You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
TYPE: enhancement
KEYWORDS: NSSL microphysics scheme; rain breakup
SOURCE: Ted Mansell (NOAA/NSSL) (@MicroTed )
DESCRIPTION OF CHANGES:
- More accurate saturation mixing ratio calculation (iqvsopt=1)
- Changed default droplet renucleation to irenuc=5, which allows extra nucleation at high supersaturation
- A default explicit rain breakup for 3-moment rain (irainbreak=2) has been added to reduce excessive median drop diameters in rain cores, which in turn increases evaporation and cool pool temperature deficits (were too warm).
- Imposed reflectivity conservation in graupel->hail conversion (ihlcnh=3) and Bigg freezing (both 2- and 3-moment)
- Two optional diagnostic outputs for supersaturation (wrt liquid/ice)
- Changed nssl_hail_on from max_dom array to singe value, because it can't be different for different domains, anyway.
LIST OF MODIFIED FILES:
Registry/Registry.EM_COMMON
doc/README.NSSLmp
dyn_em/solve_em.F
phys/module_microphysics_driver.F
phys/module_mp_nssl_2mom.F
phys/module_physics_init.F
TESTS CONDUCTED:
2. Are the Jenkins tests all passing? Yes.
RELEASE NOTE: NSSL microphysics scheme updates include 1. An explicit rain breakup for 3-moment rain (addresses issue of cold pools being too warm and drops being too large in rain cores), 2. Improved reflectivity
conservation for graupel->hail conversion and drop freezing, 3. More accurate saturation mixing ratio calculation, 4. New default droplet nucleation that controls excess supersaturation much better than
previously (and default is to always predict the number of activated CCN). The update has been submitted to CCPP repository as well.
Copy file name to clipboardExpand all lines: doc/README.NSSLmp
+25-7
Original file line number
Diff line number
Diff line change
@@ -22,16 +22,16 @@ Basic options in physics namelist:
22
22
CCN concentration + options
23
23
24
24
The legacy options (17,19,21,22) still behave as before (for now), but going
25
-
forward one should use mp_physics=18 with modifier flags:
25
+
forward one should use mp_physics=18 with modifier flags. 2025 Update, however, sets nssl_ccn_on=1 by default (keeps supersaturation much more reasonable; except for single moment).
26
26
27
27
mp_physics
28
28
= 22 ! NSSL scheme (2-moment) without hail
29
-
Equivalent: mp=18, nssl_hail_on=0, nssl_ccn_on=0
30
-
= 17 ! NSSL scheme (2-moment) with hail with constant background CCN
31
-
concentration
32
-
Equivalent: mp=18, nssl_ccn_on=0
29
+
Equivalent: mp=18, nssl_hail_on=0, nssl_ccn_on=1
30
+
= 17 ! NSSL scheme (2-moment) with hail is now the same as mp=18
31
+
Equivalent: mp=18, nssl_ccn_on=1 <- must explicitly set nssl_ccn_on=0 to
@@ -98,14 +98,23 @@ Cloud concentration nuclei (CCN) concentration is predicted as in Mansell et al.
98
98
Droplet activation option method is controlled by the 'irenuc' option (internal to NSSL module). The default option (2) depletes CCN from the unactivated CCN field. A new option (7) instead counts the number of activated CCN (nucleated droplets) with the assumption of an initial constant CCN number mixing ratio. Option 7 better handles supersaturation at low CCN (e.g., maritime) concentrations by allowing extra droplet activation at high SS.
99
99
100
100
irenuc : (nssl_mp_params namelist)
101
-
2 = ccn field is UNactivated aerosol (default; old droplet activation)
101
+
2 = ccn field is UNactivated aerosol (old default; old droplet activation)
102
102
Can switch to counting activated CCN with nssl_ccn_is_ccna=1
103
+
5 = ccn field must be ACTVIATED aerosol (new default as of Feb. 2025)
104
+
Must have nssl_ccn_on=1 for irenuc=5
105
+
Allows activation beyond limit of nssl_cccn at higher supersaturation
106
+
as an approximation of nucleation mode aerosol being activated. (Mainly
107
+
an issue for low CCN concentration with deep updrafts.)
108
+
If more strict limitation of activation is desired, use option 7.
103
109
7 = ccn field must be ACTVIATED aerosol (new droplet activation)
104
110
Must have nssl_ccn_on=1 for irenuc=7
105
111
106
112
Excessive size sorting (common in 2-moment schemes) is effectively controlled by an adaptive breakup method that prevents reflectivity growth by sedimentation (Mansell 2010). For 2-moment, infall=4 (default; nssl_mp_params namelist) is recommended. For 3-moment, infall only really applies to droplets, cloud ice, and snow.
107
113
114
+
3-moment active rain breakup (WRF 4.7.x, 2025): The 3-moment rain without explicit breakup can result in cold pools that are too warm and rain median diameters that are too large in rain cores. A bin-model-based breakup parameterization for rain was implemented to address these issues. Very low rain rates (sparse drops) are largely unaffected (e.g., maintains Zdr arc feature). The breakup coefficient (rainbreakfac) has a default value of 1.0e6 and can reasonably be increased up to around 2.5e6 if desired (nssl_mp_params). Active breakup is automatically turned on for 3-moment (irainbreak=2) but not for 2-moment. Option irainbreak=2 is not recommended for 2-moment, but a user may experiment with irainbreak=11, which breaks up large drops in the tail of the spectrum starting at D=draintail (default 10.e-3 m).
115
+
108
116
Graupel -> hail conversion: The parameter ihlcnh selects the method of converting graupel (hail embryos) to the hail category. The default value is -1 for automatic setting. The original option (ihlcnh=1) is replaced by a new option (ihlcnh=3) as of May 2023. ihlcnh=3 converts from the graupel spectrum itself based on the wet growth diameter, which generally results in fewer initiated hailstones with larger diameters (and larger mean diameter at the ground). If hail size seems excessive, try setting ihlcnh=1, which tends to generate higher hail number concentrations and thus smaller diameters.
117
+
UPDATE (4.7.x/2025): The conversion has been updated to conserve reflectivity of the new hailstones compared to the graupel. This results in new hail that is smaller than previously but prevents spurious increases in reflectivity. (Active for both 2- and 3-moment)
109
118
110
119
The June 2023 (WRF 4.6) update introduces changes in the default options for graupel/hail fall speeds and collection efficiencies. The original fall speed options (icdx=3; icdxhl=3) from Mansell et al. (2010) are switched to the Milbrandt and Morrison (2013) fall speed curves (icdx=6; icdxhl=6). Because the fall speeds are generally a bit lower, a partially compensating increase in maximum collection efficiency is set by default: ehw0/ehlw0 increased to 0.9. One effect is somewhat reduced total precipitation and cold pool intensity for supercell storms.
111
120
@@ -138,6 +147,15 @@ Snow self-collection (aggregation) has been curbed in the 4.6 version by reducin
138
147
Snow reflectivity formerly had a default setting that turned on a crude bright band enhancement (iusewetsnow=1). This is now turned off by default (iusewetsnow=0)
139
148
These snow parameters can be accessed through the nssl_mp_params namelist.
140
149
150
+
Saturation mixing ratio (WRF 4.7.x, 2025): New formulation (iqvsopt=1) is more consistent with other microphysics schemes. Previously (iqvsopt=0), the quantity e/(p-e) was approximated as e/p, but the new default restores the full equation and uses slightly more accurate (Bolton) coefficients for the saturation (wrt liquid) tables.
151
+
152
+
New options (Feb. 2025) (not enabled by default):
153
+
- Option (nsplinter=1001) for ice crystal production by drop freezing/shattering (Sullivan et al. 2018)
154
+
- Option (incwet = 1) to treat wet growth only for D > Dwet rather than all or nothing; results in greater hail production due to maintaining dry growth at D < Dwet
0 commit comments