A Comparison of HWRF Six-Hourly 4DEnVar and Hourly 3DEnVar Assimilation of Inner Core Tail Dopper Radar Observations for the Prediction of Hurricane Edouard (2014)

Six-hourly three-dimensional ensemble variational (3DEnVar) (6H-3DEnVar) data assimilation (DA) assumes constant background error covariance (BEC) during a six-hour DA window and is, therefore, unable to account for temporal evolution of the BEC. This study evaluates one-hourly 3DEnVar (1H-3DEnVar) six-hourly 4DEnVar (6H-4DEnVar) methods analyses forecasts hurricanes with rapidly evolving Both BEC in hybrid EnVar system. In order compare these methods, experiments are conducted by assimilating inner core Tail Doppler Radar (TDR) wind Hurricane Edouard (2014) running Weather Research Forecasting (HWRF) model. most metrics, 1H-3DEnVar 6H-4DEnVar verify better than 6H-3DEnVar. produces thermodynamic 1H-3DEnVar. reflectivity shows that structure forecasts. For first 24–48 h intensity forecast, forecast performs verified against best track. Degraded found be associated storm center location as result underdispersive spread. Removing improves