Assimilating Conventional and Doppler Radar Data with a Hybrid Approach to Improve Forecasting of a Convective System

A hybrid ensemble adjustment Kalman filter—three-dimensional ensemble—variational (EAKF-En3DVar) system is developed to assimilate conventional and radar data, and is applied to a convective case in Colorado and Kansas, USA. The system is based on the framework of the Weather Research and Forecasting model’s three-dimensional variational (3DVar) and Data Assimilation Research Testbed. A two-step assimilation procedure with a shorter length scale and analysis cycle is used to reduce analysis noise in radar data assimilation. Results show that the hybrid experiment assimilating only conventional data improves the quantitative precipitation forecast (QPF) and quantitative reflectivity forecast over those of a 3DVar experiment, and the improvements are also evident after assimilating radar data. The assimilation of radar data substantially improves the QPF up to seven hours, with either the 3DVar or hybrid method. The hybrid experiment assimilating both conventional and radar data forecasts a more accurate convective system in terms of structure, spatial extent and intensity and produces increased low-level cooling and mid-level warming in the convective region. These improvements are attributable to an improved forecast background field of wind, temperature and water vapor mixing ratio, with maximum root mean square error reduction at the tropopause and near the surface.