program crashed while run real.exe

[omoakinyoola73]

After successfully run geogrid, ungrib and metgrid

i was trying to run real.exe but the program crashed with the following
[jakinyoola@login2 wrfprd]$ real.exe
starting wrf task 0 of 1
[jakinyoola@login2 wrfprd]$
View attachment namelist.input

 

[mgduda]

I've moved this topic from the MPAS-Atmosphere section of the forum to here since the question concerns the real.exe program.

 

[kwerner]

Hi,
If you take a look at the end of your rsl.error.0000 file, it likely has the reason for the failure. Try to address the problem and run it again. If you are still having problems, please send me the rsl.error.* files and a new namelist.input file (if you updated it). Please also let me know the version of WRF you're running. Thanks!

 

[omoakinyoola73]

I opened the file rsl.error.0000 but unable to identify the exact variable in the Domain and Chem parameter in order to correct the error

the attached file are nameliat.input, nameliat.wps and rsl.error files
and the wrfchem version: WRFCHEM-3.9.1-pnc-gccmpich on CHPC

with kind regards
Julius

 

[kwerner]

Julius,
It looks like your namelist.input file includes several lines in which you've added additional columns, when those particular variables should only have a single value (i.e., not a value for each domain). For e.g., I see this in the namelist:

time_step                = 310,506,
time_step_fract_num      = 0,
time_step_fract_den      = 1,
max_dom                  = 2,
e_we                     = 100,142,
e_sn                     = 57,73,
e_vert                   = 28,28,
p_top_requested          = 5000,5000,
num_metgrid_levels       = 27,

The variable time_step should only have a value for d01. The same is true for p_top_requested.

There are possibly others like this, and for the chem section of the namelist. To determine whether the variables should have a single or multiple values, take a look at the WRF/Registry/Registry.EM file. Here is the line for time_step

rconfig   integer time_step               namelist,domains      1            -1       ih   "time_step"

and here is the line for e_we

rconfig   integer e_we                    namelist,domains      max_domains    32      irh    "e_we"          ""      ""

Notice e_we has a column that says "max_domains." This indicates a value is expected for each domain. Whereas time_step simply has a "1" in that column, indicating only a single value is required. You may have to use the registry.chem file for some of the chemistry variables.