Abnormal 2m temperature pattern with WRF V4.4.1

[alainF]

Hello Everyone,
After the WRF upgrade from V4.3.1 to V4.4.1, I get an unrealistic cold temperature pattern when heavy rains occur. The pattern is shown on the attached document.
I have put on the cloud a complete set of files called "Saved_4.4.1-10-20-2022.tar.bz2" including namelists and the input wrf files generated by real.exe.
In some region in blue (see document), the temperature drops as low as 2°C although the real temperature was 16°C to 20°C !
I never saw this before.
I have used the same namelist for the previous 4.3.1 version and it went well.
As you will see in the namelist, the land surface model is NOAH-MP, radiation = CAM, and MP = 28.
I have also put the grib results named atlantic_d01 and atlantic_d03.

Thank you for your help.

 

[alainF]

For information, I have tried with use_aero_icbc = .false. and it gives the same results
I also tried RRTMG for short and long wave model but same results.

 

[Ming Chen]

NoahMP has been changed a lot in WRFV4.4. I am not sure whether these changes have resulted in 2-m temperature change.
Please confirm that WRFv4.3.1 and WRFv4.4.1 are run with the same namelist options and driven by the same input data.

 

[alainF]

Hello,
Yes the namelists and the grib input data were identical while running both versions, 4.4.1 and 4.3.1.