Surface friction for high-res runs using 3D TKE-Diffusion

Topics related to current and future physics in the WRF as well as any problems you may have.

Surface friction for high-res runs using 3D TKE-Diffusion

Postby PBLer » Tue May 20, 2008 3:02 pm

I am running WRFV3 with a high-resolution nest (500 m grid spacing). For this nest, I turned off the PBL scheme and used the 3D TKE-diffusion option instead. This simulation looks as if there is no surface friction, since the max winds are located at the surface. In contrast, the parent nest (with 1.5 km grid spacing) used the MYJ PBL scheme and produced a barrier jet located at the top of the PBL. Does anyone know the proper physics options to include surface friction when using the 3D TKE-diffusion option? Here are the physics and dynamics namelist options that I used for the high-res nest:

&physics
mp_physics = 8,
ra_lw_physics = 1,
ra_sw_physics = 1,
radt = 3,
sf_sfclay_physics = 0,
sf_surface_physics = 3,
bl_pbl_physics = 0,
bldt = 0,
cu_physics = 0,
cudt = 3,
isfflx = 2,
ifsnow = 0,
icloud = 1,
surface_input_source = 1,
num_soil_layers = 6,
ucmcall = 0,
mp_zero_out = 1,
mp_zero_out_thresh = 1.e-8,
maxiens = 1,
maxens = 3,
maxens2 = 3,
maxens3 = 16,
ensdim = 144,
slope_rad = 0,
topo_shading = 0,
/

&dynamics
w_damping = 1,
diff_opt = 2,
km_opt = 2,
base_temp = 290.,
damp_opt = 0,
zdamp = 5000.,
dampcoef = 0.01,
khdif = 0,
kvdif = 0,
smdiv = 0.1,
emdiv = 0.01,
epssm = 0.1,
time_step_sound = 4,
h_mom_adv_order = 5,
v_mom_adv_order = 3,
h_sca_adv_order = 5,
v_sca_adv_order = 3,
non_hydrostatic = .true.,
pd_moist = .false.,
pd_scalar = .false.,
pd_chem = .false.,
pd_tke = .false.,
mix_full_fields = .true.,
mix_isotropic = 0,
mix_upper_bound = 0.1,
tke_drag_coefficient = 0.001,
tke_heat_flux = 0.,
/

thanks,
joe
PBLer
 
Posts: 38
Joined: Wed May 07, 2008 5:11 pm

Re: Surface friction for high-res runs using 3D TKE-Diffusion

Postby jimmyc » Wed May 21, 2008 1:18 pm

are you sure the nest dynamic options are being set? Not sure if this applies to version 3, but do all the dynamic options need a specification for the nest?

I dont know how the model works when you use a pbl scheme on one grid and on the nest use the dynamic options. Someone else was asking about this in another post, but never responded. Might want to make sure that the dynamic option is being called on the nest. Pleas epost back if you see where the code is which runs the conditional for the choice between PBL and dynamic options.
The views expressed in this message do not necessarily reflect those of NOAA or the National Weather Service or the University of Oklahoma.
James Correia, Jr
jimmyc
 
Posts: 519
Joined: Tue Apr 15, 2008 1:10 am

Re: Surface friction for high-res runs using 3D TKE-Diffusion

Postby PBLer » Fri May 23, 2008 1:59 pm

I found the reason for my lack of surface friction and it was my problem, not WRF's. I had isfflx=2, which used the drag from the surface layer physics and I had sf_sfclay_physics=0 (no surface layer). I ran another simulation with with isfflx=0 with tke_drag_coef=0.001 and tke_heat_flux=0.01 and got a *reasonable* barrier jet structure (with surface drag), but the simulaton was quite noisy in regions of high Richardson number flow (which I think we should expect).

I am confident that the dynamic options are being set, but not sure if ALL the dynamic options need a specification for the nest. The wrfouts have non-zero values for TKE when I used the 3D TKE diff option (km_opt=2) and are zero otherwise.

I agree with your word of caution on going directly from a PBL scheme to an LES subgrid model. There haven't been very many studies exploring the model's capability in handling this transition of scales. To my knowledge, it seems as if there is no consensus on how best to configure the WRF for these scales. Just for comparison, I did an additional simulation using the MYJ for this hi-res domain. Despite not being designed for 500 m grid spacing, my first impression is that the model solution looks just as reasonable as the simulation using the 3D TKE diffusion. I'll verify both runs against obs to get a more meaningful assessment.
PBLer
 
Posts: 38
Joined: Wed May 07, 2008 5:11 pm

Re: Surface friction for high-res runs using 3D TKE-Diffusion

Postby PBLer » Fri May 23, 2008 3:18 pm

sorry, I meant in regions of LOW Richardson #, of course.
PBLer
 
Posts: 38
Joined: Wed May 07, 2008 5:11 pm


Return to Physics

Who is online

Users browsing this forum: No registered users and 7 guests