Problem outputting daily ocean data

[a_sengupta@unsw_edu_au]

Hi All,
Im trying to output daily ocean velocity data. I've been successful in producing daily output as defined in the pop2 namelist, but CCSM4 always crashes near the end of the first month. I have tried doing both branch and startup runs with the same result. I have also tried different output configurations using the pop2 namelist but to no avail.
Here's an example:
I create an extra output stream:
pop2.buildnml.csh:
&tavg_nml
n_tavg_streams = 4
ltavg_streams_index_present = .true.
tavg_freq_opt = 'nmonth' 'nday' 'once' 'nday'
tavg_freq = 1 1 1 1
tavg_file_freq_opt = 'nmonth' 'nmonth' 'once' 'nday'
tavg_file_freq = 1 1 1 1
tavg_stream_filestrings = 'nmonth1' 'nday1' 'once' 'nday2'
tavg_start_opt = 'nstep' 'nstep' 'nstep' 'nstep'
tavg_start = 0 0 0 0
tavg_fmt_in = 'nc' 'nc' 'nc' 'nc'
tavg_fmt_out = 'nc' 'nc' 'nc' 'nc'
tavg_contents ='$tavg_contents_filename'
ltavg_nino_diags_requested = .true.
tavg_infile ='${output_h}restart.end'
tavg_outfile ='$output_h'
ltavg_has_offset_date = .false. .false. .false. .false.
tavg_offset_years = 1 1 1 1
tavg_offset_months = 1 1 1 1
tavg_offset_days = 2 2 2 2
ltavg_one_time_header = .false. .false. .false. .false.

and in gx1v6_tavg_contents I set
4 UVEL
4 VVEL
4 WVEL

I run for a month and correctly get the daily output files:
CASE.pop.h.nday2.0001-01-02.nc to CASE.pop.h.nday2.0001-01-30.nc


The first error Im getting in the ccsm log is:
POP aborting...
(write_nstd_netcdf) Attempt to write undefined field in netCDF write

It looks like the problem occurs near the end of the month when trying to write CASE.pop.h.0001-01.nc. While the file is created it has a time size of 0.

Any help would be much appreciated. Im happy with any output configuration as long as it produces daily UVEL, VVEL and WVEL
Many thanks
Alex

 

[bates]

Hi Alex,

There are certain variables (like MOC, N_HEAT, N_SALT) that are calculated by the model in tavg.F90. These are calculated from other model variables. All variables used to calculate each of these needs to be in the same stream. To get MOC, you potentially need all of the following:

WVEL
VVEL
WISOP
VISOP
WSUBM
VSUBM

When you moved WVEL and VVEL to another stream, MOC could not be calculated. You cannot have one variable in more than one stream. You have 2 choices:

1. Move all of the variables above into your new stream. This means that you'll be calculating MOC every day (which may take more time), and writing all of these variables to a daily stream will certainly take more space.

2. If you don't care whether you have MOC output or not, you can turn off that calculation by setting 'moc_requested = .false.' in the transports namelist (transports_nml). Depending on the version you are using, this could be located in different places, but most likely, you have a file called 'pop2.buildnml.csh' in your /BuildConf directory. This is where you'll find the namelists for POP2.

There are other such calculations that might depend on the velocity components. Start with this one, and if you encounter a problem again, write back and we'll work through the next one.


Susan
Ocean Model Science Liaison

 

[jhe@rsmas_miami_edu]

Hi,
I guess I might have similar problems
I'm trying to add daily snapshot SST and IFRAC field to my output. But the model doesn't seem to respond to my changes.
Here is how I did it.
In 'pop2.buildnml.csh' file, I set

&history_nml
history_freq_opt = 'nday'
history_freq = 1
history_outfile = '${output_h}s'
history_fmt = 'nc'
history_contents = '$history_contents_filename'
/

My 'gx1v6_history_contents' file is like this,

7
SHGT
UBTROP
VBTROP
UVEL
VVEL
TEMP
SALT
SST
SST2
IFRAC

The model runs smoothly and gives me snapshot output, but it doesn't contain SST, SST2 and IFRAC. I guess I missed something when I set up the model. Can anyone please help? Thank you very much!
Jie

 

[bates]

Hi Jie,

The 7 at the top of your list below in gx1v6_history_contents is a counter for the number of variables that follow. Since you added 3 variables, this 7 should be changed to a 10.

Susan

 

[asemenov@alaska_edu]

Hi all,I tend to get colder ocean surface waters in the North Atlantic and as a result there is more sea ice in the Arctic than it is in the remote sensing data. I know there is no way to fix it easily, but I was hoping that you could provide a hint where to look in the codes to add or change coefficients related to heating of the North Atlantic water. Thanks a lot,
Jason