Take me to the WRF WG5 Forum.

Table of Contents


        John M. Brown (lead)
          NOAA/Forecast Systems Laboratory

        Jian-Wen Bao
          NOAA/Environmental Technology Laboratory

        Tom Black
          NOAA/NCEP/Environmental Modelling Center

        Shuhua Chen
          University of California at Davis and
          Air Force Weather Agency

        Jimy Dudhia
          NCAR/Mesoscale and Microscale Meteorology Division

        Song-You Hong
          Yonsei University, Seoul, Korea and
          NOAA/NCEP/Environmental Modeling Center

        Xin-Zhong Liang
          Illinois State Water Survey
          Department of Natural Resources and University of Illinois at Urbana-Champaign

        Dave Stauffer
          Pennsylvania State University

        Wei-Kuo Tao
          NASA/Goddard Space Flight Center/Laboratory for Atmospheres

        Ming Xue
          OU/Center for the Analysis and Prediction of Storms


1. Incorporate physics packages into WRF.  Specific tasks include

  • selection of physics packages for WRF
  • design of physics modules and their interface to WRF
  • coding physics packages or porting them from other models
  • testing physics modules in WRF framework.

    2. Understand strengths, weaknesses and limitations of individual physics 
        packages as well as their coupling and feedbacks.

    3. Make recommendations regarding the configuration of physics 
        suites to be tested in idealized simulations and in real-data simulations and 

    4. Encourage development of new schemes or improvements (accuracy, 
        consistency, efficiency) to existing physics schemes for application in WRF.


    We define model physics as code describing those processes (Table 5.1) not
    explicitly included in the basic dynamical and thermodynamical equations
    describing the earth's atmosphere. These processes are either too complicated
    to be explicitly included in the model based on their most fundamental physics
    laws (e.g. radiation and microphysics), or finer in scale than can be
    adequately represented by realizable grid resolutions (sub-grid scale
    turbulence, PBL transport).  Yet, their effects on the resolvable scale flows
    and on the sensible weather (e.g., precipitation amount) have to be properly
    included for a model to accurately predict atmospheric behavior
    for NWP purposes.  Simplifications are typically made and
    variables (parameters) on the resolvable scales are often used in treating
    these processes; the resulting schemes are usually referred to as physics

    Table 5.1  Physical processes to be incorporated into WRF.
    Generation of PBL turbulence and related transports (including non-local effects but essentially dry surface-based processes.)
    Surface-atmosphere exchanges (momentum, heat,moisture, might eventually include other quantities, either land or water surfaces) including dependencies on surface and subsurface processes.
    Generation of subgrid-scale turbulence and related transports above the PBL (resulting in primarily local diffusion)
    Convection (non-local fluxes aided by condensation, including "shallow" convection)
    Radiation (short and long wave, atmospheric and surface effects)
    Cloud and precipitation physics (local processes including fallout)


    We have currently implemented and are testing the options listed below in table 5.2. The schemes in the column "In" have already been implemented into the WRF model. The schemes in the "Under Development" column are being worked on currently. This table has been updated for release 2.0 in May 2004.

    Table 5.2  WRF Version 2.0 Physics Options and Packages
    Physics In Under
    • Kessler
    • Lin (Purdue)
    • WSM3 simple ice
    • WSM5 mixed-phase
    • Eta (Ferrier)
    • WSM6 graupel
    • Goddard microphysics (Tao et al.)
    • Reisner graupel
    Convective Parameterization
    • Kain-Fritsch (new)
    • Betts-Miller-Janjic
    • Grell ensemble
    Long-Wave Radiation
    • RRTM
    • GFDL (Eta)
    • CAM
    Short-Wave Radiation
    • Dudhia short wave
    • GFDL (Eta)
    • GSFC
    • RRTM
    • CAM
    Surface layer
    • MM5 Similarity theory
    • Janjic (Eta)
    Land-Surface layer
    • 5-layer soil temperature
    • Noah LSM
    • RUC LSM
    • CLM
    Boundary layer
    • Yonsei (YSU)
    • Mellor-Yamada-Janjic
    Subgrid eddy diffusion
    • Constant diffusion
    • Stress/deformation form (with TKE)
    • Stress/deformation form (Smagorinsky)
    • Horizontal Smagorinsky (mesoscale)


    • Continue to evaluate physics performance in real-time WRF forecasts.
    • Continue to work on adding options for WRF Research Version release.


    • Test physics in idealized settings.
    • Select and test physics suites in NWP testbed periods.
    • Finalize physics modules for Research Version release.
    • Design method to handle contributed physics modules within WRF.

    Interaction with other WRF Working Groups:

    • Dynamic Model Numerics Working Group: Interface of physics with dynamical solver (vertical coordinates, basic variables, time-stepping scheme).
    • Software Architecture, Standards, & Implementation Working Group: Physics coding standards (array and loop ordering, special variables and constants, subroutine interfaces).
    • Standard Initialization Procedures Working Group: Use of real data input for NWP tests.
    • 3D-VAR Data Assimilation System Working Group: ?
    • 4D-VAR Data Assimilation System Working Group: Adjoint codes for physics.
    • Post Processing Working Group: Model output fields and visualization.
    • Model Testing & Verification Working Group: Design tests of physics, use verification to identify areas for improvement.
    • Web Site, Workshops, and Model Support Working Group: Interact with outside institutions to add other physics options.
    • Operational Implementation Working Group: Selection of physics, optimization issues.

    Send email to for comments on this page.