Weather Research and Forecasting Model (WRF)

The Weather Research and Forecasting Model (WRF) is a multi-scale numerical weather prediction system designed to meet the needs of atmospheric research and weather forecasting. It is equipped with a data assimilation system and software structure facilitating parallel computing and system scalability. This model provides a wide range of cross-scale meteorological applications, from tens of meters to thousands of kilometers.

WRF mode allows researchers to generate atmospheric simulations based on actual data (observations, analyses) or idealized conditions. The two cores of the WRF model are designed for different purposes, namely ARW (Advanced Research WRF) and NMM (Non-hydrostatic Mesoscale Model). The two are the same in WRF architecture but emphasize different aspects. ARW is developed by NCAR's MMM (Mesoscale and Microscale Meteorology Division), which focuses on research and can be adjusted to static stability mode. It can be applied to ideal case simulation, parameterization research, data assimilation research, real-time numerical weather forecast and analysis, typhoon simulation, regional climate research, coupling model application, teaching, etc. The major physical processes of WRF-ARW can be divided into microphysics, cumulus parameterization, planetary boundary layer, surface process, and radiation. NMM, developed by the NCEP of NOAA (National Oceanic and Atmospheric Administration), focuses on non-hydrostatic stability and real-time forecasting, which can be applied to real-time numerical weather forecasting and analysis, parameterization research, coupled model applications, teaching, etc.

The WRF model is designed with a standardized language with complete numerical calculation, data assimilation techniques, multiple moving grids, and physical processes (especially convective and mesoscale precipitation processes). The grid design of WRF adopts the Arakawa C staggered grid method on the horizontal plane. This grid design staggers the velocity field at the left, right, upper, and lower half grids of the mass field/thermocline/chemistry, presenting a staggered distribution grid framework. The eta(η) coordinates in the vertical direction are equivalent to the terrain coordinates of MM5 model sigma(σ) and can be used in inconsistent grid intervals.