRUCLM45BGC.0241-01-01.ne30np4_g1v6_simyr1850_c140111.nc
  (GETFIL): using
  /lustre/atlas1/cli900/world-shared/cesm/inputdata/lnd/clm2/initdata_map/clmi.I1850CRUCLM45BGC.0241-01-01.ne30np4_g1v6_simyr1850_c140111.nc
  Reading restart dataset
  check_dim ERROR: mismatch of input dimension         15865
   with expected value         16335  for variable gridcell
  ERROR: ERROR in ncdio_pio.F90.in at line 353

B) I then reran with finidat = ' ' in user_nl_clm, and it died with (from lnd.log.XXX)

 ...
  dtime_sync=          1800  dtime_clm=          1800  mod =             0
  --WARNING-- skipping CN balance check for first timestep
  --WARNING-- skipping CN balance check for first timestep
  clm: completed timestep             0
  --WARNING-- skipping CN balance check for first timestep
  --WARNING-- skipping CN balance check for first timestep
  clm: completed timestep             1
  clm: completed timestep             2
  column nbalance error =   -1.7761375917134824E-007          755            3
  ...
  ENDRUN:
  ERROR in /autofs/nccs-svm1_home1/worley/ACPI/SVN/ACME/worleyph/V1_instr/ACME_instr/components/clm/src/biogeochem/CNBalanceCheckMod.F90 at line 396  
  ERROR: Unknown error submitted to shr_sys_abort.

[bishtgautam]

1. The incorrect dimension error (check_dim ERROR: mismatch of input dimension) is because of the use of an incorrect finidat file. This issue is same as Runtime error in ACME B-case with CLM45%SP #507.
2. I believe the column nbalance error is same as ICLM45BGC case fails on blues #316.
   • @xyk-ornl: I was under the impression PR-Making P code compatible with Century decomposition submodel  #371 should have fixed this error. Any ideas why one is getting the above error?
   • @qzhu-lbl : Do you expect that this error would be resolved by the PR that fixes issue-Incorrect N/P uptake and phosphatase activity #571?
3. @worleyph: In case you are looking for a compset to study the performance of ACME for BGC v1 experiments, I believe the BGC compset is not the correct compset to use as BGC v1 experiments will use CN compset.

[worleyph]

@bishtgautam , Thanks. I assumed that since @rljacob had taken the time to generate a compset with BGC, that this was relevant to V1. (See https://acme-climate.atlassian.net/wiki/x/V4FjAg.) I knew about the need to use "finidat = ' ' " for A_B1850, but did not realize that the same applied to A_B1850BGC.

My intent is to evaluate the "timer coverage", i.e., whether the existing timers are providing the necessary performance data, and to add/move them if not. If the BGC case will not be used in production over the next year, I'll ignore it for now.

[bishtgautam]

At this stage, I believe there is no initial land dataset at ne30_m120 resolution for any compset (SP/CN/BGC).

The A_B1850CN compset will be used for BGC v1 experiments, so give it a try to evaluate the timer coverage.

[worleyph]

Thanks. Please close this issue if it is covered by the others.

[rljacob]

I misunderstood and thought CLM45%BGC was going to be used so added it to ACME B cases.