Running Atmosphere Model/

Mgduda,

I want to think you for helping me out with a previous issue in another thread. I have searched the forum to resolve my current issue I am having running atmopshere_model. I am running the atmosphere_model for a limited area mesh for the tropical Atlantic using x4.535554. I ran the initial_atmosphere model the first run as described as config_init_case=7 with dimensions variables all set to 1. The config_static_interp=true and config_static_interp=true; my streams.initial_atmosphere input file is tropics.grid.nc to get tropics.static.nc. This compiles with no errors; I followed this by running the initial_atmosphere with config_intit_case=7 with my streams_initial_atmosphere input tropics.static.nc to create tropics.init.nc file. The second run with config_init_case=7 to create tropics.init.nc the log.initial_atmosphere had no errors.

I changed config_init_case=9 to create my boundary conditions with no errors in the log.initial_atmosphere file. Now when I changed my streams.atmosphere file to have tropics.init.nc file as my input; at the same time for boundary conditions as my namelist.initial_atmosphere file.

min/max of meshScalingDel2 = 1.00000 4.75683
min/max of meshScalingDel4 = 1.00000 107.635
Troubles finding level 100.000 above ground.
Problems first occur at (196)
Surface pressure = NaN hPa.
*** MSLP field will not be computed
Timing for diagnostic computation: 293.857 s
Timing for stream output: 299.692 s
WARNING: Variable lbc_qi not in input file.
WARNING: Variable lbc_qs not in input file.
WARNING: Variable lbc_qg not in input file.

Begin timestep 2019-09-22_00:00:00
WARNING: Variable lbc_qi not in input file.
WARNING: Variable lbc_qs not in input file.
WARNING: Variable lbc_qg not in input file.
----------------------------------------------------------------------
Updated lateral boundary conditions. LBCs are now valid
from 2019-09-22_00:00:00 to 2019-09-22_03:00:00
----------------------------------------------------------------------
--- time to update background surface albedo, greeness fraction.
--- time to run the LW radiation scheme L_RADLW =T
--- time to run the SW radiation scheme L_RADSW =T
--- time to run the convection scheme L_CONV =T
--- time to update the ozone climatology for RRTMG radiation codes
--- time to apply limit to accumulated rainc and rainnc L_ACRAIN =F
--- time to apply limit to accumulated radiation diags. L_ACRADT =F
--- time to calculate additional physics_diagnostics =F

I compiled the WPS model and my boundary conditions are GFS 0.25. I used the same FILES to run the WRF-model for TS Karen with no problems.

Could you attach your namelist.atmosphere file and your namelist.init_atmosphere file (as it was when you created your initial conditions)?

It may also be worth checking that the 'theta' (potential temperature), 'rho' (dry air density), and 'u' (normal component of horizontal velocity) fields in your initial conditions file all appear to be within a physically reasonable range. The easiest option may be to just use ncdump on the command-line.

Greetings,

I have attached my namelist and log.out files for you to review.

Thanks,
Travis

The x4.535554 mesh refines to a minimum grid distance of about 15 km. Could you try making two adjustments in your namelist.atmosphere file:

• config_dt = 90.0
• config_len_disp = 15000.0

In general, the value of 'config_len_disp' should match the minimum grid distance in meters in your mesh. There is slightly more latitude in choosing the integration timestep, 'config_dt', but a reasonable starting value is around six times the minimum grid distance in km.

Mgduda,

I tried your suggustion changing config_dt = 90.0 and config_len_disp = 15000.0 in the namelist.atmopshere. The atmosphere model stopped again. I put the log.atmopshere.out and .atmosphere info below.

Thanks,
Travis
_______________________________________________________________________________________________________________________________
min/max of meshScalingDel2 = 1.00000 4.75683
min/max of meshScalingDel4 = 1.00000 107.635
Troubles finding level 100.000 above ground.
Problems first occur at (170)
Surface pressure = NaN hPa.
*** MSLP field will not be computed
Timing for diagnostic computation: 183.559 s
Timing for stream output: 181.966 s
WARNING: Variable lbc_qi not in input file.
WARNING: Variable lbc_qs not in input file.
WARNING: Variable lbc_qg not in input file.

Begin timestep 2019-09-22_00:00:00
WARNING: Variable lbc_qi not in input file.
WARNING: Variable lbc_qs not in input file.
WARNING: Variable lbc_qg not in input file.
----------------------------------------------------------------------
Updated lateral boundary conditions. LBCs are now valid
from 2019-09-22_00:00:00 to 2019-09-22_03:00:00
----------------------------------------------------------------------
--- time to update background surface albedo, greeness fraction.
--- time to run the LW radiation scheme L_RADLW =T
--- time to run the SW radiation scheme L_RADSW =T
--- time to run the convection scheme L_CONV =T
--- time to update the ozone climatology for RRTMG radiation codes
--- time to apply limit to accumulated rainc and rainnc L_ACRAIN =F
--- time to apply limit to accumulated radiation diags. L_ACRADT =F
--- time to calculate additional physics_diagnostics =F
______________________________________________________________________________________________________________

Program received signal SIGSEGV: Segmentation fault - invalid memory reference.

Backtrace for this error:
#0 0x1051f52be
#1 0x1051f4a1a
#2 0x7fff6a34d5fc
#3 0x1042281a0
#4 0x104240368
#5 0x10424fa3d
#6 0x104257640
#7 0x104324641
#8 0x1042d77ef
#9 0x103fd96fd
#10 0x103fda6e0
#11 0x103fd3a3d
#12 0x10453f55d
[2] 10141 segmentation fault ./atmosphere_model

These messages in your log file could be indicative of problems with the initial conditions:

Have you tried taking a quick look through your initial conditions file to verify that the 'theta', 'rho', and 'u' fields look plausible? One way to do this would be to just print them to the terminal with ncdump. Alternatively, you could try using the convert_mpas utility to interpolate your initial conditions to a regular lat-lon grid and to then browse the initial fields with, e.g., ncview.

Mgduda,

I tried using the ECMWF, Gdas1, and GFS0.25. However, I am getting Troubles finding level 100.000 above ground and Surface pressure = NaN hPa error. I compiled the WPS, and ran the data on the WRF with no problems. So, my question is can you run the Limited-Area to get your grid.nc file and then run grid_rotate with the same grid.nc file created by the limited-Area program?

Thanks for your help again.

Strange, I'm trying to replicate your situation, but then, mgduda's second fix - the one about grid distances - solves it.
I'll give it another try, and will tell you if I ever find something worth telling.

FLUIDDYNAMICS2, can you check the min/max values of the 'surface_pressure' field in your tropics.init.nc file? Does the surface_pressure field contain any NaNs?