/** \page MYNNEDMF MYNN-EDMF Boundary Layer and Shallow Cloud Scheme \section mynnedmf_descrip Description The Mellor-Yamada-Nakanishi-Niino (Nakanishi and Niino 2009 \cite NAKANISHI_2009) eddy diffusivity-mass flux (EDMF) scheme was implemented into CCPP to introduce an alternative turbulent kinetic energy (TKE)-based planetary boundary layer (PBL) scheme which could serve as a candidate PBL parameterization for future operational implementations of the Unified Forecast System. The MYNN-EDMF is currently employed in NOAA's operational Rapid Refresh (RAP; Benjamin et al.2016 \cite Benjamin_2016) and High-Resolution Rapid Refresh (HRRR) forecast systems. The original MYNN scheme was demonstrated to be an improvement over predecessor Mellor-Yamada-type PBL schemes (e.g., Mellor and Yamada 1974,1982 \cite Mellor_1974 \cite Mellor_1982) when compared against large-eddy simulation (LES) of a convective PBL (Nakanishi and Niino 2004, 2009 \cite Nakanishi_2004 \cite NAKANISHI_2009), the prediction of advection fog (Nakanishi and Niino 2006 \cite Nakanishi_2006), and for the representation of coastal barrier jets (Olson and Brown 2009 \cite olson_and_brown_2009). The MYNN scheme can be configured to function at level 2.5, 2.6 (current default) or 3.0 closure and includes a partial-condensation scheme (also known as a cloud PDF or a statistical-cloud scheme) to represent the effects of subgrid-scale(SGS) clouds on the buoyancy flux (Nakanishi and Niino 2004, 2006, and 2009 \cite Nakanishi_2004 \cite Nakanishi_2006 \cite NAKANISHI_2009). The closure constants for the original MYNN scheme were tuned to a database of LES as opposed to observational data. The MYNN-EDMF scheme has been extensively developed to improve upon the forecast skill of the original MYNN, largely driven by requirements to improve forecast skill in support of the NOAA's National Weather Service (NWS), the Federal Aviation Administration (FAA) and users within the renewable-energy industry.Specifically, fundamental changes were made to the formulation of the mixing lengths and representation of subgrid-scale(SGS) clouds, but new components have also been added to improve the representation of non-local mixing, the turbulence interaction with clouds, and the coupling to other model components (i.e., radiation). A description of the changes to the MYNN scheme are available in Olson et al.(2019) \cite https://doi.org/10.25923/n9wm-be49. \section intra_mynnpbl Intraphysics Communication - \ref SGSCLOUD_page The SGS clouds produced by the MYNN-EDMF scheme are coupled to the longwave and shortwave radiation schemes if the namelist parameter \p icloud_bl is set to 1. In this case, the SGS cloud fraction, \p CLDFRA_BL, and the SGS cloud-mixing ratio, \p QC_BL, are added to the microphysics arrays within the radiation driver (mynnrad_pre_run()). The following two steps are performed: (1) Use either Xu and Randall (1996) \cite xu_and_randall_1996 or Chaboureau and Bechtold (2005) \cite Chaboureau_2005 cloud fraction for the convection scheme subgrid clouds, chosen by the switch \p conv_cf_opt = 0: CB2005, 1: XR1996. Note that the MYNN-EDMF PBL scheme subgrid clouds input into this scheme are already partitioned into qc and qi and already have assigned cloud fractions, but now qi is further partitioned into qi and qsnow. (2) if the resolved-scale cloud liquid (\f$q_c\f$) is less than \f$10^{-6}kg kg^{-1}\f$ , and there exists a nonzero SGS cloud fraction, then the SGS components are added to their respective resolved-scale components by a temperature weighting, according to linear approximation of Hobbs et al.(1974) \cite HOBBS_1974 . Next the frozen species are crudely split into 50% ice and 50% snow below approximately 700 hPa. The percentage of snow decreases with altitude and tapers off at approximately 300 hPa: \f[ W_{snow}=min(0.5,max((P-30000,0.0)/140000.0) \f] \f[ W_{ice}=1.0-W_{snow} \f] Next the SGS cloud water, liquid, and snow are sorted as: \f$q_c\f$=QC_BL*CLDFRA_BL \f$q_i\f$=QI_BL*\f$W_{ice}\f$*CLDFRA_BL \f$q_s\f$=QI_BL*\f$W_{snow}\f$*CLDFRA_BL This allows using a single 3-D array for both SGS cloud water,ice and snow. The updated \f$q_{c}\f$,\f$q_{i}\f$, \f$q_{s}\f$and \p CLDFRA are then used as input into the radiation schemes. After exiting the radiation schemes, the original values of \f$q_c, q_i, q_s\f$ and \p CLDFRA are restored, so the SGS clouds do not impact the resolved-scale moisture budget. The MYNN-EDMF CCPP-compliant interface: \ref arg_table_mynnedmf_wrapper_run \section gen_mynnedmf_conv General Algorithm \image html MYNN-EDMF_call_order.png "Figure 1.The order of subroutines within the MYNN-EDMF (Courtesy of J.B. Olson). The green rectangles within the main subroutine (mynn_bl_driver()) represent subroutine calls. The blue rectangles represent tasks coded within the main driver. A brief description is shown on the right " width=600 \ref gen_mynn_bl_driver */