Surface pressure computation problem in real.exe

[peter]

I recently changed a few details from successful WRF 3.8.1 simulations but unexpectedly I am having problems with real.exe. I just shifted from NCEP FNL Operational Model Global Tropospheric Analyses with 1degx1deg horizontal resolution to NCEP GDAS/FNL 0.25 Degree Global Tropospheric Analyses and Forecast Grids (respectively ds083.2 and ds083.3 at rda.ucar.edu) for the initial and boundary conditions and I slightly modified the studied region. Both datasets have the same WRF Vtable (www2.mmm.ucar.edu/wrf/users/download/free_data.html). I also modified the top for the simulations from 10 hPa to 1 hPa. Preprocessing with geogrid, ungrib and metgrid was OK with WPS 4.0 on a different computer, as with my previous successful WRF runs. I am using 1 node with 64 cores when running real.exe and I get the following message

$ tail rsl.error.0043
d01 2019-09-11_00:00:00 t(i,j,k) was 0 at 165 240 1 , setting Qv to 0
d01 2019-09-11_00:00:00 t(i,j,k) was 0 at 166 240 1 , setting Qv to 0
d01 2019-09-11_00:00:00 t(i,j,k) was 0 at 167 240 1 , setting Qv to 0
d01 2019-09-11_00:00:00 t(i,j,k) was 0 at 168 240 1 , setting Qv to 0
Using sfcprs3 to compute psfc
-------------- FATAL CALLED ---------------
FATAL CALLED FROM FILE: <stdin> LINE: 7255
probs with sfc p computation
-------------------------------------------

In a similar but different issue ("not enough info for a p sfc computation") like viewtopic.php?f=6&t=294&p=27002&hilit=sfcp_to_sfcp+%3D+.true.&sid=056f98cdf4e2d29680c38dd0a757e295#p27002 I saw that sometimes a solution was setting "sfcp_to_sfcp = .true." but in my case with debug_level=1000 I only got

d01 2019-09-11_00:00:00 Warning: vapor pressure exceeds total pressure, setting Qv to 1.E-6
Using sfcprs2 to compute psfc

I would very much appreciate any idea on how this issue can be solved. I am including below namelist.wps and namelist.input

Code: Select all

&share
 wrf_core = 'ARW',
 max_dom = 2,
 start_date = '2019-09-11_00:00:00', '2019-09-11_00:00:00',
 end_date   = '2019-09-12_12:00:00', '2019-09-12_12:00:00',
 interval_seconds = 21600,
 io_form_geogrid = 2,
 opt_output_from_geogrid_path = '/home/wrf/dominios/ST08_1/',
 debug_level = 0,
/

&geogrid
 parent_id         = 1,1,
 parent_grid_ratio = 1,3,
 i_parent_start    = 1,46,
 j_parent_start    = 1,82,
 e_we          = 364,697,
 e_sn          = 382,574,
 geog_data_res = 'default','default',
 dx = 9000,
 dy = 9000,
 map_proj =  'lambert',
 ref_lat   = -52.741,
 ref_lon   = -64.634,
 truelat1  = -52.741,
 truelat2  = -52.741,
 stand_lon = -64.634,
 geog_data_path = '/home/wrf/WRF/GEOG_DATA/WPS_GEOG',
 opt_geogrid_tbl_path = '/home/wrf/dominios/ST08_1/',
 ref_x = 182.0,
 ref_y = 191.0,
/

&ungrib
 out_format = 'WPS',
 prefix = 'FILE',
/

&metgrid
 fg_name = 'FILE',
 io_form_metgrid = 2,
 opt_output_from_metgrid_path = '/home/wrf/dominios/ST08_1/',
 opt_metgrid_tbl_path = '/home/wrf/dominios/ST08_1/',
/


Code: Select all

&time_control
run_days                 = 0,
run_hours                = 36,
run_minutes              = 0,
run_seconds              = 0,
start_year               = 2019,     2019,
start_month              = 9,        9,
start_day                = 11,       11,
start_hour               = 0,        0,
start_minute             = 00,       00,
start_second             = 00,       00,
end_year                 = 2019,     2019,
end_month                = 9,        9,
end_day                  = 12,       12,
end_hour                 = 12,       12,
end_minute               = 00,       00,
end_second               = 00,       00,
interval_seconds         = 21600,
input_from_file          = .true.,   .true.,
history_interval         = 180,       60,
frames_per_outfile       = 1000,     1000,
restart                  = .false.,
restart_interval         = 5000,
io_form_history          = 2,
io_form_restart          = 2,
io_form_input            = 2,
io_form_boundary         = 2,
debug_level              = 0,
/

&domains
time_step                = 22,
time_step_fract_num      = 1,
time_step_fract_den      = 2,
max_dom                  = 2,
e_we                     = 364,      697,
e_sn                     = 382,      574,
e_vert                   = 75,       75,
p_top_requested          = 100,
num_metgrid_levels       = 34,
num_metgrid_soil_levels  = 4,
dx                       = 9000,     3000,
dy                       = 9000,     3000,
grid_id                  = 1,        2,
parent_id                = 1,        1,
i_parent_start           = 1,       46,
j_parent_start           = 1,       82,
parent_grid_ratio        = 1,        3,
parent_time_step_ratio   = 1,        3,
feedback                 = 1,
smooth_option            = 0,
/

&physics
mp_physics               = 3,        3,
ra_lw_physics            = 1,        1,
ra_sw_physics            = 1,        1,
radt                     = 30,       30,
sf_sfclay_physics        = 1,        1,
sf_surface_physics       = 2,        2,
bl_pbl_physics           = 1,        1,
bldt                     = 0,        0,
cu_physics               = 0,        0,
cudt                     = 5,        5,
isfflx                   = 1,
ifsnow                   = 0,
icloud                   = 1,
surface_input_source     = 1,
num_soil_layers          = 4,
sf_urban_physics         = 0,        0,
maxiens                  = 1,
maxens                   = 3,
maxens2                  = 3,
maxens3                  = 16,
ensdim                   = 144,
/

&fdda
/

&dynamics
w_damping                = 0,
diff_opt                 = 1,
km_opt                   = 4,
diff_6th_opt             = 0,        0,
diff_6th_factor          = 0.12,     0.12,
base_temp                = 290.,
damp_opt                 = 0,
zdamp                    = 5000.,    5000.,
dampcoef                 = 0.2,      0.2,
khdif                    = 0,        0,
kvdif                    = 0,        0,
non_hydrostatic          = .true.,   .true.,
moist_adv_opt            = 1,        1,
scalar_adv_opt           = 1,        1,
/

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

&grib2
/

&namelist_quilt
nio_tasks_per_group      = 0,
nio_groups               = 1,
/


[dcvz]

By default for WPS V4.0.1 and later, ungrib only decodes the GFS levels at or below 1 hPa. To return to the V3.8.1 behavior add "pmin = 0." to the &ungrib section of namelist.wps. Then, ungrib will decode all levels in the GFS file. (This assumes that the GFS file contains fields higher than 1 hPa). Your WRF domain (x, y, and z) needs to fit within the first guess model.

In my opinion, a 1 hPa model top is too high. WRF doesn't contain the physics to properly handle the processes at that altitude.

[peter]

Thank you very much for your suggestion. I included "pmin = 0." in the &ungrib section of namelist.wps but now I get an error at the metgrid stage.

INFORM: Field LANDSEA.mask does not have a valid mask and will not be checked for missing values
WRF_DEBUG: Warning DIM 4 , NAME num_metgrid_levels REDEFINED by var VV 35 34 in wrf_io.F90 line 2588
ERROR: Error in ext_pkg_write_field
application called MPI_Abort(MPI_COMM_WORLD, 21937) - process 0

===================================================================================
= BAD TERMINATION OF ONE OF YOUR APPLICATION PROCESSES
= EXIT CODE: 177
= CLEANING UP REMAINING PROCESSES
= YOU CAN IGNORE THE BELOW CLEANUP MESSAGES
===================================================================================

I guess there is a conflict in the number of lelels.

[dcvz]

Well, now metgrid is complaining that the V wind component (called VV in metgrid) is missing. That's because the GFS file doesn't have wind fields at the 0.4 hPa level. WPS assumes that all variables are available at all levels. The pmin namelist option was introduced in ungrib after NCEP changed the fields available in the GFS. ungrib can't extrapolate missing fields to the 0.4 hPa level.

Your options are 1) create your own fields for the missing variables and add them to the intermediate files, 2) set pmin=100 (the default) and lower your ptop_requested to 2 hPa, or 3) use another first guess model (e.g. EC) which has output levels above 1 hPa.

[peter]

Thank you very much again. I don't know exactly why, but I repeated the preprocessing without pmin and thereafter real was successful this time. I guess that for some unknown reason metgrid had initially a problem but ended apparently successful. After real anyway wrf exploded briefly after the start, but as you said it may be unable to handle the physical processes at 1 hPa. However, there are some options that I will explore. If you don't mind I have a last question in reference to your last answer: p_top = pmin must be avoided ?

[dcvz]

I haven't looked at this in detail, but I like to have ptop be lower in altitude than pmin. Maybe due to roundoff or due to interp_type the values at eta=0 may not be correct where ptop=pmin.