Update of GF aerosol treatment and tunings

[hannahcbarnes]

The treatment of aerosols in the Grell-Freitas Convective Parameterization is updated. These updates include modifying aerosol scavenging based on a method that has been used in CIMP5 and fixing how cloud condensation nuclei is calculated from aerosol optical depth.

There are also a few bug fixes and further tuning of the Grell-Freitas Convective Parameterization.

[DomHeinzeller]

Please provide a description for what this PR does. Also, I think you can remove the newly added variable trash from cu_gf_driver.F90.

[DomHeinzeller]

I don't see trash being used.

[hannahcbarnes]

I just updated the PR without that variable, thanks

[hannahcbarnes]

@DomHeinzeller How do I add a description to the commit?

[DomHeinzeller]

@DomHeinzeller How do I add a description to the commit?

In your PR, click on the three dots in the top-right corner of the first comment box, then edit.

[hannahcbarnes]

@DomHeinzeller Thanks. I updated the description. Let me know if it needs to be further modified.

[climbfuji]

@DomHeinzeller Thanks. I updated the description. Let me know if it needs to be further modified.

That's great, thanks a lot.

[DomHeinzeller]

@hannahcbarnes something is not quite right with your code changes. The following debug tests all crash:

fv3_gsd_debug 015 failed
fv3_gsd_debug 015 failed in run_test
fv3_gsd_mynnsfc_debug 017 failed
fv3_gsd_mynnsfc_debug 017 failed in run_test
fv3_gsd_noah_mynnsfc_debug 018 failed
fv3_gsd_noah_mynnsfc_debug 018 failed in run_test
fv3_gsd_dtend_debug 016 failed
fv3_gsd_dtend_debug 016 failed in run_test

The error for fv3_gsd_debug is:

  0:  af create fcst fieldbundle, name=sfc_bilinearrc=           0
  0:  in fcst,init total time:    30.2752709388733
  5: forrtl: error (73): floating divide by zero
  5: Image              PC                Routine            Line        Source
  5: fv3.exe            0000000008F7E47E  Unknown               Unknown  Unknown
  5: libpthread-2.17.s  00002B6561C38630  Unknown               Unknown  Unknown
  5: fv3.exe            0000000008561E26  cu_gf_deep_mp_cup        4143  cu_gf_deep.F90
  5: fv3.exe            00000000083E688D  cu_gf_deep_mp_cu_         955  cu_gf_deep.F90
  5: fv3.exe            00000000072B2D64  cu_gf_driver_mp_c         657  cu_gf_driver.F90
  5: fv3.exe            0000000006EE2479  ccpp_fv3_gsd_v0_p        2458  ccpp_FV3_GSD_v0_physics_cap.F90
  5: fv3.exe            0000000006927B55  ccpp_static_api_m         464  ccpp_static_api.F90
  5: fv3.exe            000000000691F1DC  ccpp_driver_mp_cc         188  CCPP_driver.F90
  5: libiomp5.so        00002B6560040A43  __kmp_invoke_micr     Unknown  Unknown
  5: libiomp5.so        00002B65600042C6  __kmp_fork_call       Unknown  Unknown
  5: libiomp5.so        00002B655FFC3BB0  __kmpc_fork_call      Unknown  Unknown
  5: fv3.exe            000000000691CEB2  ccpp_driver_mp_cc         169  CCPP_driver.F90
  5: fv3.exe            000000000340E6BD  atmos_model_mod_m         341  atmos_model.F90
  5: fv3.exe            0000000002DE7A79  module_fcst_grid_         751  module_fcst_grid_comp.F90
  5: fv3.exe            0000000001317A5F  _ZN5ESMCI6FTable1        2036  ESMCI_FTable.C
  5: fv3.exe            0000000001313420  ESMCI_FTableCallE         765  ESMCI_FTable.C
  5: fv3.exe            0000000000F0EAD6  _ZN5ESMCI3VMK5ent        2204  ESMCI_VMKernel.C
  5: fv3.exe            0000000000D146C4  _ZN5ESMCI2VM5ente        1211  ESMCI_VM.C
  5: fv3.exe            0000000001313B09  c_esmc_ftablecall         922  ESMCI_FTable.C
  5: fv3.exe            0000000000E6B0D0  esmf_compmod_mp_e        1214  ESMF_Comp.F90

Note that the only remaining debug test fv3_gsd_sar_debug passes. This test uses MYNN EDMF, but no cumulus parameterization.

[hannahcbarnes]

@DomHeinzeller Hmm, that confuses me. I am using this code without a problem. Can you tell me more about what these debug scripts are testing for. Do you have any other output / log, I can't seem to make much out of the error that is listed.

[hannahcbarnes]

@DomHeinzeller I see that an update to the GF code has changed how the variable cactiv works in the code (it is now linked to another variable conv_act). I will try to apply these changes to the variable I added (cactiv_m) and see if that helps.

[DomHeinzeller]

You need to compile the code in DEBUG mode (add DEBUG=Y to the MAKEOPT string) to get this error.

…

On Jul 13, 2021, at 7:41 AM, Hannah C Barnes ***@***.***> wrote: @DomHeinzeller <https://github.com/DomHeinzeller> I see that an update to the GF code has changed how the variable cactiv works in the code (it is now linked to another variable conv_act). I will try to apply these changes to the variable I added (cactiv_m) and see if that helps. — You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub <#97 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AN7FF5BLDFLE6CQET44MDJTTXQ7APANCNFSM5ADNS2EQ>.

[hannahcbarnes]

Thanks, Dom. I realized that my runs with this code are also not working, even though it compiles in non-debug mode. So far my idea about. "cactiv_m" is not the source of the issue. (Code still crashes when I turn that variable off)

[hannahcbarnes]

@DomHeinzeller Is this the correct command.

./compile.sh "$target" "DEBUG=Y APP=ATM CCPP=Y 32BIT=Y SUITES=FV3_GFS_v15,FV3_GFS_v16,FV3_GFS_v16_GF,FV3_GSD_noah,FV3_GSD_v0" 2 NO NO

I just ran this and did not get any compile errors

[climbfuji]

@DomHeinzeller Is this the correct command.

./compile.sh "$target" "DEBUG=Y APP=ATM CCPP=Y 32BIT=Y SUITES=FV3_GFS_v15,FV3_GFS_v16,FV3_GFS_v16_GF,FV3_GSD_noah,FV3_GSD_v0" 2 NO NO

I just ran this and did not get any compile errors

You don't need CCPP=Y anymore, but it doesn't hurt. You need to run the code now and see what you get.

[hannahcbarnes]

@DomHeinzeller Do you think there is any possibility that this problem is not linked to my commit? I just compiled and ran a version of the code where all the code I changed was completely replaced with the previous version of the code, and it still crashed with the same error that I have been getting.

[hannahcbarnes]

I have been using this code multiple times with a Mar gsl/develop code base and Judy is using the code with a June EMC code base and have not run into this problem.