Improving Short-term Precipitation Forecasting with Radar Data Assimilation and a Multiscale Hybrid Ensemble-Variational Strategy

Improving convection-permitting numerical weather prediction (NWP) relies not only on accurate convective-scale analysis that is often obtained via radar data assimilation, but also on properly representing the large-scale environments where the storms initiate, evolve, and dissipate. In this paper we present a multiscale hybrid ensemble-variational (EnVar) data assimilation strategy with an hourly rapid update, aiming to improve the analysis of both convection and its environment. In this multiscale hybrid EnVar strategy, the flow-dependent background error covariance (BEC) is computed using ensemble members updated by DART EnKF assimilating conventional data. A two-step approach is employed in the hybrid EnVar to achieve improved multiscale analysis by assimilating radar data and conventional data separately in two successive steps. In addition, this study also examines the impacts of the flow-dependent BEC generated with/without radar data assimilation on the performance of hybrid EnVar analysis and ensuing convective forecasting.

The improved performance of the multiscale hybrid EnVar strategy over 3DVar and EnKF is demonstrated using a convective rainfall case documented during the PECAN field campaign. We also illustrate the benefit of the two-step assimilation strategy in improving both the convective-scale and large-scale analyses and forecasts of convective rainfall. In addition, it is shown that the ensemble BEC generated without radar data assimilation lead to improved hybrid EnVar analysis over that with radar data assimilation by better representing uncertainties in convective environment and reducing spurious spatial and multivariate correlations.