Radar Data Assimilation for Convective-Scale Ensemble Forecasts Using the CAPS EnKF System

Radar Data Assimilation for Convective-Scale Ensemble Forecasts Using the CAPS EnKF System 
 Youngsun Jung
 Center for Analysis and Prediction of Storms (CAPS)
University of Oklahoma
Currently, radar is the only observational platform that can provide complete four-dimensional coverage of a convective system at sufficiently high-spatial and high-temporal resolutions. Assimilating radar observations is now common practice in storm-scale numerical weather prediction (NWP) models. Because of its ability in handling complex, nonlinear, physical processes in the assimilation model and in the forward observation operators, the ensemble Kalman filter (EnKF) method has been a popular choice in assimilating them. CAPS has developed an efficient hybrid MPI-OpenMP parallel algorithm for an ensemble square root filter (EnSRF) especially suitable for dense observations. This DA system has been used in many research articles. In 2015, CAPS is running fully cycled EnKF DA on a CONUS 3 km grid, assimilating observations from more than 140 WSR-88D radars, and producing about 15 ensemble forecasts during the NOAA Hazardous Weather Testbed Spring Experiment. In this talk, some research results in radar DA and a brief summary of real-time forecasts will be presented. Considerations and challenges will also be discussed.