NSSL-MP bug fix and more

[MicroTed]

Bug fix and enhancements to NSSL microphysics (mp_physics=18)

TYPE: bug fix, enhancements
module_mp_nssl_2mom.F:

• Fixed error in interface that missed unit conversion for reflectivity moments, which resulted in advection errors (and units of m^6/m^3 instead of m^6/kg) Resolves issue NSSL-MP 3-moment bug in unit conversion #2344
• Fixed reflectivity conservation in 2-moment graupel->hail conversion (could cause spurious large hail number concentrations). 3-moment now has this turned off by default. Turned off reflectivity check for drop freezing.
• Increased the default rain breakup coefficient (from 1e6 to 2.5e6) for 3-moment and rewrote the Z-rate equation to prevent negative number concentrations. The higher value of rainbreakfac helps boost cold pools, which were a bit weak.
• Added missing control flags in smallvalues for 3-moment consistency checks
• Added subroutines to compute radar reflectivity from outside of the driver
• 3-moment: turned off rescaling of Z for graupel and hail at lowest value of alpha (shape parameter), which allows larger Z to exist that normally would be require a negative shape parameter. (Rain still gets Z rescaled at alphamin)
• 3-moment: Fixed lookup table for incwet=1 (default is 0) for calculating dry/wet growth over portions of the PSD instead of all or nothing. Also fixes a minor error with incwet=1 (small effect). incwet=1 is now safe to use [no default effect]
• Fixed a non-WRF NaN issue with the ssfilt array [no default effect]
• Moves a couple inline functions to separate routines [no default effect]
• Other changes for future options
  Registry.EM_COMMON, module_physics_init.F:
• Added nssl_alphar to registry (namelist) to allow easier setting of 2-moment rain shape parameter

KEYWORDS:
3-moment; hail conversion

SOURCE: Ted Mansell (NOAA)

DESCRIPTION OF CHANGES:
Problem: For 3-moment option, the reflectivity moments were not having their units converted between advection (m^6/kg) and in-scheme (m^6/m^3), resulting in advection errors and incorrect output units

Solution:
Added appropriate in-scheme flags to multiply and divide by air density

ISSUE:
Fixes #2344

LIST OF MODIFIED FILES:
Registry/Registry.EM_COMMON
phys/module_mp_nssl_2mom.F
phys/module_physics_init.F

TESTS CONDUCTED:

RELEASE NOTE:
NSSL Microphysics: Fixes a bug in the 3-moment option that caused advection errors; Enhances cold pools in 3-moment by increasing rain breakup (and correcting reflectivity change); Fixes a 2-moment reflectivity check in the hail conversion process; Allows greater graupel/hail reflectivities in 3-moment.

[MicroTed]

The NSSL module could now be moved to a git submodule if desired, which would be consistent with pending inclusion in MPAS. I can go either way.

[weiwangncar]

The regression test results:

Test Type              | Expected  | Received |  Failed
= = = = = = = = = = = = = = = = = = = = = = = =  = = = =
Number of Tests        : 23           24
Number of Builds       : 60           57
Number of Simulations  : 158           150        0
Number of Comparisons  : 95           86        0

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

[weiwangncar]

@MicroTed You have just missed our 4.8 release - which has been tested and staged for release very soon. Since you're making the code available to MPAS, we think it would be a benefit to move your code to a git submodule so that you can maintain one version of it. We will consider it in our next release - maybe a 4.8.1.

[dudhia]

@MicroTed You have just missed our 4.8 release - which has been tested and staged for release very soon. Since you're making the code available to MPAS, we think it would be a benefit to move your code to a git submodule so that you can maintain one version of it. We will consider it in our next release - maybe a 4.8.1.

I concur with @weiwangncar here. You can decide whether we should wait for the submodule or merge this on develop first.

[MicroTed]

@weiwangncar Would it be possible to do a quick 3-line bug fix instead? But if it's too late, then never mind.

[weiwangncar]

@MicroTed We will have a bug fix release some time after the 4.8 release. At this time, it is best to wait until then. Thanks for your understanding.

[MicroTed]

@weiwangncar No worries. I suppose could queue that fix up shortly as PR and switch this one to a draft.

[weiwangncar]

@MicroTed Yes, you can make the PR now and target the develop branch. Thanks!