Error in running ./real.exe

[Rajib Lochan Ram]

I want to run a wrf urban (bep+bem) model on Gaziabad (WHICH IS AN INDIAN CITY).
I have run geogrid.exe successflly and now I want to run real.exe but i am getting error and unable to run it.

I have shared both namelist.wps and namelist.input for your referrence along with the error message screenshot received.
Requesting your help to sort the issue.

 

[kwerner]

Hi,
The error message you see is:

------ ERROR while reading namelist physics --------
Maybe here?:    &dynamics
Maybe here?:     hybrid_opt = 2,
----------------------------------------------------

Often, even if the error log mentions two lines from your namelist, the error can actually come from a line above those. I think in this case, the line giving the error is

fractional_seaice   = 0.0,

The expected value here is an integer, so you need to change this to fractional_seaice = 0.


I see numerous other issues with the namelist setup:

1) Your domains are entirely too small. Each domain should not be smaller than 100x100 grid spaces. See Why domains should be at least 100x100 grid spaces.

e_we   = 29, 55, 118,
e_sn   = 25, 43, 85,

2)

dx                                  = 2970,
dy                                  = 2970,

What is the resolution of your input data? We recommend that the first domain be no more than about a 1:5 ratio to the input data.

3)

physics_suite                       = 'URBAN',

This physics suite option does not exist. See Using Physics Suites from the WRF Users' Guide.

4)

mp_physics                          = 8, 2, 2,

All domain should use the same value for physics options, with the exception of cu_physics and the pbl, which can be turned off for some domains (e.g., they way you have cu_physics set up - cu_physics = 1.1.0 -- However, cu_physics does not need to be turned on for domains smaller than ~3km, so this would mean you should have it set to 0 for all domains)

5)

sf_surface_physics   = 2,

You need a value for each of the three domains.

6)

radt = 10

You need a value for each domain. We recommend this value be set to something similar to the integer value of the resolution of domain 01; however, it may be okay to keep it set to radt = 10, 10, 10

7)

sf_urban_physics = 2,

You need a value for each of the three domains.


There may be others, but those are the ones that stuck out to me. See the WPS Namelist Best Practices for some suggestions on setting up a reasonable domain, and you may also benefit from the WRF Namelist Best Practices.

 

[Rajib Lochan Ram]

Hi kwerner,

Thank you for your reply. Here are some of the reasons why I have used such a domain:
1) My domain size is small because of my limitations in extent of the urban morphology parameter's extent.
2) I have used 1:3 grid ratio such that Domain 3 is 330m, Domain 2 is 990m and Domain 1 is 2970m.
3) I have updated the physics suite to 'tropical' and has updated the namelist based on your suggestion too but still I am having issues in running ./real.exe. The error is constantly coming in &dynamics part and &bdy_control part. The eooro message is also attached here with the updated namelisgt.input.

Requesting your suggestions for the same.

Thanks and regards,
Rajib

​

 

[kwerner]

2) I have used 1:3 grid ratio such that Domain 3 is 330m, Domain 2 is 990m and Domain 1 is 2970m.

And what resolution is the input data (i.e., the first guess external model input - like GFS, ERA5, for example)?

Will you please attach the full real.exe rsl.error.0000 file? I'd like to be able to see the whole file, and not just the section that is in the screenshot. Thank you!

 

[Rajib Lochan Ram]

And what resolution is the input data (i.e., the first guess external model input - like GFS, ERA5, for example)?

Will you please attach the full real.exe rsl.error.0000 file? I'd like to be able to see the whole file, and not just the section that is in the screenshot. Thank you!

I am using 0.25 degrees 6 hour interval GFS data for Input.
Attached rsl.error.0000, rsl.out.0000, geogrid.tbl, namelist.wps and namelist.input for your reference.

 

[kwerner]

Thanks for sending those. These are the new error messages:

  ------ ERROR while reading namelist physics ------
Maybe here?:     &dynamics
Maybe here?:      hybrid_opt                          = 2,
  ------ ERROR while reading namelist dynamics ------
Maybe here?:     &bdy_control
Maybe here?:      periodic_x                          = .true.,=
  ------ ERROR while reading namelist bdy_control ------
Maybe here?:     &bdy_control
Maybe here?:      periodic_x                          = .true.,=
-------------- FATAL CALLED ---------------
FATAL CALLED FROM FILE:  <stdin>  LINE:   11680
ERRORS while reading one or more namelists from namelist.input.
-------------------------------------------

1) fractional_seaice is still listed as a "real" value, with a decimal. This can only be an integer value. I realize in my above message, I did put a period at the end of the sentence, which may have confused you to think I mean to keep the period after the zero. The value should be

fractional_seaice      = 0

2) You have the parameter 'gwd' in the namelist, but no value is assigned. It's just sitting there by itself. Please either assign a value or remove that line, altogether.

3) You have an extra "=" in this line. Please remove it.

periodic_x     = .true.,=

 

[Rajib Lochan Ram]

Thank you for the suggestions I have successfully ran ./real.exe but now i am facing issues in running ./wrf.exe.

I have 24 CPU Cores but I ran the code with 16 cores using mpirun -np 16 ./wrf.exe but the following error in rsl.out.0000 and rsl.error.0000 is seen. Attaching the updated namelist.input, namelist.wps and rsl.error.0000 and rsl.out.0000 files for your reference.

Requesting your help for the same:

 

[kwerner]

I'm glad to hear you were finally able to get past the real.exe errors. Since you are now experiencing a new issue, please open a new thread to discuss it. For easier search/readability, we prefer to keep each thread specific to individual issues. Thanks!

 

[Rajib Lochan Ram]

Ok thanks