real.exe not generating wrfbdy_d01 file, causing wrf.exe to fail with error "error opening wrfbdy_d01 for reading. IERR = -1021"

[spencerp]

Hello, I am trying to run the WRF model following these steps. Everything is going to plan, until I go to run real.exe.
The tutorial states to run real.exe, and it should generate: a wrfbdy_d01 file, and wrfinput_d0* for each domain. After I run real.exe (with mpirun -np 2 ./real.exe) I see the below logs in the rsl.out.0000 file, and I get the two wrfinput files for my two domains, however, I do not get the wrfbdy_d01 file.

...
d02 2024-03-01_00:00:00 Timing for output          2 s.
d02 2024-03-01_00:00:00 Timing for loop #    1 =          5 s.
d01 2024-03-01_00:00:00 real_em: SUCCESS COMPLETE REAL_EM INIT

Therefore, when I run wrf.exe in the next step, it fails, and the below error is found in the rsl.error.0000 file:

...
Timing for Writing wrfout_d01_2024-03-01_00:00:00 for domain        1:   24.20236 elapsed seconds
-------------- FATAL CALLED ---------------
FATAL CALLED FROM FILE:  <stdin>  LINE:    1841
 med_latbound_in: error opening wrfbdy_d01 for reading. IERR =        -1021
-------------------------------------------
application called MPI_Abort(MPI_COMM_WORLD, 1) - process 0

In my data/ folder I have three gfs gribfiles, gfs.t00z.pgrb2.0p25.f000 gfs.t00z.pgrb2.0p25.f003 gfs.t00z.pgrb2.0p25.f006. They are the 0h, 3h. and 6h forecast for a single datetime.
Attached are my namelist.wps and namelist.input files. I hope that it is just a silly type in one of those files causing the error.
Any help is appreciated, thanks!

 

[spencerp]

I figured out the problem I was having. The three GFG gribfiles I was using all have the same initialization datetime (2024,03,01,00:00z) however, they are for the 0h, 3h, and 6h forecast respectfully. Therefore, I needed to modify the two namelist files to have an end hour of 06 instead of 00!

 

[GewellLlorin]

Hello, I would like to report a similar error reported in this thread. I also run into an issue generating the wrfbdy_01 file from real.exe but in my case, the real.exe process doesn't even finish executing before ending with the error:

-------------- FATAL CALLED ---------------
FATAL CALLED FROM FILE: <stdin> LINE: 915
REAL: error opening wrfbdy_d01 for writing
-------------------------------------------

I'm unsure if I can apply the same solution the previous user did as I did check if my namelist details and input files aligned correctly. I've attached for now my namelist.wps, namelist.input, and rsl.error.0000 files for reference.

Any advice would be much appreciated, thank you very much!

 

[Ming Chen]

Can you set max_dom =1, then try again? We would like to know whether this is a data issue or a memory issue. I don't think there is anything wrong with the code.

 

[GewellLlorin]

Hello! I applied the suggestion but encountered the same error message (see log file attached.)

As a separate update, I attempted a simulation using the same input file on another server's WRF system that I previously used. It ran with no problems so I believe it is something with the installation of WRF on my new computer. However, re-configuring and recompiling WRF following the official tutorials doesn't seem to solve the problem.

Are there other things I could try out?

Thank you!

 

[Ming Chen]

I am suspicious this is an issue related to libs installation. Can you post the link you use for WRF installation?

 

[GewellLlorin]

Thank you for your quick response. I used the installation script for WRFv4.5.2 found on this page: GitHub - bakamotokatas/WRF-Install-Script: Install WRF with required libraries