Issues outputting microphysics variables to wrfout...

[nluchett]

Hi all, I am trying to output snow and graupel melt, as well as rain evaporation from the Thompson Microphysics scheme and write it to the wrfout files.. I have been unsuccessful so I was wondering if anyone can look at my source code mods and perhaps help me de -bug? I've attached several screenshots of my mods as well as an error.txt containing the compile.log errors.

Based on my limited experience working within the Thompson mp scheme source code, I found the prs_sml (snow melt), prr_gml (graupe melt), and prv_rev (rain evaporation) to be the variables I needed to write out to wrfoutput (not entirely confident in the rain evaporation term... the other two I think are for sure correct). I first created new entires in the Registry.EM_COMMON (NTL_Registry_EM_edits.png). I think I did this correctly, however I wasn't confident in the IO field. Should this be "h" for history?

Next, I edited the Solve_em.F (NTL_Solve_em_edits.png) which I think is correct... I then made edits to the Thompson scheme itself by first editing the mp_gt_driver routine to include the variables from the Registry I created (NTL_Thompson_mp_gt_driver_edits.png), defined the variables (NTL_thompson_define_variables_out_of_subroutine.png), initialized the 3d variables (NTL_thompson_intitalize_3D_mp_variables_edits.png), called the 1D variables in the call mp_thompson routine (NTL_thompson_call_mp_thompson_edits.png) and finally converted some units in the source/sink terms (NTL_thompson_convert_sink_units.png).

Lastly, I edited the module_microphysics_driver.F to add the new 3D variables (NTL_microphysics_driver_add_3d_var.png), define the variables (NTL_microphysics_driver_define_var - Not too sure if this one is necessary or not....), and finally to modify the call mp_gt_driver (NTL_microphysics_driver_call_mp_gt_driver.png).

I've attached an error.txt which shows the compile.log errors I'm receiving. It appears the errors are coming from the microphysics_driver.F module... "../phys/module_microphysics_driver.f90(747): error #6631: A non-optional actual argument must be present when invoking a procedure with an explicit interface. [SNOWML]"

I'm really not sure how to interpret this error as this is all really new to me. Does anyone have any suggestions?

Thanks again for the help!

Take care,

Nick Luchetti

 

[nluchett]

Bumping to add more screen shots due to 5 attachment limit....

 

[nluchett]

Bump again to add a 1 more screen shots of mods..

 

[kwerner]

Hi Nick,
Instead of screen shots of parts of the files, can you simply attach the code files that you modified (e.g., attach module_microphysics_driver.F), along with the full compile log? Please also let me know which version of the model you are making your modifications from so that I can compare to the default.

Thanks!

 

[nluchett]

Hi Kwerner,
Absolutely. See attached for each of the files I modified (Solve_em.F, module_microphsyics_driver.F, thompson_mp, and Registry.EM_Common), as well as the entire compile.log. I'm modifying WRF version 4.0.3.

 

[nluchett]

Here is compile.log file since I can only attach 5 per post. You can search "nick luchetti" in each of these files to find where I made my edits.

Thank you in advance!

Nick

 

[kwerner]

Thanks for sending those. That was helpful. There are 3 things that need to be fixed in your files:

module_microphysics_driver.F
line 488:

REAL, DIMENSION(ims:ie, kms:kme, jms:jme), INTENT(OUT):: snowml, graupml, rain_evp

There is an 'm' missing in the dimension - it should be

ims:ime

You need to add your 3 new variables in the both calls to "mp_gt_driver", so you they should also be added around line 30:

snowml=snowml,                      &
                     graupml=graupml,                    &
                     rain_evp=rain_evp,                  &

solve_em.F
There is a mis-spelling for "rain_evp"
line 3731:

  &        , rain_evp=grid%rain_evap                                   &

You will need to remove the 'a' from the right side.

After making those changes, you SHOULD be able to build.

 

[qinjm18]

Hi Nick,
Have you solved the problem？
There was a problem when compiling.
Please check out compile.log in the attachment.

Thanks!