Anyone knows how to use wrf-les?

[junius.Wang]

Hello everyone, I want to know if you have a relatively complete WRF-LeS tutorial. Or is it better to apply to LES? I found a lot of documents, but there is no specific setting. I am a novice, so I need some hand-to-hand solutions. If it can have a namelist reference, it is even better. It needs your help very much!

 

[junius.Wang]

can any ones help me?

 

[Ming Chen]

Hi Junius,

I am sorry to tell that there is no tutorial specifically for WRF-LES.
For the default settings, you may refer to the namelist designed for ideal LES simulation.
However, real-data LES simulation is more complicated issue. You need to get information from the literature and correctly set the phys/dynamics options.
Unfortunately we also don't have sufficient experiences for LES simulations.

 

[Deleted member 3607]

Hi Junius,

I am sorry to tell that there is no tutorial specifically for WRF-LES.
For the default settings, you may refer to the namelist designed for ideal LES simulation.
However, real-data LES simulation is more complicated issue. You need to get information from the literature and correctly set the phys/dynamics options.
Unfortunately we also don't have sufficient experiences for LES simulations.

https://www2.mmm.ucar.edu/wrf/users/tutorial/presentation_pdfs/202101/dudhia_physics_pbl_turbulence.pdf

That might help

 

[junius.Wang]

https://www2.mmm.ucar.edu/wrf/users/tutorial/presentation_pdfs/202101/dudhia_physics_pbl_turbulence.pdf

That might help

Thanks for your help!

I did some tests myself, using three nested layers of pbl and one nested layer of les. And placed a wind farm inside the les nesting layer. But I don't know if the result is very strange because of the small locale (given my limited computing resources) and the short simulation time (24h).
1. The data visualization of u v is not real and not coherent, and the wind speed of u will not change after one degree.
2. Although there is a wind farm in my les layer, I can't find any wind farm information from the result file, which is very different from the result without les nesting layer. The power generation is also 0, but it can be clearly seen from the rsl_error that there are fans inserted into the grid.

I'll give the namelist.input file, and I'd be very grateful if you could provide any advice.

Thanks again for your reply and look forward to hearing from you again.

 

[junius.Wang]

 

[rdumais63]

Hi Junius. Have you been able to make any progress since April? I know it has been awhile since you posted. If you still need some ideas I would be able to share some of my experiences in getting up and going with WRF-LES. There are alot of things to consider but it is doable. You absolutely need a compute cluster to really make this work.

 

[rdumais63]

Also.. try looking at Jimy Dudhia's 2021 NCAR presentation titled "Overview of WRF Physics Boundary Layer and Turbulence". It is a great place to see how certain physics and treatment of turbulence & diffusion should be handled when jumping from mesoscale to LES nesting in WRF (especially in how to handle the turbulence grey area of NWP).

 

[junius.Wang]

Thanks so much and I do need some suggestions.
I use WRF-Les for atmospheric studies. At present, I hope to get atmospheric data with relatively high accuracy, but I know that this is difficult, and how to add disturbance effectively is a tricky issue.

What is your research on WRF-les? Do you have some suggestions to provide me? Any reminder will be grateful

 

[rdumais63]

Hi Junius. I have no experience running idealized simulations with WRF-LES or adding specific perturbations or disturbances. All of my work to this point has been in attempting to nest WRF from mesoscale to LES using one-way interactive nesting for real-case simulations. I also have the opton in my work to run FDDA but am not applying for LES at the moment. Typically I use either GFS 1/2 deg or NAM 218 fields (from NCEP) to initialize and supply time-dependent lateral boundary tendencies. Right now my main issues have been in terms of developing gradients of variables within the boundary relaxation zones of nests when applying vertical nesting. I have another recent forum item opened about this issue. I have attached a sample namelist of one configuration I am playing with now experimentally. Ignore some parameters in the namelist you may not recognize as related to, for example, radar assimilation as they are things added by our group for specific purposes and are not part of the standard WRF repository.