Assimilation of Polar Orbiter and Geostationary Satellite Data in the Rap and HRRR Models: Recent Upgrades and Ongoing Work with New Observation Sets.

Version 4 of the Rapid Refresh (RAP) will be operationally implementation at NCEP in May 2018. The RAP uses the Gridpoint Statistical Interpolation (GSI) hybrid variational/Ensemble Kalman Filter (EnKF) data assimilation system including an advanced satellite radiance assimilation package developed at NCEP, with ensemble information for the regional assimilation application obtained from the 80-member global ensemble data assimilation system.

We will report on the extensive radiance assimilation enhancements included in RAPv4 and describe work on RAPv5. For RAPv4 a large number of new radiance assimilation types were added and extensive use was made of direct broadcast data. A key focus for RAPv5 is use of the new and improved GOES-16 datasets (including ABI, GLM, and improved AMV winds) in both the RAP and HRRR.

A series of radiance upgrades has been tested and included in the RAP version 4, including adding new data (e.g., ATMS, CrIS, IASI, SEVIRI, AIRS etc.) with channel selection, supplementing the standard data feed with direct broadcast data, and delaying partial cycles for improved real-time data coverage. With the new radiance updates, positive impact has been seen in our retrospective runs. We are conducting a comprehensive data impact study for each individual and/or combined data set within the framework of RAP version 4 through retrospective runs with short-range forecast verifications. In addition, we will be examining assimilation of the ABI data from GOES-16 in RAPv5 and testing assimilation of lightning, AMV, and cloud-top cooling-rate data in HRRRv4. At the conference, we will report on progress to date in all aspects of our geostationary and polar orbiter satellite assimilation within RAP and HRRR.