Save history file only for specified domains

[Zetison]

Is it possible to save history files only for specified domain to reduce storage consumption? I guess one could use history_interval to have less frames for the other domains, but maybe there exists a more clean way to obtain a solution here?

 

[Deleted member 3607]

This thread may help you.

Setting a time to start writing history files from

Hi, Is it possible to set a starting time for writing history files? For example, for a total of 24-hour long simulation, I'm only interested in the hours between 18-22. My goal is to save disk space. Thanks in advance

forum.mmm.ucar.edu

Also these are the options for the namelist.input file for wrf and wps.

WRF/README.namelist at master · wrf-model/WRF

The official repository for the Weather Research and Forecasting (WRF) model - WRF/README.namelist at master · wrf-model/WRF

github.com

WPS/namelist.wps.all_options at master · wrf-model/WPS

The official repository for the WRF Preprocessing System (WPS) - WPS/namelist.wps.all_options at master · wrf-model/WPS

github.com

 

[Zetison]

Thanks for your answer. I had a look into that thread and it did not work out for me, it seems to me that these "inputout..." settings control the input data "wrfinput_d<domain>_<date>" instead of the history output data files. Also, when I try to add these options I get errors (the namelist is not well described on this imo).
------ ERROR while reading namelist time_control ------
Maybe here?: inputout_begin_y = 0, 0, 0, 0,
Maybe here?: inputout_begin_mo = 0, 0, 0, 0,

I'm running WRF Model Version 4.3

 

[kwerner]

@Zetison,
If you are only interested in writing history output for one of your domains, you can set the other to zero. For e.g.,
history_interval = 0, 60
This would not write any history files for d01, and would write them every hour for d02.

 

[Zetison]

Setting history_interval (or history_interval_s) equal to zero for a domain seems not to give any simulation at all. Attached is the output (no errors produced). Did I miss something here?

 

[kwerner]

Hi,
I believe this is because you're using adaptive time-stepping, and you have "step_to_output_time = .true." Can you set that to .false. and see if it fixes the issue?

 

[Zetison]

Yes, that gave at least a simulation, but the history is still written to file for the other domains (but only two frames for some reason) with history_interval=0.

 

[kwerner]

I'm not sure why it would be still writing to the wrfout* files for d01, d02, and d03. I just ran a test with a similar set-up and I'm only getting history files for the domain that isn't set to 0. Can you run the test again, and then issue this command from your running directory

ls -ls >& ls.txt

and then attach that ls.txt file, along with the full rsl.out.0000 file? Thanks!