Effect of damp_opt on plume simulation?

[Again]

Hi, I am new in wrf and trying to use wrf_les to simulate real plume, when I set damp_opt=0, the plume looks good(i think, not sure):
But after i set damp_opt=3, the plume looks strange:

I referred to this paper:

Does this mean that when setting damp_opt=3, I have to modify the Coriolis parameters and set the damping layers and other parameters, otherwise the generated plume is incorrect? Is damp_opt=3 necessary in simulating plumes? Because other papers do not mention damp_opt.Appreciate any reply or information : )

 

[Ming Chen]

damp_opt = 3 only works for real-data case. Please turn it off for your ideal case.

 

[Christian_]

damp_opt = 3 only works for real-data case. Please turn it off for your ideal case.

Hello Ming Chen,

as this damp_opt setting in the namelist and your statement confuses me, may I ask two follow-up questions?

The description of damp_opt in the readme.namelist looks like this:
damp_opt = 0, ! upper level damping flag
0 = without damping
1 = with diffusive damping, maybe used for real-data cases
(dampcoef nondimensional ~0.01-0.1)
2 = with Rayleigh damping (dampcoef inverse time scale [1/s] e.g. .003; idealized case only
not for real-data cases)
3 = with w-Rayleigh damping (dampcoef inverse time scale [1/s] e.g. .2;
for real-data cases)

- Even though one can derive from this list that damp_opt = 2 should be used for idealized cases (if one wants damping at all), I have seen that the default setting in the namelist.input for the idealized cases em_hill2d_x and em_convrad is indeed damp_opt = 3. Why does it work for these ideal cases and not for others like em_les?

- Just to clarify:
Does damp_opt = 1 do what is described in section 4.4.1 ("Absorbing Layer Using Spatial Filtering") of the "Description of the Advanced Research WRF Model Version 4" from July 2021 (should be the newest one)? And damp_opt = 2 is related to section 4.4.3 ("Traditional Rayleigh Damping Layer") and damp_opt = 3 to 4.4.2 ("Implicit Rayleigh Damping for the Vertical Velocity")? If that is true, then there is this sentence in 4.4.2: "We recommend its use in both large-scale and small-scale applications, and in idealized and real-data applications." which raises again the question why we should not use damp_opt = 3 for idealized cases.

Regards,
Christian

 

[Ming Chen]

Christian,

The two ideal cases (em_hill2d_x and em_convrad) both have high model top (30km) , while for em_les the model top is pretty low. damp_opt = 3 is designed for real-data case, although for some ideal cases with very high model top, it can also work. Please take a look at the paper below:

Klemp, J. B., J. Dudhia, and A. Hassiotis, 2008: An Upper Gravity Wave Absorbing Layer forNWP Applications. Mon. Wea. Rev.., 136, 3987–4004.

damp_opt = 1 is described in section 4.4.1 ("Absorbing Layer Using Spatial Filtering") of the "Description of the Advanced Research WRF Model Version 4" from July 2021, damp_opt = 2 is related to section 4.4.3 ("Traditional Rayleigh Damping Layer") , and damp_opt = 3 to 4.4.2 ("Implicit Rayleigh Damping for the Vertical Velocity").

 

[Christian_]

Christian,

The two ideal cases (em_hill2d_x and em_convrad) both have high model top (30km) , while for em_les the model top is pretty low. damp_opt = 3 is designed for real-data case, although for some ideal cases with very high model top, it can also work. Please take a look at the paper below:

Klemp, J. B., J. Dudhia, and A. Hassiotis, 2008: An Upper Gravity Wave Absorbing Layer forNWP Applications. Mon. Wea. Rev.., 136, 3987–4004.

damp_opt = 1 is described in section 4.4.1 ("Absorbing Layer Using Spatial Filtering") of the "Description of the Advanced Research WRF Model Version 4" from July 2021, damp_opt = 2 is related to section 4.4.3 ("Traditional Rayleigh Damping Layer") , and damp_opt = 3 to 4.4.2 ("Implicit Rayleigh Damping for the Vertical Velocity").

Thanks for your clarification, Ming.
Although I still find it a bit confusing that the use of damp_opt = 3 is generally recommended even for idealized cases in section 4.4.2 in the description.

 

[Ming Chen]

Hi Christian,
I share the same confusion with you, and honestly I don't fully understand why it works for ideal cases of high model top but fails when the model top descends to lower levels. I guess I need to revisit the paper and get some further in-depth knowledge of this option. Sorry for not being able to be more helpful on this issue.