Impact of Assimilating Scatterometer Vector Winds on High-Resolution Hurricane Model Prediction

Impacts of the QuikSCAT surface winds on hurricane model initialization and prediction are investigated using a high-resolution, vortex following, nested-grids model. A new initialization scheme has been developed, which includes a PBL model retrieved sea level pressure (SLP) field from the surface wind and a four-dimensional variational method (4DVar) to assimilate both the SLP and QuikSCAT winds to obtain the initial model fields. The method is tested in simulations of three Atlantic hurricanes. It is found that the improved initial fields had a significant impact on Hurricanes Floyd, Isabel, and Fabian. The improvement in model predicted hurricane intensity is in part depended on the initial conditions from the global model analysis fields.