Testing and Development of the Data Assimilation System for Rapid Refresh Forecast System

The Rapid Refresh Forecast System (RRFS) is NOAA’s next generation regional convection-allowing ensemble forecast system under development for the National Weather Service. RRFS is jointly developed by EMC and GSL along with the wider Unified Forecast System (UFS) community. The RRFS incorporates an hourly cycled deterministic forecast system, an hourly cycled EnKF data assimilation system and a 5 member ensemble forecast system. The deterministic forecast system includes an hourly spin-up cycle and an hourly product cycle. The spin-up cycle starts from 3-h GFS forecast at 03 and 15z and cycled for 6 hours. GDAS ensembles are only used in hybrid 3dEnvar in the first spin-up cycles when starting RRFS, otherwise, 30 RRFS ensembles will be used. The product cycle also does hybrid 3DEnVar data assimilation with 30 RRFS 3-km ensemble members. The deterministic forecast system provides an 18 hour forecast hourly and 60 hour forecast every 6 hours. The hourly cycled EnKF is initialized from GEFS ensemble members twice a day at 06z and 18z with assimilation of both conventional and radar reflectivity observations. The ensemble forecast is initialized from EnKF members with additional multi-physics perturbation for 60 hour forecasts at 00, 06, 12 and 18z. Toward the implementation of the first version of RRFS, within the constraints of the limited computational resources at NCEP, RRFS will be tested with real-time parallel experiments and retrospective experiments. The forecast performance will be reported at the conference.