Scheduled Downtime
On Friday 21 April 2023 @ 5pm MT, this website will be down for maintenance and expected to return online the morning of 24 April 2023 at the latest

RH field from met_em files does not transfer to QVAPOR or Q2 in wrfinput

F

frediani

New member
Hi,

I have prepared WPS files where the 3D RH field appears correct. However, when I run real.exe, the vapor fields in the initial conditions (QVAPOR and Q2) contain only missing values, while the corresponding fields in the boundary conditions (e.g. QVAPOR_B* in wrfbdy) look fine.

Is there a way to make real.exe process RH from metgrid into initial conditions? Is there a namelist option that I can include or remove to make it work?

I tried a few different versions of real.exe (4.6.0, 4.5.0, 4.3.) with no success. If it's helpful, my files are here: /glade/work/frediani/Projects/UFS-fire/wrf-test-case/wrf-LastChance-3km

Thanks
Maria
 
Maria,
I can repeat your issue, but I am not sure yet what is wrong.
I will continue to look at this problem and get back to you once I find something.
It may take some time and thank you for your patience.
 
Hi Ming, the 3D atmospheric fields are from RAP. I had to use GFS for the soil variables (only for the initial conditions).
The RAP files are here: /glade/campaign/ral/wsap/frediani/LastChance/NCEI-HAS/
My WPS folder is this:
/glade/work/frediani/Projects/UFS-fire/wrf-test-case/wps-LastChance-3km

Note: I have seen this happen before. The difference was that I only had RH in my Vtable, which caused WPS to produce RH fields instead of SPECHUMD. I fixed it by adding SPECHUMD to the Vtable, which led WPS to produce a SPECHUMD field (instead of RH), so real.exe was then able to create the vapor fields. This old simulation was using HRRR inputs.

In my current Vtable, I have both, RH and SPECHUMD, but WPS is producing RH. I've seen a couple of github threads that led me to believe real.exe should be able to handle either (Use Qv or specific humidity in interpolation if they are available by weiwangncar · Pull Request #1959 · wrf-model/WRF, Enable initialization of ARW using QV (mixing ratio) by mgduda · Pull Request #163 · wrf-model/WRF). Maybe I can try to use this new option use_sh_qv.
 
I tried setting this new option use_sh_qv = True in v4.6.0 but it didn't work. I still don't have QVAPOR in wrfinput and in addition, it disabled QVAPOR in wrfbdy.
 
Maria.
I am working on the issue you reported. Would you please change the permission of RAP data file saved in

/glade/p/ral/wsap/frediani/LastChance/NCEI-HAS/

Thank you.
 
Hi Ming, the files are in campaign now: /glade/campaign/ral/wsap/frediani/LastChance/NCEI-HAS/
Please let me know if you can't access them.
 
Sorry about that Ming. I tried it with a different command now and also made a copy of the file here: /glade/derecho/scratch/frediani/for_ming
 
Hi Maria,

I guess I found a solution to this issue:

Please don't include specific humidity in your intermediate files, i.e, delete the lines in Vtable for SPECHUMD, then rerun ungrib.exe. In this case, REAL will derive QVAPOR based on RH.

I have the sample Vtable and RAP files saved at

/glade/derecho/scratch/chenming/WPS-Forum,

and the wrfinput file saved at

/glade/derecho/scratch/chenming/WRF-Forum

Please take a look and let me know if you have questions.
 
Hi Ming, thank you so much for helping me make this work!
I did an ncdump -h of both met_em files and I can see that both of them have the 3D RH field, but one of them also has a 2D SPECHUMD, which is likely the cause of the issue in real.exe. This seems to be the opposite of what I encountered with HRRR a few years ago! I guess the answer here is to make sure the vapor variables are consistent in the 2D and 3D, i.e. either SH or RH, not both.
 
Maria,

I am not sure why you need to have separate 2D and 3D fields? In my test, I simply put them in the same single file, --- both ungrib and metgrid work fine in this way.

My understanding is that we need to have RH in the intermediate files. REAL doesn't process spechumd to produce qvapor, which is why you see qvapor == 0 everywhere.

Please let know if you have more questions.

Ming
 
You must log in or register to reply here.
Top