This page provides an overview of the major updates in CCPP Physics v7.0.0.
General Changes:
Suites GFS_v17_p8, HRRR, RRFS_v1beta, and RAP, which were supported with CCPP v6.0.0, have been phased out.
GFS Scale-aware SAS Cumulus Schemes:
The updates between GFSv16 and GFSv17 are carefully outlined in Bengtsson and Han (2024)(submitted to Weather and Forecasting). The main updates include:
- Implementation of a positive definition mass-flux scheme and a method for removing the negative tracers (Han et al. 2022 [82])
- Introduction of a new closure based on a prognostic evolution of the convective updraft area fraction in both shallow and deep convection (Bengtsson et al. 2022 [16])
- Introduction of 3D effects of cold-pool dynamics and stochastic initiation using self-organizing cellular automata stochastic convective organization scheme (not supported in SCM; Bengtsson et al. 2021 [15])
- Introduction of environmental wind shear effect and subgrid TKE dependence in convection, to seek improvements in hurricane forecast prediction (Han et al. 2024 [83])
- Introduction of stricter convective initiation criteria to allow for more CAPE to build up to address a low CAPE bias in GFSv16 (Han et al. 2021 [81])
- Reduction of convective rain evaporation rate to address a systematic cold bias near the surface in GFSv16 (Han et al. 2021 [81])
Grell-Freitas Scale and Aerosol Aware Convection Scheme:
- Update for aerosol-awareness (experimental and not supported)
- Scale-awareness turned off when explicit microphysics is not active anywhere in the column
- Convection is completely suppressed at grid points where the MYNN PBL sheme produces shallow convection
- Radar reflectivity considers mass flux PDF as well as whether scale-awareness is turned on at the gird point in equation
GFS Scale-aware TKE-EDMF PBL Scheme:
- Implementation of a parameterization for environmental wind shear effect in the GFS TKE-EDMF PBL and cumulus schemes to reduce the negative hurricane intensity biases.
- Entrainment rates are enhanced proportionally to the sub-cloud or PBL mean TKE (turbulent kinetic energy) when TKE is larger than a threshold value
- Increased entrainment rate as a function of vegetation fraction and surface roughness length to enhance the underestimated CAPE forecasts in the GFS
- See also
- Han et al.(2024) [83]
MYNN-EDMF PBL Scheme:
- Small increase of buoyancy length scale in convective environment
- Patch for ensuring non-zero cloud fractions for all grid cells where cloud mixing ratio is greater than 1e-6 or ice mixing ratio is greater than 1e-9
Subgrid-scale (SGS) Clouds Scheme:
- Bug fix for cloud condensate input into RRTMG radiation
- New code section for use with SAS cumulus scheme
- Cloud fraction now computed as a mix between the area-dependent form and the modified Chaboureau and Bechtold (2005) [30] form
- Adjusted limit for the boundary flux functions
Thompson Cloud Microphysics Scheme:
- Ice generation supersaturation requirement reduced from 0.25 to 0.15 to generate more ice at the upper levels and reduce the outgoing longwave radiation bias
- Cloud number concentration divided into two parts (over land and others). Number concentration over ocean reduced to a smaller number (50/L) from its previous default (100/L). Both changes were made to increase the downward shortwave radiative flux and reduce the negative bias off coastal regions including the Southeast Pacific
- Small fixes to the minimum size of snow and collision constants
- Note
- The above improvements were tested with the non-aerosol option (in GFS_v17_p8_ugwpv1 suite), so results with the aerosol-aware Thompson (in HRRR_gf suite) may vary.
NSSL Cloud Microphysics Scheme:
- Updated with 3-moment option
Radiation
- Coupling of GOCART aerosols with radiation (iaer=2011)
- Convective cloud water (liquid water + ice water) added to the calculations of cloud water path and ice water path for radiation cloud properties
NoahMP Land Surface Model:
- Option for using the unified frozen precipitation fraction in NoahMP
- Diagnostic 2-meter temperature and humidity now based on vegetation and bare-ground tiles (new namelist option iopt_diag)
- Bug fixes for GFS-based thermal roughness length scheme
- New soil color dataset introduced to improve soil albedo to reduce the large warm bias found in the Sahel desert
- Wet leaf contribution factor is included
- Leaf-area index now depends on momentum roughness length
RUC Land Surface Model:
- Initialization of land and ice emissivity and albedo with consideration of partial snow cover
- Initialization of water vapor mixing ratio over land ice
- Initialization of fractions of soil and vegetation types in a grid cell
- Changes in the computation of a flag for sea ice: set to true only if flag_ice=.false. (atmosphere uncoupled from the sea ice model)
- Separate variable for sea ice, for example: snowfallac is replaced with snowfallac_ice
- Solar angle dependence of albedo for snow-free land
- Stochastic physics perturbations (SPP) introduced for emissivity, albedo and vegetation fraction
- Coefficient in soil resistance formulation (Sakaguchi and Zeng, 2009 [175]) raised from 0.7 to 1.0 to increase soil resistance to evaporation
- Computation of snow cover fraction and snow thermal conductivity updated
MYNN Surface-layer Scheme
- Reintroduce friction velocity averaging over water to reduce noise in 10-m winds in the hurricane regime
Unified Gravity Wave Physics Scheme:
- Replacement of the resolution-dependent effective grid spacing (cdmbgwd) with a constant (=6dx)
- Removal of the planetary boundary layer height in determining the reference level
- Weakening of the momentum stress over land ice to reduce the negative wind biases
- Introduction of a launching level to avoid the underestimation of the blocked stress
- Introduction of damped breaking to prevent the wind reversal in the lower troposphere
- Suppression of gravity wave breaking in the upper atmosphere (over 7.5 hPa) to avoid numerical instability
- Revision in sub-grid orography data considering the mathematical definition of moments
1.12.0