WRF, developed primarily by the National Center for Atmospheric Research (NCAR) and the National Centers for Environmental Prediction (NCEP). Other contributors include the Forecast Systems Laboratory (FSL), CAPS, and the Air Force Weather Agency (AFWA).
WRF is a unified research and operations model. It replaces ETA at NCEP in 2006, and RUC in 2007-2008. It is the successor to MM5, which is no longer being developed or maintained.
WRF uses the fully-compressible, nonhydrostatic equations, and is conservative for scalar variables. The horizontal coordinate is the Arakawa staggered C-grid, while in the vertical a terrain-following mass coordinate is used.
WRF employs advanced numerics. The time integration is 3rd-order Runge-Kutta in time with split acoustic/gravity mode handler, while 2nd- to 6th-order integration is used in the horizontal and vertical. Sub-gridscale turbulence, divergence damping, and external mode filtering are available.
Both 1-way and 2-way nesting is supported, with an option for moving nests. The grid ratio can be any integer.
A number of microphysical schemes are available, including Thompson (6 classes plus ice number concentration). Available cumulus parameterizations include Kain-Fritsch (Eta version enhanced for shallow convection), Betts-Miller-Janic, and Grell-Devenyi ensemble.
Surface layer schemes based on similarity theory have been adapted from Eta and MM5. Multi-layer Land-surface models include Noah and RUC. Planetary boundary layer schemes include MRF, Yonsei University (YSU), and Mellor-Yamada-Janic (MYJ).
WRF executes efficient execution on a range of computing
platforms (distributed and shared memory, vector and scalar
computing). The scientific code is separated from the
parallelization and other architecture-specific codes. WRF is
highly modular, single-source Fortran-90 code for maintainability.
It supports multiple dynamics solvers and physics modules.
WRF’s model coupling API enables it to be coupled with other
models such as ocean, and land models.
Custom web-based interface for managing mesoscale forecasts.
