How to suppress false convection in the upper layer of the model and avoid w-cfl problem

[husiyang22]

hello everyone! I run an idealized case for several times and the w-cfl problems always happen... I don't know if it is related to my eta_levels sets?
below are my namelist:
&time_control
run_days = 100,
run_hours = 0,
run_minutes = 0,
run_seconds = 0,
start_year = 2024,
start_month = 01,
start_day = 01,
start_hour = 0,
start_minute = 00,
start_second = 00,
history_interval = 60,
frames_per_outfile = 1,
restart = .false.,
restart_interval = 1440,
io_form_history = 2
io_form_restart = 2
io_form_input = 2
io_form_boundary = 2
/

&domains
time_step = 10,
time_step_fract_num = 0,
time_step_fract_den = 1,
max_dom = 1,
s_we = 1,
e_we = 513,
s_sn = 1,
e_sn = 513,
s_vert = 1,
e_vert = 65,
dx = 3000,
dy = 3000,
ztop = 28000.,
eta_levels = 1.000000, 0.996000, 0.992074, 0.987000, 0.982000, 0.976000, 0.969000, 0.962800, 0.957000, 0.950000, 0.942500, 0.934000, 0.925000,0.917000,
0.909300, 0.887500, 0.875600, 0.863600, 0.851500, 0.839300, 0.817000, 0.804600,
0.792100, 0.779500, 0.766800, 0.753000, 0.741100, 0.728000, 0.715000,0.701800, 0.686500, 0.671000,
0.658000, 0.634400, 0.615600, 0.596700, 0.574600, 0.554400, 0.534000, 0.511500,
0.491800, 0.470000, 0.45000, 0.429800, 0.409400, 0.388800, 0.365000, 0.340000,
0.325000, 0.308000, 0.280000, 0.261000, 0.241000, 0.220000, 0.200000,
0.180000, 0.160000, 0.140000, 0.120000, 0.100000, 0.080000,
0.060000, 0.040000, 0.020000, 0.000000, !

/

&physics
mp_physics = 6,
ra_lw_physics = 4,
ra_sw_physics = 4,
radt = 3,
sf_sfclay_physics = 1,
sf_surface_physics = 1,
bl_pbl_physics = 1,
bldt = 0,
cu_physics = 0,
cudt = 0,
bucket_mm =10.,
bucket_J =1.e8,
/

&fdda
/

&dynamics
hybrid_opt = 0,
rk_ord = 3,
diff_opt = 1, 1, 1,
km_opt = 4, 4, 4,
damp_opt = 3,
dampcoef = .2,
zdamp = 12000.,
w_damping=1,
khdif = 0,
kvdif = 0,
smdiv = 0.1,
emdiv = 0.01,
epssm = 0.2,
time_step_sound = 6,
h_mom_adv_order = 5,
v_mom_adv_order = 3,
h_sca_adv_order = 5,
v_sca_adv_order = 3,
mix_full_fields = .true.,
non_hydrostatic = .true.,
pert_coriolis = .false.,
/

&bdy_control
periodic_x = .true.,
symmetric_xs = .false.,
symmetric_xe = .false.,
open_xs = .false.,
open_xe = .false.,
periodic_y = .true.,
symmetric_ys = .false.,
symmetric_ye = .false.,
open_ys = .false.,
open_ye = .false.,
/

&grib2
/

&namelist_quilt
nio_tasks_per_group = 0,
nio_groups = 1,
/
&ideal
ideal_case = 3
/
I would appreciate a lot if you can give some suggestions!

 

[Ming Chen]

Which ideal case are you running? How did you create initial condition for this case? How long did the model run before it crashed? In addiiton to CFL violation, did you see other error messages?

 

[husiyang22]

Dear Chen Ming:

Which ideal case are you running?

I'm running the em_convrad

How did you create initial condition for this case?

I use a temperature and humidity profile got by a 100-days small region simulation of radiation convection equilibrium.
Besides, I set the solar radiation fixed at 413.9W/m2, and change the delt to 0.1K instead of 1K, which is the amplitude of the initial random temperature pertubation of lowest 10 levels.

How long did the model run before it crashed?

The mode stopped running after 11 hours. I obtained approximately 17days simulation results

In addiiton to CFL violation, did you see other error messages?

so far, I did not see other error

 

[husiyang22]

Which ideal case are you running? How did you create initial condition for this case? How long did the model run before it crashed? In addiiton to CFL violation, did you see other error messages?

the current situation is: I dont know how to set the eta_levels to a proper set of levels. And it seems like the ideal_case em_convrad dont support to automatic generation by given e_vert

 

[Ming Chen]

Can you try with the default eta_levels provided in the namelist.input for em_convrad?

 

[husiyang22]

Can you try with the default eta_levels provided in the namelist.input for em_convrad?

Hello, Dear Chen Ming.
I have tried the 35 levels in default setting and successfully complete simulation in a small region. However, the vertical levels is too less, i need at least 49 levels in vertical direction

 

[husiyang22]

Dear Chen Ming:
So far, I have create a 49 and 52 levels setting of eta_levels:
The first 49 levels one is generate evenly spaced:
[1.0, 0.9791666667, 0.9583333333, 0.9375, 0.9166666667, 0.8958333333, 0.875, 0.8541666667, 0.8333333333, 0.8125, 0.7916666667, 0.7708333333, 0.75, 0.7291666667, 0.7083333333, 0.6875, 0.6666666667, 0.6458333333, 0.625, 0.6041666667, 0.5833333333, 0.5625, 0.5416666667, 0.5208333333, 0.5, 0.4791666667, 0.4583333333, 0.4375, 0.4166666667, 0.3958333333, 0.375, 0.3541666667, 0.3333333333, 0.3125, 0.2916666667, 0.2708333333, 0.25, 0.2291666667, 0.2083333333, 0.1875, 0.1666666667, 0.1458333333, 0.125, 0.1041666667, 0.0833333333, 0.0625, 0.0416666667, 0.0208333333, 0.0]
The second 52 levels one is generate by reading a different real_case wrfout ZNU I create many months ago:
[0.99814224 0.9942312 0.98990947 0.9851363 0.9798674 0.97405493
0.96764696 0.96058786 0.9528178 0.944273 0.93488586 0.92458475
0.91329455 0.90093756 0.8874333 0.8727 0.8566556 0.83921903
0.82031244 0.79986286 0.7778051 0.7540846 0.72866094 0.70151174
0.6726363 0.6420604 0.6098408 0.5760689 0.5408751 0.5044272
0.4669199 0.42857814 0.389669 0.35050026 0.31141704 0.2727963
0.23503801 0.19932318 0.16729245 0.13933377 0.11492942 0.09362759
0.07503383 0.05880387 0.0446372 0.03227152 0.02147788 0.01205643
0.00383273]
Can you give me some suggestion of these two levels setting? Thank you very much.

 

[Ming Chen]

One concern I have on the setting of 49 levels is that, the vertical resolution in PBL is too coarse, which may affect the simulation of surface and PBL processes, leading to unreasonable results. On the contrary, the setting of 52 levels has very high resolution in the lower atmosphere. This is my personal opinion based on real-data cases. Sorry that I don't have many experiences running ideal cases. Note that the default option provided in the officially released code is often the optimum option. If you need to make any changes, probably you can refer to the literature and find more information.

 

[husiyang22]

Dear Chen Ming:
Thank you for your suggestion.
Given that it is difficult to find a specific eta_levels definitions in relevant literature, I decided to find some threads which have eta_levels definitions and just copy them...
Good luck to me...