ERROR : The mandatory field TT was not found in any input data

[alainF]

Hello everyone,
Sometimes, I get the error message written in the subject of this post. This message is issued at the end of metgrid.
I underline that it happens randomly maybe once a week, while there is one run per day. The consequence is that wrf stops after the first wrf.out file has been sent. I have sent 3 files in nextcloud : gfs.f000, FILE 2022-09-11_00, met_em.d01.2022-09-11_00.nc
When I read the met_em file with Panoply, I can see a TT field with 34 levels from metgrid. When I read the intermediate FILE with the utility in WPS/util, I can see 34 levels in the field TT. What is really missing ? I don't know. It looks weird because the problem does not happen every time I run the simulation.
WRF 4.3.1
WPS 4.3
Thanks for any help.

 

[Ming Chen]

I guess you forget to attach the datafile. Please upload at least gfs.f000 for me to take a look.
By the way, where did you download the gfs.f000 file?

 

[alainF]

Thank you for your reply.
Actually the files are too big to be attached on this forum. So I have uploaded on NextCould server, as it is written on the "Including files" item of the forum.
The files are named : gfs.f000, FILE 2022-09-11_00, met_em.d01.2022-09-11_00.nc
I don't know how to copy the corresponding links.
I hope that you will find them.

 

[Ming Chen]

I am sorry I cannot find your files in next cloud. We have WRF folder there, in which I can see "Forum upload" and "forum download". But there is no folder named "Including files".

 

[alainF]

Hello Ming Chen,
Now files gfs.f000, FILE 2022-09-11_00, met_em.d01.2022-09-11_00.nc are available on the server
Thank you for your help

 

[Ming Chen]

Hi,
I have downloaded the three files. Would you please send me your namelist.wps and namelist.input to take a look? You can simply attach them in the post. Thanks.

 

[alainF]

Here they are.

 

[alainF]

..But the date values inside the namelist files are not the right values, because these files are templates. I have a script that changes the date values.
The correct date is 2022-09-11 00Z

 

[Ming Chen]

Hi,
I have ungribbed your gfs.f000 file and run metgrid.exe to process the data. All work just fine and I cannot repeat the error you reported.
I also run REAL to process the met_em file, and REAL successfully run to the end.
Are you able to double check your issue and make sure it is related to the data and case you posted?

 

[alainF]

Did you also run wrf.exe ?
I have attached the namelist files that were really used. I forgot that I have changed target_cfl and target_hcfl values (sorry...). Could you check with these files please ?
If you need other files, now they are all gone. So we will need to wait for another crash.

 

[Ming Chen]

I cannot run wrf.exe because I only ungrib the single file of gfs.000 you attached.

Also, with limited resources I am not able to run specific cases for users.