using vertical nest in ideal case "em_les"

[leelaycn]

I have tried the ideal case "em_les" under vertical nest condition. The preprocesss "ideal.exe" runs "successfully" with nothing about nest domian 02.
According to the vertical nest paper " A New Vertical Grid Nesting Capability in the Weather Research and Forecasting (WRF) Model", I choose the easiest way, that means " vert_refine_method = 1" while " vert_refine_fact" and "e_vert" are calculated by the instruction in the paper. It tunrs out that no nest domain info is generated by the "ideal.exe".
My question: What should I do to run the vertical nest case in an ideal case by setting "vert_refine_method=1"?

WRF edition: v4.4.1

 

[kwerner]

Hi,
I apologize for the delay in response. Thank you for being patient with us.
In looking at the code, it doesn't look like the ideal program is set up to use this option. If you take a look in dyn_em/module_initialize_real.F, you can see the code that checks for a "vert_refine_method" in the namelist. Unfortunately the only way you could get this into the ideal case is to try to add it into the dyn_em/module_initialize_ideal.F file. There may be other files where code should be modified, as well, so you may have to search around a bit and run some trials after modifying. If you decide to do this, you'll need to recompile the code. If you don't modify any registry files, it should be okay to just recompile (without issuing 'clean -a' or reconfiguring). If you find a successful method, feel free to share it here so other users may be able to use it in the future!

 

[leelaycn]

Hi,
I apologize for the delay in response. Thank you for being patient with us.
In looking at the code, it doesn't look like the ideal program is set up to use this option. If you take a look in dyn_em/module_initialize_real.F, you can see the code that checks for a "vert_refine_method" in the namelist. Unfortunately the only way you could get this into the ideal case is to try to add it into the dyn_em/module_initialize_ideal.F file. There may be other files where code should be modified, as well, so you may have to search around a bit and run some trials after modifying. If you decide to do this, you'll need to recompile the code. If you don't modify any registry files, it should be okay to just recompile (without issuing 'clean -a' or reconfiguring). If you find a successful method, feel free to share it here so other users may be able to use it in the future!

Thank you, kwerner.

I checked the file "module_initialize_ideal.F ". It is what you said that there is no vertical nest initializing in em_les.
Still, the main program "wrf.exe" runs successfully in the vertical nest case of em_les. It turns out that the vertical nest method used in the ideal case em_les is a two-way nest method and the ideal program does not need to generate the nest domain in advance (which usually generated by ideal.exe). Once the wrf.exe runs into the time when the nest domain is on, the wrf.exe will automatically interpolate the domain from the coarse domain.
I got the nest domain output and the result is vertically interpolated.

 

[kwerner]

That is great news, and thank you for updating this. I wasn't aware of that, so it's good to know for the future and hopefully will help other users!