Some auxiliary output files have double the time steps

[gabriel_bromley]

Hello,

I am running WRF and saving some variables to a file at hourly intervals. Some of the netcdf files have 59 time steps instead of 24. I am assuming it's some sort of corrupted file problem. Has this problem come up before? The files in question are at:

/glade/p/univ/umsb0001/model_runs/control_umd/sfc

and this is one example of a file that is a problem:

sfc_vars_d01_2011-08-13_00\:00\:00.nc

While this one is correct:

sfc_vars_d01_2011-08-12_00\:00\:00.nc

Thanks,

Gabe

 

[kwerner]

Gabe,
I've not seen this before, but when I look at the times listed in the 'sfc_vars_d01_2011-08-13_00:00:00.nc' file, I see these weird empty quotes at the end.

Times =
  "2011-08-13_00:00:00",
  "2011-08-13_01:00:00",
  "2011-08-13_02:00:00",
  "2011-08-13_03:00:00",
  "2011-08-13_04:00:00",
  "2011-08-13_05:00:00",
  "2011-08-13_06:00:00",
  "2011-08-13_07:00:00",
  "2011-08-13_08:00:00",
  "2011-08-13_09:00:00",
  "2011-08-13_10:00:00",
  "2011-08-13_11:00:00",
  "2011-08-13_12:00:00",
  "2011-08-13_13:00:00",
  "2011-08-13_14:00:00",
  "2011-08-13_15:00:00",
  "2011-08-13_16:00:00",
  "2011-08-13_17:00:00",
  "2011-08-13_18:00:00",
  "2011-08-13_19:00:00",
  "2011-08-13_20:00:00",
  "2011-08-13_21:00:00",
  "2011-08-13_22:00:00",
  "2011-08-13_23:00:00",
  "",
  "",
  "",
  "",
  "",
  "",
  "",

(plus more - I didn't include them all)
Can you let me know which version of WRF you're using, whether it's been modified at all, and can you point me to the directory where WRF is built and you're running the case? I'd like to take a look at the namelist, output files, etc. Thanks!

 

[gabriel_bromley]

Hello Kelley,

My run directory is here: /glade/scratch/gbromley/control_umd/wrf/run

I copied the WRF files from the WRFHELP precompiled code directories. It's WRFv4.2. There are no special options besides some extra aux output.

I discovered that if I just cut out the first 24 time steps, everything seems fine, which is awesome since that means the data is probably fine overall. I can probably write a script to do this for all files. If I can stop this from happening with future files though would be great (it's still happening with the most recent aux out files).

Thanks,

Gabe

 

[kwerner]

Gabe,
I'm glad you've figured out a workaround for now. I apologize for taking so long to get back to you, but I've been trying to find ways to reproduce it and it's been unsuccessful. My latest attempt used the following:
wrfv4.2
your namelist (with dates modified)
your wrfrst_d01_2011-10-20_00:00:00 file
I used your met_em* files to create a wrfbdy_d01 file for that time period.
I ran a restart through the 23rd, since it looks like your file "sfc_vars_d01_2011-10-22_01:00:00.nc" has the issue.
But my times in each sfc* file look okay.

Is there anything you can think of that I would need to do differently? If not, then perhaps next time you come across the problem, save the following files so that I can try to reproduce with your exact files:
wrfbdy_d01 for that time period
wrfrst* for that time period
namelist.input for that time period
rsl.* files for your simulation
sfc* files for your simulation

Thanks!