Add EFFC variable to WRF-Chem output in morrison microphysics

[Anie_Lal]

Hi all,

I want to output EFFC (Cloud droplet effective radius) to the WRF-Chem output file.

These are the steps I followed:

1. Added 'EFFC' in the MP_MORR_TWO_MOMENT subroutine in phys/module_mp_morr_two_moment.F file similar to rainprod/evapprod in line 575.
2. Added 'EFFC' in line 140 and added 'REAL, DIMENSION(ims:ime, kms:kme, jms:jme), INTENT(OUT):: EFFC' in line 580 in phys/module_microphysics_driver.F.
3. Added EFFC=grid%EFFC in line 3830 in dyn_em/solve_em.F.
4. Added
state real EFFC ikj misc 1 - \
h "EFFC" "Cloud droplet effective radius" "um"
in Registry/Registry.EM_COMMON file.

Recompiled by ./compile em_real.

Now I am getting EFFC in the model output, but the spatial pattern are not looking good. They are coming like few lines on the boundaries of the domain.
Please find the figures attached.

The run was only for a day (25 timesteps) to check if EFFC was generated in the output.

What could be the issue? Am I doing the code changes correctly?

Looking forward to any kind of help.

Thank you
Anie

 

[Anie_Lal]

Hi all,

I tried with a 6 day simulation to see if enough spin up improves the spatial EFFC. But the pattern is same.

Looking forward to any kind of help.

Thank you

 

[HollyChen]

excuse me,
do you know the answer of this question?
could you please share it ?
thank you!

 

[keke1417]

I am also interested in this question!

 

[Anie_Lal]

Hi @HollyChen @keke1417,

I regret to inform you that I was unable to resolve the issue.

I had reached out directly to Dr. Hugh Morrison, and I am sharing his comments here (please find attached).

I would greatly appreciate it if we could work together to find a solution. If anyone comes across any insights or leads, please feel free to share them here.

Thank you for your support.

 

[HollyChen]

Thank you so much for your response.
After some attempts, I found that we need to rename EFFC in MP_MORR_TWO_MOMENT, such as EFF_C
Then define its type: REAL, DIMENSION(ims:ime, kms:kme, jms:jme), INTENT(INOUT):: EFF_C
Finally add this code around Line 844 to transfer 1D arrays back into 3D arrays: EFF_C(i,k,j) = EFFC1D(k)
The other steps remain the same with you, except for changing the variable name.
Finally, we can obtain the spatial distribution of EFFC.

But I have other questions.
I also used NSSL_2mom scheme to get the cloud droplet effective radius but found that the spatial distribution of those two was different.
Based on your response, I guess that the computation method of NSSL_2mom is inconsistent with Morrison_2mom.