History outputs

[TCNasa]

I'm having trouble getting the *.h history file outputs for my CAM case. I did a startup and tried a restart run, but I'm only getting restart outputs.
Any suggestions?

 

[mlevy]

What is the output of ./xmlquery -p DOUT_S? If DOUT_S=TRUE then the short-term archiver is enabled and all the history files will be moved to subdirectories of DOUT_S_ROOT. If that is not the case, let me know and I will move this to the CAM forum.

 

[TCNasa]

DOUT_S is TRUE, but I looked in my output/archive/$case/atm/hist and there are no files there.

 

[mlevy]

Can you run ./xmlchange DOUT_S=FALSE to turn off the archiver and then re-run? I wonder if you are having an issue with the short-term archiver (perhaps similar to Running CESM2.2 for a particular short run)

 

[TCNasa]

OK I will try that

 

[TCNasa]

While I wait on that, will a startup run generate history files?

 

[mlevy]

It should - how long are you running for? (what is the output of ./xmlquery -p STOP?) I think most components use the monthly stream for the bulk of the output, but also provide daily streams... so as long as you are running for more than a day you should get some history output.

 

[TCNasa]

So running for a few hours wouldn't produce any history? But 2 days for more would?

 

[mlevy]

Yes, I think you want to simulate at least a day to generate history output

 

[TCNasa]

Tried the DOUT_S = FALSE and a 2 day run. Still no *.h history outputs.

 

[mlevy]

From your case directory, could you run tail -n 15 CaseStatus and paste the results here? I'd also be interested to see the results of ./xmlquery -p STOP. The CaseStatus file contains information about the case -- every time you build the model or run the model, it is logged in that file (along with information about whether it completed successfully).

 

[TCNasa]

bn14:/CERES/sarb/caldwell/CESM2.2/cime/scripts/cam6-6hours {200} tail -n 15 CaseStatus


-, not checked out --> fi_20200609


2022-03-01 12:18:36: case.build success


---------------------------------------------------


2022-03-01 12:18:43: case.submit starting


---------------------------------------------------


2022-03-01 12:18:47: case.run starting


---------------------------------------------------


2022-03-01 12:18:51: model execution starting


---------------------------------------------------


2022-03-01 16:00:35: model execution success


---------------------------------------------------


2022-03-01 16:00:35: case.run success


---------------------------------------------------


2022-03-01 16:01:02: case.submit success


---------------------------------------------------

 

[TCNasa]

bn14:/CERES/sarb/caldwell/CESM2.2/cime/scripts/cam6-6hours {201} ./xmlquery -p STOP





Results in group run_begin_stop_restart


STOP_DATE: -999


STOP_N: 2


STOP_OPTION: ndays

 

[mlevy]

So it looks like the model ran successfully yesterday afternoon:

2022-03-01 12:18:47: case.run starting
---------------------------------------------------
2022-03-01 12:18:51: model execution starting
---------------------------------------------------
2022-03-01 16:00:35: model execution success
---------------------------------------------------
2022-03-01 16:00:35: case.run success

And is, indeed, running for two days. The next two things to look at are the output of ls in your run directory and the output of ./xmlquery -p DOUT in your case directory.

 

[TCNasa]

bn14:/CERES/sarb/caldwell/CESM2.2/cime/output/cam6-6hours/run {221} ls


timing cam6-6hours.mosart.r.2000-01-01-21600.nc ice.log.220301-113704.gz


atm_in cam6-6hours.mosart.r.2000-01-03-00000.nc ice.log.220301-121847.gz


atm.log.220301-111906 cam6-6hours.mosart.rh0.2000-01-01-21600.nc ice_modelio.nml


atm.log.220301-113704.gz cam6-6hours.mosart.rh0.2000-01-03-00000.nc lnd_in


atm.log.220301-121847.gz CASEROOT lnd.log.220301-113704.gz


atm_modelio.nml cesm.log.220301-111906 lnd.log.220301-121847.gz


cam6-6hours.cam.r.2000-01-01-21600.nc cesm.log.220301-113704.gz lnd_modelio.nml


cam6-6hours.cam.r.2000-01-03-00000.nc cesm.log.220301-121847.gz mosart_in


cam6-6hours.cam.rh0.2000-01-01-21600.nc cism.config ocn.log.220301-113704.gz


cam6-6hours.cam.rh0.2000-01-03-00000.nc cism_in ocn.log.220301-121847.gz


cam6-6hours.cam.rs.2000-01-01-21600.nc cpl.log.220301-111906 ocn_modelio.nml


cam6-6hours.cam.rs.2000-01-03-00000.nc cpl.log.220301-113704.gz rof.log.220301-113704.gz


cam6-6hours.cice.r.2000-01-01-21600.nc cpl.log.220301-121847.gz rof.log.220301-121847.gz


cam6-6hours.cice.r.2000-01-03-00000.nc cpl_modelio.nml rof_modelio.nml


cam6-6hours.cism.initial_hist.2000-01-01-00000.nc docn_in rpointer.atm


cam6-6hours.cism.r.2000-01-03-00000.nc docn.streams.txt.prescribed rpointer.drv


cam6-6hours.cism.tavg_helper.0000-00-00-00000.nc drv_flds_in rpointer.glc


cam6-6hours.clm2.r.2000-01-01-21600.nc drv_in rpointer.ice


cam6-6hours.clm2.r.2000-01-03-00000.nc esp_modelio.nml rpointer.lnd


cam6-6hours.clm2.rh0.2000-01-01-21600.nc finidat_interp_dest.nc rpointer.ocn


cam6-6hours.clm2.rh0.2000-01-03-00000.nc glc.log.220301-113704.gz rpointer.rof


cam6-6hours.cpl.r.2000-01-01-21600.nc glc.log.220301-121847.gz seq_maps.rc


cam6-6hours.cpl.r.2000-01-03-00000.nc glc_modelio.nml wav_modelio.nml


cam6-6hours.docn.rs1.2000-01-01-21600.bin iac_modelio.nml


cam6-6hours.docn.rs1.2000-01-03-00000.bin ice_in

 

[TCNasa]

bn14:/CERES/sarb/caldwell/CESM2.2/cime/scripts/cam6-6hours {223} ./xmlquery -p DOUT





Results in group run_data_archive


DOUT_S: FALSE


DOUT_S_SAVE_INTERIM_RESTART_FILES: FALSE





Results in group run_dout


DOUT_S_ROOT: /CERES/sarb/caldwell/CESM2.2/cime/output/archive/cam6-6hours

 

[mlevy]

Is it possible that you aren't producing any daily output? I thought CAM and CTSM both had daily output by default, but another option would be to run for a full month and see if the monthly output appears. The .rh files:

cam6-6hours.mosart.rh0.2000-01-03-00000.nc
cam6-6hours.cam.rh0.2000-01-03-00000.nc
cam6-6hours.clm2.rh0.2000-01-03-00000.nc

Are restart history files - they indicate that you have ended the run before those history files were due to be written.

What compset are you using, and have you made any modifications to this case or are you running out of the box?

 

[TCNasa]

I'm using F2000climo as a compost. start date of 2000-1-1 running for 2 days.

 

[TCNasa]

I mean compset. No mods just specifying the start date and run period

 

[mlevy]

I have verified on the NCAR system that F2000climo does not have any daily output by default. I ran for a month, though, and was able to produce history files from the atmosphere, land, sea ice, and river runoff models:

$ ls -1 *.h*.nc
F2000climo.forum_question.cam.h0.0001-01.nc
F2000climo.forum_question.cice.h.0001-01.nc
F2000climo.forum_question.clm2.h0.0001-01.nc
F2000climo.forum_question.mosart.h0.0001-01.nc

So it looks like you just need to run longer: ./xmlchange STOP_N=1,STOP_OPTION=nmonths should be sufficient to generate the above four files. Apologies for misleading you, I was under the impression that most of the CESM components produced at least some daily output.

It took me ~90 seconds to run the 2-day simulation, but most of that time must have been in initialization because it only took 9 minutes to run for a month