4.1
Status report on the Rapid Refresh
Stephen S. Weygandt, NOAA/ESRL/GSD, Boulder, CO; and J. M. Brown, S. G. Benjamin, T. G. Smirnova, M. Hu, D. Dévényi, K. J. Brundage, J. B. Olson, W. R. Moninger, C. R. Alexander, G. Grell, B. D. Jamison, S. E. Peckham, and T. L. Smith
After more than 15 years of service as the "situational awareness" and very short-range numerical weather prediction forecast model at the National Oceanic and Atmospheric Administration's National Centers for Environmental Prediction, the Rapid Update Cycle is scheduled to be replaced in the near future.
The RR will perform the same hourly updating and situational awareness function as the RUC, but over a much larger, North American domain. For the analysis, we are using the Gridpoint Statistical Interpolation (GSI) 3dVAR procedure currently used for the North American Mesoscale and Global Forecast System at NCEP. The RUC hydrometeor analysis has been added to the GSI (Hu et al, this conf.), and other modifications to the GSI needed for more effective use of surface and aircraft data are under development and testing (Devenyi et al, IOAS). These include improved quality control using platform-dependent history of observation minus 1-h forecast background differences, accounting for disparity between model and actual terrain height in computing the surface-observation innovations, and spreading the influence of the surface observations vertically when the 1-h forecast background exhibits a surface-based mixed layer. The RR model component will be the Advanced Research (NCAR mass coordinate) version of the Weather Research and Forecast model (WRF-ARW), but with physics that is similar to that used in RUC: the Rapid Radiative Transfer Model longwave and Dudhia shortwave radiation, updated versions of the Land-Surface Model (RUC LSM) and the Thompson mixed-phase bulk microphysics, and, most likely, the Mellor-Yamada-Janjic surface and boundary-layer schemes for vertical mixing.
This talk is intended as a report on progress toward the planned initial 2010 operational implementation, and will describe the configuration of the RR that is currently in testing. Aspects of current performance will be discussed, and examples of recent forecasts will be shown.
Session 4, Nowcasting and Modeling, Part 1
Tuesday, 19 January 2010, 11:00 AM-12:00 PM, B314
Next paper