J2.3
The 20-km version of the RUC
Stanley G. Benjamin, NOAA/OAR/FSL, Boulder, CO; and G. A. Grell, S. S. Weygandt, T. L. Smith, T. G. Smirnova, B. E. Schwartz, D. Kim, D. Devenyi, K. J. Brundage, J. M. Brown, and G. S. Manikin
In spring 2001, a new version of the Rapid Update Cycle (RUC) will be implemented at NCEP, with extensive changes to the analysis and forecast model as well as the doubling of horizontal resolution from the previous 40km resolution. This is the first major overhaul of the RUC since April 1998. This paper will summarize these differences and changes in forecast and analysis characteristics resulting from the changes. The assimilation changes include a 3d variational analysis, assimilation of GOES cloud-top data, and improved treatment of near-surface data. The 20-km RUC model includes a new version of the Grell convective parameterization, significant modifications to the MM5/RUC mixed-phase cloud microphysics, and much more detailed treatment of the land surface. The paper will summarize the integrated effects of these changes -- in particular, improved near-surface and precipitation forecasts.
The RUC is still the only operational assimilation system with a 1h intermittent forward cycle, but now with 20-km resolution. This means that the forecast errors that the hourly assimilation is trying to correct have still smaller spatial scales, while the data are of essentially the same resolution as before. Results will be presented comparing accuracy of RUC 1h forecasts vs. 3h/6h forecasts for different variables and levels as a means to investigate how much aliasing from high-resolution high-frequency assimilation may be affecting results.
Joint Session 2, Mesoscale Data Assimilation (Parallel with Session 3)
Tuesday, 31 July 2001, 8:00 AM-9:29 AM
Previous paper