An Update of the Recent Development of the Hybrid EnVar Data Assimilation and Convection-Allowing Ensemble System Design to Improve the Operational Hazardous Weather Forecast over the Continental United States

This presentation will provide an update on the recent development of the hybrid ensemble-variational data assimilation and ensemble forecast system targeted on improving convection allowing prediction of hazardous weather over the continental US. First, we will discuss the recent development of the hybrid EnVar for convective scale radar data assimilation. In particular, we will present further development of the static background error covariance for direct assimilation of radar reflectivity within the GSI based hybrid EnVar system. With this development, the hybrid DA system now fully takes advantage of both the static covariance and ensemble covariance for storm scale data assimilation. Experiments in HRRR and WoF contexts have shown positive impacts of the inclusion of the static covariance and will be presented in the conference. Secondly, we will discuss several advancements on convection allowing ensemble system design. We will discuss how to optimally perturb initial condition errors in a multi-scale scenario and the impact of using a single dynamic core multi-physics approach vs multi-dynamic core approach to sample model errors.