… into feature/mpas_nuopc_init

…nopy2

Adding Integrated Canopy Effects to CCPP/PBL Scheme

… into feature/mpas_nuopc_init

Fixes ufs-community#300.

…an66:

a) As the current large horizontal momentum diffusivity (Kh) at
large-scale and meso-scale (LSMS) looks too large, Kh is computed at zl
and modified as Kh=Kq=0.4*elmh*sqrt(tke),
elmh=min(1.5*sqrt(zlup*zldn,500), where zlup is parcel's upward travel
length and zldn is parcel's downward travel length with tke momentum.
b) Scale-aware non-local mass flux (xmf, xmfd) is now given as
xmf=(1-sigma)^2 * (sigma*wu) rather than the current xmf=pfnl*c1*wu,
where wu is updraft velocity, c1 is a constant, sigma is updraft
fraction, and pfnl is partition function for nonlocal flux. The constant
c1 is the updraft fraction in LSMS and it should be sigma in small
scales as sigma becomes larger than c1 in small scales. Also, the
scale-awareness parameter, (1-sigma)^2, is more physically based
following Arakawa and Wu's study than the empirical parameter, pfnl.
c) Refinement of scl (height of 50% of max tke): max tke can be
occasionally found in Jet stream and then scl would be too large. In the
update, we look for max tke below 2.5km height and then compute scl and
limit to 2.5km (i.e., scl=min(max(scl,500),2500))
d) ri(gradient Richardson number) is computed as ri=bf/shr2 (1dtke)
rather than ri=bf/(def_1+def_2), where bf is buoyancy, because the
parameters given as a function of ri (such as Prandtl number) is derived
based on ri=bf/shr2.
e) TTE-EDMF (total turbulent energy based EDMF scheme) is combined into
sa3dtke scheme as it shows slightly better GFS forecast skill than
tke-based scheme. The logical parameter to conduct TTE-EDMF is
'tte_edmf'.

Combination of 264,265,291 (MERRA2 3-hr, photochem diag, prog closure SFS)

… into feature/mpas_nuopc_init

… into HEAD

…k that w3emc is using the expected precision

…k that w3emc is using the expected precision

…o 1000m for the 3DTKE option.

…tke_update

Fixes ufs-community#300.

ufs-dev ufs-community#295

…o_20251119

address unreasonable large F0 and F1 in entrainment calculation

Combination of NCAR#1175 and NCAR#1176

…composites_pre.F90

This PR changes the option relative_path in the CCPP metadata to dependencies_path as discussed in NCAR/ccpp-framework#685.

NCAR/main sync

…_gjf

Update of CDEPS inline sync without sfc_data

…cle_non6

Allow ihr to be non-multiples of 6

items once and reuse, rather than recomputing in each call to RUCLSM.

…_gjf2

fix bug in GFS_surface_composites_pre.F90