Improvements to numerical forecasts of cloud, precipitation, precipitation type, and aircraft icing are clearly dependent on improving initial conditions for those forecasts. The Rapid Update Cycle at NCEP uses the bulk mixed-phase cloud microphysics scheme from the NCAR/Penn State MM5 model, with 5 hydrometeor types explicitly forecast. To improve the cloud/precipitation forecasts in models such as the RUC with more advanced cloud microphysics schemes, it is necessary to develop forward models from remotely sensed data to the 3-d hydrometeor and water vapor distribution (not to mention the further goal of modifying wind fields such that vertical motions are also consistent).

With the implementation of the 20km RUC at NCEP this spring, an initial cloud/hydrometeor assimilation technique using GOES cloud-top pressure data is being introduced. The scheme uses the 3-dimensional RUC hydrometeor fields as a background, and includes both cloud building and clearing. This scheme will be described, including its effect on RUC forecasts. The effects have been a consistent improvement on short-range cloud forecasts, even out to 12 h, and a modest improvement for precipitation forecasts.

The next step in development of the RUC cloud/hydrometeor analysis is to add radar reflectivity data. Problems associated with this step are significant, but consistency with the GOES cloud data can be used as part of the quality control. Initial testing of the scheme adding radar to satellite data to initialize RUC hydrometeor fields will be presented at the conference.