Question about Vertical Levels for Model and Observation Comparison

[Naser]

Hello,


I have a question regarding vertical levels in WRF.
I need to compare my weather vertical observations with model output. I have two types of observational data:

1. Data based on height (meters)
2. Data based on pressure (hPa)

When setting up the model in the namelist.input, we define e_vert (number of vertical levels), and optionally, we can specify eta_levels ranging from 0 to 1 for vertical coordinate distribution.

After running the model, I use ARWpost to extract variables, where I can choose to interpolate model outputs to either height or pressure levels by setting interp_levels = 1.

My questions are:

1. If I want to compare vertical observations with model output, do I need to set e_vert equal to the number of observation levels?
2. Do I have to define eta_levels, or is it optional?
3. In general, do you have experience with such comparisons, and could you please share how you usually handle this process?

I would really appreciate your advice on this.

Thank you very much, and I look forward to your response.

Best regards,
Naser

 

[Ming Chen]

Please see my answers below:

Hello,


I have a question regarding vertical levels in WRF.
I need to compare my weather vertical observations with model output. I have two types of observational data:

1. Data based on height (meters)
2. Data based on pressure (hPa)

When setting up the model in the namelist.input, we define e_vert (number of vertical levels), and optionally, we can specify eta_levels ranging from 0 to 1 for vertical coordinate distribution.

After running the model, I use ARWpost to extract variables, where I can choose to interpolate model outputs to either height or pressure levels by setting interp_levels = 1.

My questions are:

1. If I want to compare vertical observations with model output, do I need to set e_vert equal to the number of observation levels?

No, you don't need to do so. This is because WRF model vertical coordinate is terrain-following and you cannot specify the levels exactly the same as observation levels.

1. Do I have to define eta_levels, or is it optional?

You can skip specifying eta_levels in namelist.input, and REAL program will calculate eta levels based on ptop and e_vert.

1. In general, do you have experience with such comparisons, and could you please share how you usually handle this process?

NCAR develops a model evaluation tool package for verification of WRF. Please see

dtcenter.org

dtcenter.org

This is software package open to public. It is a very useful tool for model verification.

I would really appreciate your advice on this.

Thank you very much, and I look forward to your response.

Best regards,
Naser

 

[Naser]

Dear Chen,
Thank you for your reply.

No, you don't need to do so. This is because WRF model vertical coordinate is terrain-following and you cannot specify the levels exactly the same as observation levels.

Since I am working with FDDA, I provide Little-R vertical observations for my FDDA run.
Each observation station has a specific height.

My question is:
How does the model apply these observations during the simulation?
And, how can we compare observations at a specific height with the model values at that same height?


Thank you very much, and I look forward to your response.
Naser