Streaks in the output field

[smeech84]

What could be causing these streaks in the attached image? This image shows WRF output of the variable U10 for a simulation run over 2003 7/25 @18:00Z centered on Oklahoma City. I used CFSR data as initial conditions with the following name list options. The model was run from 7/23 - 7/28 so there should be proper spin up prior to this time.

Additionally, there also seems to be some non-physical discontinuities along the border... Has anyone else seen this type of behavior? I have been struggling with issues like this for a while now and would really like to find a solution. Thanks!

&time_control
run_days = 0,
run_hours = 120,
run_minutes = 0,
run_seconds = 0,
start_year = 2003, 2003, 2003, 2003,
start_month = 7, 7, 7, 7,
start_day = 23, 23, 23, 23,
start_hour = 0, 0, 0, 0,
start_minute = 00, 00, 00, 00,
start_second = 00, 00, 00, 00,
end_year = 2003, 2003, 2003, 2003,
end_month = 7, 7, 7, 7,
end_day = 28, 28, 28, 28,
end_hour = 0, 0, 0, 0,
end_minute = 00, 00, 00, 00,
end_second = 00, 00, 00, 00,
interval_seconds = 10800,
input_from_file = .true., .true., .true., .true.,
history_interval = 60, 60, 5, 360,
frames_per_outfile = 1, 1, 1, 1,
auxinput11_interval_s = 180, 180, 180, 180,
auxinput11_end_h = 6, 6, 6, 6,
restart = .false.,
restart_interval = 60,
io_form_history = 2,
io_form_restart = 2,
io_form_input = 2,
io_form_boundary = 2,
debug_level = 0,
force_use_old_data = T,
/

&domains
use_adaptive_time_step = .true.
step_to_output_time = .true.
time_step = 150,
time_step_fract_num = 0,
time_step_fract_den = 1,
max_dom = 3,
e_we = 151, 151, 151,
e_sn = 151, 151, 151,
e_vert = 37, 37, 37, 37,
dx = 10000, 3333.3, 1111.1,
dy = 10000, 3333.3, 1111.1,
p_top_requested = 5000,
num_metgrid_levels = 38,
num_metgrid_soil_levels = 4,
sfcp_to_sfcp = .true.,
grid_id = 1, 2, 3, 4,
parent_id = 0, 1, 2, 3,
i_parent_start = 1, 51, 51, 28,
j_parent_start = 1, 51, 51, 22,
parent_grid_ratio = 1, 3, 3, 3,
parent_time_step_ratio = 1, 3, 3, 3,
feedback = 0,
smooth_option = 0,
eta_levels = 1.000000, 0.996200, 0.989737, 0.982460, 0.974381,
0.965422, 0.955498, 0.944507, 0.932347, 0.918907,
0.904075, 0.887721, 0.869715, 0.849928, 0.828211,
0.804436, 0.778472, 0.750192, 0.719474, 0.686214,
0.650339, 0.611803, 0.570656, 0.526958, 0.480854,
0.432582, 0.382474, 0.330973, 0.278674, 0.226390,
0.175086, 0.132183, 0.096211, 0.065616, 0.039773,
0.018113, 0.000000,
/

&physics
mp_physics = 8, 8, 8, 8,
progn = 1, 1, 1, 1,
ra_lw_physics = 1, 1, 1, 1,
ra_sw_physics = 1, 1, 1, 1,
radt = 10, 10, 10, 20,
sf_sfclay_physics = 1, 1, 1, 1,
sf_surface_physics = 2, 2, 2, 2,
bl_pbl_physics = 1, 1, 1, 1,
bldt = 0, 0, 0, 0,
cu_physics = 0, 0, 0, 0,
cudt = 5, 5, 5, 0,
sf_urban_physics = 0, 0, 0, 0,
isfflx = 1,
ifsnow = 1,
icloud = 1,
surface_input_source = 1,
num_soil_layers = 4,
num_land_cat = 21,
maxiens = 1,
maxens = 3,
maxens2 = 3,
maxens3 = 16,
ensdim = 144,
cu_rad_feedback = .false.,
cugd_avedx = 1,
mp_zero_out = 2,
mp_zero_out_thresh = 1.e-8,
cu_diag = 0,
slope_rad = 1, 1, 1,
/

&dynamics
w_damping = 1,
diff_opt = 1,
km_opt = 4,
base_temp = 290.,
damp_opt = 3,
zdamp = 5000., 5000., 5000., 5000.,
dampcoef = 0.15, 0.15, 0.15, 0.2,
khdif = 0, 0, 0, 0,
kvdif = 0, 0, 0, 0,
non_hydrostatic = .true., .true., .true., .true.,
moist_adv_opt = 1, 1, 1, 4,
scalar_adv_opt = 1, 1, 1, 4,
tke_adv_opt = 1, 1, 1, 4,
/

&bdy_control
spec_bdy_width = 5,
spec_zone = 1,
relax_zone = 4,
specified = .true., .false., .false., .false.,
nested = .false., .true., .true., .true.,
/

&namelist_quilt
nio_tasks_per_group = 0,
nio_groups = 1,
/

&fdda
obs_nudge_opt = 0,0,0,0,0
max_obs = 150000,
fdda_start = 0., 0., 0., 0., 0.
fdda_end = 99999., 99999., 99999., 99999., 99999.
obs_nudge_wind = 1,1,1,1,1
obs_coef_wind = 8.E-4,8.E-4,8.E-4,6.E-4,6.E-4
obs_nudge_temp = 1,1,1,1,1
obs_coef_temp = 8.E-4,8.E-4,8.E-4,6.E-4,6.E-4
obs_nudge_mois = 1,1,1,1,1
obs_coef_mois = 8.E-4,8.E-4,8.E-4,6.E-4,6.E-4
obs_rinxy = 240.,240.,180.,180,180
obs_rinsig = 0.1,
obs_twindo = 0.6666667,0.6666667,0.6666667,0.6666667,0.6666667,
obs_npfi = 10,
obs_ionf = 2, 2, 2, 2, 2,
obs_idynin = 0,
obs_dtramp = 40.,
obs_prt_freq = 10, 10, 10, 10, 10,
obs_prt_max = 10
obs_ipf_errob = .true.
obs_ipf_nudob = .true.
obs_ipf_in4dob = .true.
obs_ipf_init = .true.
/

 

[kwerner]

Hi,
You have a lot of non-traditional options in your namelist file. It's okay to use all that you're using, but in order to try to track down the problem, it's best to start with mostly default namelist options and see if you're able to run that. If not, then we know there's likely something wrong with the data. If you are able to run with a simple namelist, then you can start adding some of your other options back in - one or a few at a time - until you find something that causes this. I'm attaching a simplified version of your namelist. Give it a try and I'd recommend only running for as long as your need to - the first time you notice the streaks in your original simulation.

I have a couple of questions:
1) Is there a reason you have feedback turned off? You could try turning it on to see if that helps at all.
2) What version of the model are you using? If it's anything older than 4.0, can you also try with a newer version to see if you still see the problem?

 

[smeech84]

Thanks for the response!

The back story behind the name list options is that I was trying to duplicate an old WRF simulation that someone had done with WRF version 3.4.1 years ago. All we had were the wrfout files, not a namelist so I had to try to piece it together based on that alone.

Based on your suggestion I ran it again with the new namelist options you provided. One run with feedback on and one without. Unfortunately, they still look very similar (attached).

I don't think I'm doing anything fancy in the set up. Just using regular CFSR data, no obsgrid, no nudging. It is interesting that other times around this period seemingly look OK. There's about 14 hours of this streaky type of behavior from maybe 7/25 14:00Z to 7/26 04:00Z (simulation runs from 7/23 to 7/28).

Any other suggestions are appreciated!

Thanks!

 

[smeech84]

Another update:

I ran the same case again using the recommendations from here:
https://www2.mmm.ucar.edu/wrf/users/namelist_best_prac_wrf.html
https://www2.mmm.ucar.edu/wrf/users/namelist_best_prac_wps.html

I stripped down the extra name list options I had left over and only used the above. Still the same type of streaks appear during the same times. Although these new simulations are different, they still have the border discontinuities and seemingly non-physical streaks in them.

This is all run on the Cheyenne cluster using the CFSR data located at /gpfs/fs1/collections/rda/data/ds093.0 using the pgbhnl and flxf06 data.

 

[kwerner]

When you say that other times around this time look okay - does this mean you're using the exact same type of input data, but different dates/times, and then the exact same namelist set-up (only changing date/time settings), and it turns out okay? If that is the case, then it seems like the problem lies with the input data and something is wrong with that particular data. If that's the case, unfortunately you would need to contact the data source to see if they are aware and have any suggestions.

 

[smeech84]

Thank you for the reply.

What I mean is during this continuous WRF simulation period from 7/23 - 7/28, from about 7/25 14:00Z to 7/26 04:00Z it displays this streaky type of behavior. The times around this period, still within this continuous WRF simulation (7/23 - 7/28), look seemingly OK.

 

[kwerner]

Thanks for clarifying. Can you take a look at your input data and see if there is anything different about the data during the "bad" (streaked) times versus other times?