Update general UPP for GTG4

[YaliMao-NOAA]

1. Change gtg_algo() interface in MDLFLD.f
2. Add CIT(:,:,:) array for In-Cloud Turbulence, ALLOCATE_ALL.f VRBLS3D_mod.f DEALLOCATE.f
3. Modify CMakeLists.txt for updated/new GTG source code files.
4. Add stub code for the new GTG source files in general UPP
5. Modify interpolation in MISCLN.f for RRFS GTG
6. Modify jobs/J_NCEPPOST to run WAFS downstream only
   Add scripts/exwafs_atmos_nceppost.sh
   Add ush/wafs_nceppost.sh
7. Add GEFS WAFS control files (.xml .txt)
8. (As an update later on) Add gtg_on option
9. Remove WAFS fields on ICAO standard atmospheric levels from post_avblflds.xml
10. Remove MDL2STD_P.f which does interpolation on ICAO standard atmospheric levels which is not ICAO required.

[WenMeng-NOAA]

@YaliMao-NOAA Do you have an UPP issue linked this PR? If not, please open a new one.

[YaliMao-NOAA]

@WenMeng-NOAA I opened an issue #763

[WenMeng-NOAA]

@YaliMao-NOAA Please add a change log with your name in the fortran code files where if change log sessions are available.

[YaliMao-NOAA]

@YaliMao-NOAA Please add a change log with your name in the fortran code files where if change log sessions are available.

You mean the code change history?

[WenMeng-NOAA]

@YaliMao-NOAA See the example at
https://github.com/NOAA-EMC/UPP/blob/develop/sorc/ncep_post.fd/INITPOST_NETCDF.f#L42

[YaliMao-NOAA]

@WenMeng-NOAA I added change logs to the Fortran source code files.
Also I added gtg_on option to the namelist

[WenMeng-NOAA]

@YaliMao-NOAA My UPP with gtg code build process failed as:

CMake Error at sorc/ncep_post.fd/CMakeLists.txt:194 (add_library):
  Cannot find source file:

    post_gtg.fd/gtg_cdf_mod.F90

  Tried extensions .c .C .c++ .cc .cpp .cxx .cu .mpp .m .M .mm .h .hh .h++
  .hm .hpp .hxx .in .txx .f .F .for .f77 .f90 .f95 .f03 .ispc


CMake Error at sorc/ncep_post.fd/CMakeLists.txt:194 (add_library):
  No SOURCES given to target: upp


CMake Generate step failed.  Build files cannot be regenerated correctly

There is no file "gtg_cdf_mod.F90" from gtg code checkout with the revision d0ad324.

[YaliMao-NOAA]

Sorry, I should have update UPP_GTG develop branch. I am requesting a viewer to approve a merge pull request of GTG code.

[WenMeng-NOAA]

@YaliMao-NOAA Please verify my UPP standalone test results at /u/wen.meng/ptmp/gfs.2019083000 on Cactus and let me know if you saw issues.

[WenMeng-NOAA]

@YaliMao-NOAA The UPP RT failed in all test case as:

+ mpiexec -l -n 96 -ppn 48 /u/wen.meng/ptmp/yali/UPP/exec/upp.x
nid003523.cactus.wcoss2.ncep.noaa.gov 2: forrtl: severe (174): SIGSEGV, segmentation fault occurred
Image              PC                Routine            Line        Source          
upp.x              0000000000F4F36A  Unknown               Unknown  Unknown
libpthread-2.31.s  0000150A256408C0  Unknown               Unknown  Unknown
upp.x              0000000000861FAA  mdl2p_                    373  MDL2P.f
libiomp5.so        0000150A26A6D3F3  __kmp_invoke_micr     Unknown  Unknown
libiomp5.so        0000150A269F1937  __kmp_fork_call       Unknown  Unknown
libiomp5.so        0000150A269B5533  __kmpc_fork_call      Unknown  Unknown
upp.x              000000000081986E  mdl2p_                    322  MDL2P.f

I guess you need to add new option "gtg_on" in these kind of routine for skipping gtg related computations.

[YaliMao-NOAA]

The test result looks good to me /u/wen.meng/ptmp/gfs.2019083000

As to the failure of SIGSEGV on your RT, you are right, gtg_on needs to be turned on by 'gtg_on=.true.' Do you think I should include 'gtg_on' option to avoid this kind of errors in mdl2p?

[WenMeng-NOAA]

The test result looks good to me /u/wen.meng/ptmp/gfs.2019083000

As to the failure of SIGSEGV on your RT, you are right, gtg_on needs to be turned on by 'gtg_on=.true.' Do you think I should include 'gtg_on' option to avoid this kind of errors in mdl2p?

@YaliMao-NOAA Yes, you might also check other gtg related routines.

[YaliMao-NOAA]

I modified the code, and added a new issue related to this PR. #764

[WenMeng-NOAA]

I modified the code, and added a new issue related to this PR. #764

@YaliMao-NOAA Can you note your new changes at the beginning session of this PR, such as

1. Change gtg_algo() interface in MDLFLD.f
2. Add CIT(:,:,:) array for In-Cloud Turbulence, ALLOCATE_ALL.f VRBLS3D_mod.f DEALLOCATE.f
   ...

[WenMeng-NOAA]

@YaliMao-NOAA You need to tune your branch further. Here are the latest errors from the UPP RT for nmmb:

upp.x              00000000008626A8  mdl2p_                    387  MDL2P.f
libiomp5.so        000014EA165423F3  __kmp_invoke_micr     Unknown  Unknown
libiomp5.so        000014EA164C6937  __kmp_fork_call       Unknown  Unknown
libiomp5.so        000014EA1648A533  __kmpc_fork_call      Unknown  Unknown
upp.x              0000000000819CBC  mdl2p_                    335  MDL2P.f
upp.x              0000000000612D37  process_                   78  PROCESS.f
upp.x              000000000056D7FC  MAIN__                    701  WRFPOST.f

The whole runtime log file can be found /u/wen.meng/noscrub/ncep_post/post_regression_test_new/jobs/out.nmmb.test.
You might run the UPP RT by yourself for debugging. The UPP RT procedure can be found at
https://docs.google.com/document/d/1gtk1DBxEa9GsaWbxrxtB4Ot7U_lNIHjLXaU2F6FyD7M/edit

[YaliMao-NOAA]

@WenMeng-NOAA It turns out gtg_interpolation was not initialized in MDL2P.f, I fixed it. Thank you for recommendation to do the regression test by myself. All regression tests passed. FYI, for rtma test, should it be 'cd $workdir' instead of '$homedir/jobs'?

[WenMeng-NOAA]

@YaliMao-NOAA With new version of gtg4 implemented, do you think the gtg files gtg.config.gfs gtg_imprintings.txt under UPP/parm should be removed?

[WenMeng-NOAA]

The UPP RT is completed on WCOSS2 with no changed results.

[WenMeng-NOAA]

@FernandoAndrade-NOAA You may start the UPP RT on Hera and Orion. There will no baseline update with this PR.

[FernandoAndrade-NOAA]

Apologies for the delay, Orion UPP RTs have completed successfully with no changes to results. Hera should finish momentarily.

[FernandoAndrade-NOAA]

Hera UPP RTs have completed as well with no changes to results.

[WenMeng-NOAA]

@gspetro-NOAA Please be aware of the updates in parm/post_avblfld.xml in this PR for your new release of the UPP user guide later.

[WenMeng-NOAA]

This PR is ready for merging.