WRF and WRFDA: ---- Error : Cannot use moist theta option with old data

[aaronhill]

I am attempting to run WRF with a perturbed wrfinput file that was generated by da_wrfvar in the WRFDA suite. Both WRF and WRFDA are version 4.6.1, downloaded and compiled from the GitHub repository, but I receive the following error when running wrf.exe:

DYNAMICS OPTION: Eulerian Mass Coordinate
alloc_space_field: domain 1 , 32737276 bytes allocated
med_initialdata_input: calling input_input
Input data is acceptable to use: wrfinput_d01
CURRENT DATE = 1999-05-03_00:00:00
SIMULATION START DATE = 1999-05-03_00:00:00
Timing for processing wrfinput file (stream 0) for domain 1: 1.29903 elapsed seconds
Max map factor in domain 1 = 1.08. Scale the dt in the model accordingly.
---- WARNING : Older v3 input data detected
-------------- FATAL CALLED ---------------
FATAL CALLED FROM FILE: <stdin> LINE: 642
---- Error : Cannot use moist theta option with old data
-------------------------------------------
taskid: 0 hostname: dec1963

I also confirmed that the input file (wrfinput_d01) used version 4.6.1 from the metadata:

:TITLE = " OUTPUT FROM WRFDA V4.6.1" ;

Other forum posts and online resources suggest you need to change two namelist options: force_use_old_data = T. or use_theta_m = 0 to set the simulation to use dry theta. Neither of these solutions seemed to work.

When I change the use_theta_m flag in the dynamics namelist, I receive the error:

DYNAMICS OPTION: Eulerian Mass Coordinate
alloc_space_field: domain 1 , 32737276 bytes allocated
med_initialdata_input: calling input_input
Input data is acceptable to use: wrfinput_d01
---- ERROR: Input file use_theta_m = 1
---- ERROR: Namelist use_theta_m = 0
---- ERROR: use_theta_m values must be consistent
CURRENT DATE = 1999-05-03_00:00:00
SIMULATION START DATE = 1999-05-03_00:00:00
NOTE: 1 namelist vs input data inconsistencies found.
-------------- FATAL CALLED ---------------
FATAL CALLED FROM FILE: <stdin> LINE: 1307
NOTE: Please check and reset these options
-------------------------------------------
taskid: 0 hostname: dec1590

I'm unsure how to resolve this issue since I am using the same version of WRFDA as WRF, so there "shouldn't" be any WRFv3 data being used (unless I am misinterpreting where the v3 data resides in the input file). Is it possible that my issue is in the met_em data since the simulation is for a case study in 1999?

I've included my two rsl.out files and the namelist also included for reference

 

[kwerner]

Hi, Apologies for the delay. Are you still experiencing this issue? Not having experience with WRFDA (or using WRFDA-generated input), I'm not certain why this is happening. Clearly because the wrfinput file was generated with a setting of "use_theta_m=1," it won't be possible to just change that setting prior to running WRF. The location in the code where the original error arises is in WRF/dyn_em/start_em.F (around line 582). I wonder if you just comment out the section that gives the error and stops the model, if that will allow the model to move forward? If you want to give that a try, comment out these lines:

CALL wrf_debug ( 0 , '---- WARNING : Older v3 input data detected' )
IF ( grid%use_theta_m .NE. 0 ) THEN
   CALL wrf_error_fatal ( '---- Error : Cannot use moist theta option with old data' )
END IF

After that, save the file and then recompile WRF (but no need to clean or reconfigure the model since you're not changing the registry file). Then run WRF to see if that helps.