Replace "derf()" with standard "erf()" in SBM MP scheme

[davegill]

TYPE: bug fix

KEYWORDS: derf, SBM

SOURCE: Nicolas Baldeck (OpenMeteoData)

DESCRIPTION OF CHANGES:
Problem:
Using derf for a double precision error function is not standard, and causes an error
with PGI 19.10:

441044-PGF90-S-0038-Symbol, derf, has not been explicitly declared (module_mp_SBM_polar_radar.f90)
441136- 0 inform, 0 warnings, 1 severes, 0 fatal for calc_dc_wet_snow

Solution:
The Fortran standard allows the TKR (type, kind, rank) of an argument to determine which
intrinsic function is being used. The erf function is part of the Fortran 2008 standard, and
has been previously available for GNU, Intel, and PGI.

The scheme's original developer has approved these mods.

ISSUE:
Fixes #1245 "PGI complains about Derf symbol in module_mp_SBM_polar_radar.F"

LIST OF MODIFIED FILES:
modified: phys/module_mp_SBM_polar_radar.F

TESTS CONDUCTED:

1. Before mods, the PGI compiler complained. After fix, code builds and runs with PGI 19.
2. With GNU 9 and Intel 19 compilers, code builds and runs as before. It is possible that older (non 2008 compliant)
   compilers will fail, but that standard is 12 years old, so not too much sympathy.
3. Status of jenkins is all pass.

[davegill]

@JS-WRF-SBM @dudhia @weiwangncar
Koby,
Please review this modification

[davegill]

@nbald
Nikolas,
Does this fix the problem?

[davegill]

Jenkins results:

Please find result of the WRF regression test cases in the attachment. This build is for Commit ID: cd0152159adcf3916aa708074baa77a7c9885e36, requested by: davegill for PR: https://github.com/wrf-model/WRF/pull/1246. For any query please send e-mail to David Gill.

    Test Type              | Expected  | Received |  Failed
    = = = = = = = = = = = = = = = = = = = = = = = =  = = = =
    Number of Tests        : 19           18
    Number of Builds       : 48           46
    Number of Simulations  : 166           164        0
    Number of Comparisons  : 105           104        0

    Failed Simulations are: 
    None
    Which comparisons are not bit-for-bit: 
    None

[JS-WRF-SBM]

@JS-WRF-SBM @dudhia @weiwangncar
Koby,
Please review this modification

@davegill @nbald
Thanks for contacting me on this.
From my end the fix is OK - as it builds and run with Intel 19.0.1 as well.

[davegill]

@JS-WRF-SBM
Koby,
Thanks for the quick response!

[nbald]

I confirm it builds now with PGI 19.10

@davegill I am not affiliated with Edgehill University. You can attribute to Nicolas Baldeck (OpenMeteoData)

[nbald]

Is this a duplicate? #1183

[davegill]

@nbald
Nicolas,
well, yes it is!
I am closing this PR.

[davegill]

Look at #1183
This was found again. Perhaps we need to do minor releases more quickly. :)