Evaluation of Bogus Vortex Techniques Using a Four-dimensional Variational Data Assimilation System

The effectiveness of techniques for creating " bogus" vortices in numerical simulations of hurricanes is examined by using the Penn State/NCAR nonhydrostatic mesoscale model ( MM5) and its adjoint system. A series of four-dimensional variational data assimilation experiments is conducted to generate initial vortices for matured hurricanes by assimilating surface pressure and wind information into the mesoscale model. Several different strategies are tested for improving the vortex representation. In addition, available satellite remote sensing data are being assimilated to refine the vortex and provide improved environmental conditions. Results from several cases from the 1998 hurricane season will be presented.