FSL will conduct a comparison of 10-km model forecasts from the RUC and WRF models during the IHOP (International H2O Project) experiment to be run in real time. Both of these models will be run with initial conditions from the RUC 20-km cycle run at FSL, including assimilation of special observations. The intent of this comparison is to assess the suitability for replacement of the hydrostatic RUC model with the non-hydrostatic mass-coordinate WRF model in a future version of the Rapid Update Cycle. Accordingly, we designate this model as RUCWRF for this application and set of experiments.

A case considered to be representative for the model intercomparison will be selected from those run during the IHOP period in May-June 2002. The case study will include a detailed examination of successful and unsuccessful aspects of the model forecasts. An investigation will be made to determine specific model or initial condition issues responsible for performance.

The RUC version will be approximately equal to that being implemented into operations at NCEP in April 2002. The data assimilation system will run in a national-scale 20-km hourly cycle, with assimilation of some special observations available for IHOP, where possible. The 10-km models will be run over a regional domain covering the central US with about 2000 km on each side. The hydrostatic RUC model will be the same as that in the 20-km RUC at NCEP, except that it will be run at 10-km horizontal resolution. A 10-km RUC has already been run successfully for 24-h forecasts during early 2001 and early 2002 over the western US as part of the PACJET experiment. The RUCWRF model to be run will be the WRF mass-coordinate model with a dynamic core developed at NCAR. It will include, to the extent possible, the same physical parameterizations as run in the hydrostatic RUC model, including the Grell-Devenyi convection parameterization and the RUC / Smirnova land-surface model.