Advanced Tests of GSI Hybrid 4-D and 3-D Ensemble-variational Data Assimilation For Rapid Refresh

Abstract

NOAA’s Rapid Refresh forecast system (RAP) is an hourly-updated regional data-assimilation and forecasting system that uses the Weather Research and Forecasting (WRF-ARW) model with 13-km horizontal grid spacing and the Gridpoint Statistical Interpolation (GSI) analysis package. A 3 dimensional Ensemble Variational (3DEnVar) hybrid assimilation procedure within GSI has been applied since the RAP version 2. The GSI 3DEnVar analysis has significantly improved RAP forecast for upper air wind, moisture, and temperature even though the ensemble perturbations used in 3DEnVar analysis were from GFS EnKF global ensemble forecasts. In the recent operation update of GFS, a GSI-based hybrid 4 dimensional Ensemble Variational (4DEnVar) procedure was implemented. The GFS results show the benefit from using hybrid 4DEnVar method for 6-hourly continue cycling global data assimilation system. This talk is to report GSD efforts in improving the GSI hybrid 3DEnVar configurations for RAP operation and testing of applying the GSI hybrid 4DEnVar method in RAP system. Because RAP is an hourly cycling system and has very short observation cut off window, the configuration of the hybrid 4DEnVar will need to be tuned with RAP retrospective experiments conducted to evaluate the impact of hybrid 4DEnVar method for rapid update applications.