fix inconsistent types related with noahmp and log functions

[uturuncoglu]

This PR aims to fix couple of minor issue related to the inconsistent data types in the CCPP/physics. If user wants to compile NoahMP and related files outside of the CCPP package, the compiler triggers errors since some of the subroutines uses real data type rather than real(kind=kind_phys), which kind_phys is set to R8 by default under CCPP.

This is related with the following issue: #750

[climbfuji]

Thanks for these bug fixes. I think that they won't impact the UFS at all, because the files you are changing are compiled with default double precision reals in the UFS.

[climbfuji]

@uturuncoglu I tested these changes, and I get compile errors (with Intel, haven't triedd GNU yet):

ifort: command line remark #10120: overriding '-O0' with '-O1'
ifort: warning #10182: disabling optimization; runtime debug checks enabled
/glade/work/heinzell/fv3/ufs-weather-model/ufs-weather-model-develop-20210929-cleanup-ccpp/intel/FV3/ccpp/physics/physics/sflx.f(3997): error #5082: Syntax error, found '&' when expecting one of: :: ) , : * <END-OF-STATEMENT> ; . (/ + - ] /) ' ** / // > PRIVATE ...
            df = log( (psis*gs2/lsubf) * ( (1.0 + ck*swl)**2.0 )      &
----------------------------------------------------------------------^
compilation aborted for /glade/work/heinzell/fv3/ufs-weather-model/ufs-weather-model-develop-20210929-cleanup-ccpp/intel/FV3/ccpp/physics/physics/sflx.f (code 1)
make[2]: *** [FV3/ccpp/physics/CMakeFiles/ccpp_physics.dir/physics/sflx.f.o] Error 1
make[2]: *** Waiting for unfinished jobs....
make[1]: *** [FV3/ccpp/physics/CMakeFiles/ccpp_physics.dir/all] Error 2
make: *** [all] Error 2

[climbfuji]

Please fix the compile errors.

[uturuncoglu]

@climbfuji sure. it is probably due to the line continuation. I need to move it to the next line.

[uturuncoglu]

@climbfuji btw, I am also using intel but compiling those sources did not reveal any issue. ccpp might use different flags than I am using.

[climbfuji]

@climbfuji btw, I am also using intel but compiling those sources did not reveal any issue. ccpp might use different flags than I am using.

Good point. I think there are some hardcoded form flags for certain files. Generally, .f/.F files should be fixed form, .f90/.F90 files free form.

[uturuncoglu]

@climbfuji I think I fixed it it was spacing issue. The & need to be at column 73. How can I test it. I think If I compile the UFS, it will show up. Right?

[climbfuji]

Thanks, Ufuk! The code now compiles with both 32bit and 64bit fv3atm (physics is always 64bit in the UFS). I pulled your PR into #752. Please keep your PR open, it should be flagged as merged (and you of course get all the credit for it) when #752 gets merged (a week or longer from now). Thanks again for catching and fixing those problems.

[uturuncoglu]

@climbfuji I confirmed that control_noahmp RT test passes without any issue. Thanks for catching that.