Improving Vortex Initialization and Hurricane Forecasting through 3dEnVar and 4dEnVar Hybrid Data Assimilation Methods

This presentation summarizes our current research results in improving the vortex initialization with both hurricane weather research and forecasting (HWRF) model with NCEP GSI-based ensemble-variational hybrid data assimilation systems (e.g, both 3dEnVar and 4dEnVar). A series of results will be presented to demonstrate the effectiveness of various data assimilation configurations in improving vortex initialization and numerical prediction of track and intensity of hurricanes. Specifically, results are summarized in the following aspects:

1) the influence of enhanced background error covariance terms;

2) 3dEnVar versus 4dEnVar;

3) assimilation of the inner core tail Doppler radar (TDR) radial velocity observations;

4) assimilation of radar reflectivity data with cloud analyses

5) assimilation of GPM satellite radiances and satellite-derived AMVs;

Problems, challenges, recent progress in improving hurricane vortex initialization will be discussed.