How to NOT output the tavg files?

[whannah]

I don't need any of the time averaged output from POP2, so I'm trying to turn them off. It seems like it should be as simple as setting: tavg_file_freq_opt = 'never','never','never'But this doesnt work! I still get the tavg files for the default settings of 'nmonth', 'nday1', and 'once' I've tried a few other random things, and I've searched through the documentation. Does anyone know how to do this? thanks,Walter 

 

[jedwards]

It's unfortunately a bit complicated:     Add the following to Sourcemods/src.pop2/ocn.base.tavg.csh (and make sure the file is executable)#!/bin/csh -fcat >! $CASEBUILD/pop2conf/base.tavg.nml

 

[whannah]

jedwards, I tried both of your suggestions,but neither is working.

The first method gives >  "ERROR: BSF must be requested in tavg_contents file"
I tried setting "ldiag_bsf" but both true and false settings gave more errors. I didnt note what they were. I can try again and write them down.

The second method gave an error about not finding the "rof_in" namelist, which I found very confusing.

Any other ideas?

 

[jedwards]

What version of CESM are you using and what compset are you trying to build?

 

[whannah]

I'm using version 1.1.1 because I'm doing experiments with the super-parameterization (SP-CAM). Also for the first method, if I use ldiag_bsf = .false. I get:FATAL ERROR  in tavg_mocBut if I use ldiag_bsf = .true. I get:ERROR: BSF must be requested in tavg_contents file  

 

[mlevy]

As Nancy mentioned, you need to turn off several diagnostic options if you are not writing history files. I believe this is a full list of additions you need in user_nl_pop2:
ldiag_bsf = .false.
 ldiag_global_tracer_budgets = .false.
ldiag_velocity = .false.
diag_gm_bolus = .false.
ltavg_nino_diags_requested = .false.
moc_requested = .false.
n_heat_trans_requested = .false.
n_salt_trans_requested = .false.
diag_cfl_freq_opt = 'never'
diag_global_freq_opt = 'never'
diag_transp_freq_opt = 'never'

 

[njn01]

 I just noticed a problem with your original attempt to disable the pop2 tavg files.  You were on the right track, but you modified the file frequency, not the field frequency.  That is, you settavg_file_freq_opt = 'never','never','never'but you should have settavg_freq_opt  = 'never','never','never' This is certainly an easy mistake to make -- just about as easy as it was for me to miss your mistake in the first place...  

 

[njn01]

If you totally eliminate the ocn tavg files, you will also need to turn off pop2 diagnostics.  You can't set ldaig_bsf=.false. without also setting moc_requested = .false., for example.  There are others like this, but I don't have a comprehensive list. (ldiag_ variables are obvious candidates).I would have thought that setting  tavg_freq_opt = 'never' 'never' never'would have worked to shut down the tavg output without problems, but if it didn't, then the following might:tavg_freq = 9999 9999 1tavg_freq_opt = 'nyear' 'nyear' 'once'

 

[whannah]

Thanks Nancy. Your solution is pretty clear. I figured the monthly files might be nice to have afterall, so I decided to use 

tavg_freq_opt = 'nmonth' 'never' 'never'

I ran about 60 days, and looked at what files made it through the short term archive:ls -l scratch/archive/CESM_ZM_CTL-CPL_0.9x1.25_gx1v6_00/ocn/hist/-rw-r--r-- 1 whannah ncar    3 Jun 24 12:47 CESM_ZM_CTL-CPL_0.9x1.25_gx1v6_00.pop.dd.1900-01-02-00000
-rw-r--r-- 1 whannah ncar 3.1K Jun 24 13:20 CESM_ZM_CTL-CPL_0.9x1.25_gx1v6_00.pop.dd.1900-01-16-00000
-rw-r--r-- 1 whannah ncar 3.1K Jun 24 13:56 CESM_ZM_CTL-CPL_0.9x1.25_gx1v6_00.pop.dd.1900-02-05-00000
-rw-r--r-- 1 whannah ncar 9.6K Jun 24 12:52 CESM_ZM_CTL-CPL_0.9x1.25_gx1v6_00.pop.do.1900-01-02-00000
-rw-r--r-- 1 whannah ncar  14K Jun 24 13:22 CESM_ZM_CTL-CPL_0.9x1.25_gx1v6_00.pop.do.1900-01-16-00000
-rw-r--r-- 1 whannah ncar  28K Jun 24 14:01 CESM_ZM_CTL-CPL_0.9x1.25_gx1v6_00.pop.do.1900-02-05-00000
-rw-r--r-- 1 whannah ncar    3 Jun 24 12:47 CESM_ZM_CTL-CPL_0.9x1.25_gx1v6_00.pop.dt.1900-01-02-00000
-rw-r--r-- 1 whannah ncar  856 Jun 24 13:20 CESM_ZM_CTL-CPL_0.9x1.25_gx1v6_00.pop.dt.1900-01-16-00000
-rw-r--r-- 1 whannah ncar  856 Jun 24 13:56 CESM_ZM_CTL-CPL_0.9x1.25_gx1v6_00.pop.dt.1900-02-05-00000
-rw-r--r-- 1 whannah ncar 4.1K Jun 24 12:52 CESM_ZM_CTL-CPL_0.9x1.25_gx1v6_00.pop.dv.1900-01-02-00000
-rw-r--r-- 1 whannah ncar 5.8K Jun 24 13:22 CESM_ZM_CTL-CPL_0.9x1.25_gx1v6_00.pop.dv.1900-01-16-00000
-rw-r--r-- 1 whannah ncar  12K Jun 24 14:01 CESM_ZM_CTL-CPL_0.9x1.25_gx1v6_00.pop.dv.1900-02-05-00000
-rw-r--r-- 1 whannah ncar 1.1G Jun 24 13:20 CESM_ZM_CTL-CPL_0.9x1.25_gx1v6_00.pop.h.1900-01.nc
-rw-r--r-- 1 whannah ncar 1.1G Jun 24 13:56 CESM_ZM_CTL-CPL_0.9x1.25_gx1v6_00.pop.h.1900-02.nc
-rw-r--r-- 1 whannah ncar 114M Jun 24 13:20 CESM_ZM_CTL-CPL_0.9x1.25_gx1v6_00.pop.hs.CESM_ZM_CTL-CPL_0.9x1.25_gx1v6_00.1900-02-01.nc
-rw-r--r-- 1 whannah ncar 114M Jun 24 13:56 CESM_ZM_CTL-CPL_0.9x1.25_gx1v6_00.pop.hs.CESM_ZM_CTL-CPL_0.9x1.25_gx1v6_00.1900-03-01.ncThe TAVG files don't seem to be there, but those monthly *.h.* files are ~1GB! I don't understand why those have to be so big by default.I'm digusted with this whole issue so I'm gonna keep running these experiments and deal with the excess output later. Thanks again for all the help though. 

 

[njn01]

The *pop.h.* files are the "TAVG files".  You have two of them in your example, and yes, they are large, because they contain information that scientists find useful when analyzing the ocean model results.  Note that the long-term archiver compresses these files to a more manageable size before they are stored long term.

 

[whannah]

ok, I understand. Thanks.