Performance of RUC13 and WRFRUC13 forecasts for the AIRS-2 11 November 2003 icing case

The Rapid Update Cycle (RUC) operational configuration at NCEP was upgraded on 28 June 2005. The RUC continues occupy the "situation awareness" niche at NCEP, producing hourly analyses and hourly short-range (out to 9 or 12h) forecasts based on these analyses. But with this upgrade, horizontal resolution is improved to 13km horizontal grid spacing (from 20km). There were also upgrades to all aspects of the RUC analysis, model and postprocessing. Because of the importance of RUC in providing guidance for aviation hazards, emphasis continues to be placed on improving the NCAR mixed-phase, bulk microphysics scheme that has been used in the model since 1998. The new RUC13 includes the Thompson et al (2004, Monthly Weather Review) upgrade to the original Reisner scheme, including making the zero-intercept for the inverse-exponential size distribution of snow particles a function of temperature instead of mixing ratio, and allowing representation of drizzle by introducing a variable zero-intercept for the inverse-exponential size distribution of raindrops. The Thompson et al scheme tends to produce more supercooled liquid water, much less graupel, and somewhat more snow than earlier versions of the NCAR microphysics used in the RUC.

Current plans call for the RUC to be replaced in late 2007 by a new "Rapid Refresh" system now under development at the Forecast Systems Lab. Two major changes from the present RUC will be 1) expansion of the domain to cover all of North America and adjacent ocean areas, and 2) use of some version of WRF as the forecast model in place of the current hydrostatic RUC model. We are now testing a candidate version of WRF (the "Advanced Research WRF) at FSL and expect to be running this test version with the Thompson et al microphysics later this year. This version, known as the WRFRUC13, uses RUC initial and lateral boundary conditions on a horizontal grid and domain identical to the RUC13. Its performance is being evaluated against the RUC13 on a continuing basis. One of the systematic differences we note is that the WRFRUC produces substantially more grid-scale precipitation than does the RUC, a difference we expect will continue when we run the 2 models with identical microphysics.

At the conference we will discuss results of the 11 November 2003 AIRS-2 case from both RUC13 and our pre-Rapid Refresh test version of WRF, with emphasis on the mixed-phase microphysics as it relates to icing. Comparisons will be made with observations as well as between comparable runs using the 2 models. Although the boundary layer and radiation physics differ between the 2 models, this does afford an opportunity to see how WRF (at least the ARW) and RUC compare at identical horizontal and similar vertical resolutions, and with identical initial and lateral boundary conditions.