Topography experiment

[Aurora_LagosD]

Hello everyone,

I am carrying out an experiment with WRF to study the role of a mountain in South America on the precipitation of the surrounding region.. For this, I have run a control simulation with the mountain in the topography, and an experiment simulation where I have modified the topography to remove the mountain. The topography without the mountain was saved in binary files. Its corresponding index file was also created and added to the GEOGRID.TBL. The geogrid.exe completed successfully.

The simulation ran successfully, but I have a question: I understand that both simulations start with the same initial conditions, but the boundary conditions in the experiment without the mountain should differ since the atmosphere will be affected by the absence of the mountain. How can I address this issue?

The boundary conditions are imposed every 6 hours. I believe, although without any specific reference, that increasing this time interval might partially solve this "problem."

Aurora

 

[kwerner]

Hi Aurora,

The simulation ran successfully, but I have a question: I understand that both simulations start with the same initial conditions, but the boundary conditions in the experiment without the mountain should differ since the atmosphere will be affected by the absence of the mountain. How can I address this issue?

Can you be specific about what problem you're seeing? I agree with the statement you made, but it's unclear what is happening when you run wrf. Thanks!

 

[Aurora_LagosD]

Sorry for the poor explanation of my question. English is not my native language.

I have no problem with running WRF, rather it is a theoretical question.

I am doing a simulation with real topography and another with a fictitious topography without a mountain (which has a maximum height of 3000 meters above sea level). The simulation without the mountain used boundary conditions calculated by FNL, so even though the topography does not have the mountain, the boundary conditions are from an atmosphere affected by mountain.

My question is, how is this inconsistency usually addressed? In my opinion, both simulations should have the same initial conditions, but the boundary conditions should be different. I think that WRF in the experiment without the mountain, should develop its own atmosphere and somehow feed off boundary conditions it calculates by itself, not from FNL's, because in FNL the atmosphere is affected by the existence of the mountain. I don't know if there is a way to make WRF calculate its own boundary conditions. Another possibility is to use an interval_seconds of several days, so WRF creates its own atmosphere almost only from the initial conditions. Another option I am thinking of is to use a domain large enough so that these lateral boundary conditions are not affected by the mountain, but the lower and upper boundary conditions would be influenced by the mountain and I don't know how to approach that.

I hope you can now understand my question.

 

[kwerner]

Thank you so much for the explanation. Yes, now I understand what you are asking. Without having done this before, I *think* the best option is to make the domain large enough that the absence of the mountain is not an issue. Assuming, for example, that the air flow is west to east, most of the important boundary conditions should be coming in through the west boundary. Although some variables are 3D and go up vertically through the atmosphere, they are still initialized far to the west of the mountain, so I don't think it will be an issue. Does that make sense?