Bug fix for z0t

[ChunxiZhang-NOAA]

This is a corresponding PR to ccpp-physics PR#23. It fixed a bug related to the definition of z0t. A new variable ztmax is added.

[uturuncoglu]

@ChunxiZhang-NOAA It seems that action is failing because GitHub Action runner is using new version of OS and spack package installer is failing. I am looking into it. I could also test your changes manually under UFS. BTW, it might be nice to add this new variable to the history file. You just need it to add just after this line,

noahmp/drivers/nuopc/lnd_comp_io.F90

Line 2245 in 0729e88

call fld_add("pblh" ,"height of pbl" ,"m" ,ptr2d=ptr2d, v1r8=noahmp%model%pblh)

. Of course, this also changes the BLs. So, if you already tested and there is no BL change then I could do it later.

[ChunxiZhang-NOAA]

@uturuncoglu I added the variable to the history file as you suggested. I tested it in the UFS, and it changed the BLs.

[ChunxiZhang-NOAA]

@uturuncoglu @barlage Please review this PR ASAP. This PR is scheduled to be committed today.

[barlage]

These changes look good to me.

[ChunxiZhang-NOAA]

@barlage Thank you!

[uturuncoglu]

@ChunxiZhang-NOAA I am getting following errors from container with GNU compiler,

/test/app/noahmp/src/module_sf_noahmplsm.f90:1:0:

    1 | #define CCPP
      | 
Warning: "CCPP" redefined
<command-line>:1:0:

note: this is the location of the previous definition
/test/app/noahmp/src/module_sf_noahmplsm.f90:2137:132:

 2137 |          l_rsurf = (-zsoil(1)) * ( exp ( (1.0 - min(1.0,sh2o(1)/parameters%smcmax(1))) ** parameters%rsurf_exp ) - 1.0 ) / ( 2.71828 - 1.0 )
      |                                                                                                                                    1
Error: Line truncated at (1) [-Werror=line-truncation]
/test/app/noahmp/src/module_sf_noahmplsm.f90:2137:132:

 2137 |          l_rsurf = (-zsoil(1)) * ( exp ( (1.0 - min(1.0,sh2o(1)/parameters%smcmax(1))) ** parameters%rsurf_exp ) - 1.0 ) / ( 2.71828 - 1.0 )
      |                                                                                                                                    1
Error: Expected a right parenthesis in expression at (1)
/test/app/noahmp/src/module_sf_noahmplsm.f90:2138:132:

 2138 |          d_rsurf = 2.2e-5 * parameters%smcmax(1) * parameters%smcmax(1) * ( 1.0 - parameters%smcwlt(1) / parameters%smcmax(1) ) ** (2.0+3.0/parameters%bexp(1))
      |                                                                                                                                    1
Error: Line truncated at (1) [-Werror=line-truncation]
/test/app/noahmp/src/module_sf_noahmplsm.f90:2138:132:

 2138 |          d_rsurf = 2.2e-5 * parameters%smcmax(1) * parameters%smcmax(1) * ( 1.0 - parameters%smcwlt(1) / parameters%smcmax(1) ) ** (2.0+3.0/parameters%bexp(1))
      |                                                                                                                                    1
Error: Syntax error in expression at (1)
/test/app/noahmp/src/module_sf_noahmplsm.f90:8869:132:

 8869 |                    smceqdeep = parameters%smcmax(nsoil) * ( -parameters%psisat(nsoil) / ( -parameters%psisat(nsoil) - dzsnso(nsoil) ) ) ** (1./parameters%bexp(nsoil))
      |                                                                                                                                    1
Error: Line truncated at (1) [-Werror=line-truncation]
/test/app/noahmp/src/module_sf_noahmplsm.f90:8869:132:

 8869 |                    smceqdeep = parameters%smcmax(nsoil) * ( -parameters%psisat(nsoil) / ( -parameters%psisat(nsoil) - dzsnso(nsoil) ) ) ** (1./parameters%bexp(nsoil))
      |                                                                                                                                    1
Error: Expected a right parenthesis in expression at (1)
/test/app/noahmp/src/module_sf_noahmplsm.f90:8881:132:

 8881 |            smceqdeep = parameters%smcmax(nsoil) * ( -parameters%psisat(nsoil) / ( -parameters%psisat(nsoil) - dzsnso(nsoil) ) ) ** (1./parameters%bexp(nsoil))
      |                                                                                                                                    1
Error: Line truncated at (1) [-Werror=line-truncation]
/test/app/noahmp/src/module_sf_noahmplsm.f90:8881:132:

 8881 |            smceqdeep = parameters%smcmax(nsoil) * ( -parameters%psisat(nsoil) / ( -parameters%psisat(nsoil) - dzsnso(nsoil) ) ) ** (1./parameters%bexp(nsoil))
      |                                                                                                                                    1
Error: Syntax error in expression at (1)
f951: some warnings being treated as errors
make[2]: *** [CMakeFiles/noahmp.dir/build.make:283: CMakeFiles/noahmp.dir/src/module_sf_noahmplsm.f90.o] Error 1
make[1]: *** [CMakeFiles/Makefile2:83: CMakeFiles/noahmp.dir/all] Error 2
make: *** [Makefile:136: all] Error 2

[barlage]

Does GNU have a column 132 truncation requirement?

[ChunxiZhang-NOAA]

@uturuncoglu It looks the issue is not related to this PR. The error message shows the truncation error. If can be solved with compiling options.

[ChunxiZhang-NOAA]

Fortran90+ has 132 line length limit. Intel compiler usually doesn't strictly force it. Looks like gnu does. Maybe it will help by adding this compiler option: -ffree-line-length-none

[barlage]

In a future update, we can break this line.

[ChunxiZhang-NOAA]

@uturuncoglu Could you try the compiler option -ffree-line-length-none? I believe it will resolve the issue with this option.

[uturuncoglu]

@ChunxiZhang-NOAA I compiled with -DCMAKE_Fortran_FLAGS="-ffree-line-length-none -ffixed-line-length-none" and those error disappears (I think if you could you could fix those lines and we don't need to worry about the compiler flags, this also requires some changes in GitHub action that I fixed later if we need to use those flags) but following came up

Scanning dependencies of target noahmp
[  5%] Building Fortran object CMakeFiles/noahmp.dir/drivers/ccpp/machine.F.o
/test/app/noahmp/drivers/ccpp/machine.F:9:47:

    9 |       integer, parameter :: kind_sngl_prec = 4                          &
      |                                               1
Error: Syntax error in data declaration at (1)
/test/app/noahmp/drivers/ccpp/machine.F:24:53:

   24 |       integer, parameter :: kind_phys = kind_dbl_prec
      |                                                     1
Error: Symbol 'kind_dbl_prec' at (1) has no IMPLICIT type
/test/app/noahmp/drivers/ccpp/machine.F:29:48:

   29 |       integer, parameter :: kind_io8 = kind_phys                        &! Note kind_io8 is not always 8 bytes
      |                                                1
Error: Symbol 'kind_phys' at (1) has no IMPLICIT type; did you mean 'kind_io8'?
/test/app/noahmp/drivers/ccpp/machine.F:35:53:

   35 |       integer, parameter :: kind_dyn  = kind_dbl_prec
      |                                                     1
Error: Symbol 'kind_dbl_prec' at (1) has no IMPLICIT type
/test/app/noahmp/drivers/ccpp/machine.F:39:25:

   39 |       real(kind=kind_phys), parameter :: mprec = 1.e-12           ! machine precision to restrict dep
      |                         1
Error: Symbol 'kind_phys' at (1) has no IMPLICIT type
/test/app/noahmp/drivers/ccpp/machine.F:40:25:

   40 |       real(kind=kind_phys), parameter :: grib_undef = 9.99e20     ! grib undefine value
      |                         1
Error: Symbol 'kind_phys' at (1) has no IMPLICIT type
make[2]: *** [CMakeFiles/noahmp.dir/build.make:205: CMakeFiles/noahmp.dir/drivers/ccpp/machine.F.o] Error 1
make[1]: *** [CMakeFiles/Makefile2:83: CMakeFiles/noahmp.dir/all] Error 2
make: *** [Makefile:136: all] Error 2

[barlage]

@uturuncoglu any reason why these are showing up now? as far as I know, none of this code has been changed in this PR.

[uturuncoglu]

@barlage i am not sure. Maybe adding -DCMAKE_Fortran_FLAGS="-ffree-line-length-none -ffixed-line-length-none" to compile triggered these. I think it would be nice to fix long lines and try again without those extra flags.

[uturuncoglu]

@ChunxiZhang-NOAA @barlage Okay. I fixed the long lines and compile without extra flags. It builds without any issue and also runs under my Docker container. So, we just need to fix those long lines and split them using &.

[barlage]

I guess the question also applies to the original error since this long line is in the original version too.

[ChunxiZhang-NOAA]

@uturuncoglu Did you try to only use -ffree-line-length-none?

[uturuncoglu]

@ChunxiZhang-NOAA it seems it is building fine with only this option.

[ChunxiZhang-NOAA]

@uturuncoglu Good to know. I think it is because -ffixed-line-length-none makes the "&" character at column 73 legal but "&" in the context is illegal in Fortran77.

[uturuncoglu]

@ChunxiZhang-NOAA Thanks. I think this is fine. I could make changes in the action once I fix the issue with the updated runner.

[uturuncoglu]

@ChunxiZhang-NOAA I could merge it now, if it is fine for you.

[ChunxiZhang-NOAA]

@uturuncoglu The UFS commit process for PR#1394 has started one hour ago. It would be great if you can create a new PR for the changes in the actions.

[uturuncoglu]

@ChunxiZhang-NOAA Yes. No worries about action. I could fixed later.

[ChunxiZhang-NOAA]

@uturuncoglu Ok, thank you!

[jkbk2004]

All tests are done on ufs-wm pr #1493 side. Can we start merging in this pr?