Dynamical 4D vortex initialization for GFDL's global high-resolution hurricane modeling system

A physically consistent and dynamically balanced 4D vortex initialization scheme that is suitable for hurricane prediction with a high-resolution global model will be described. Although this technique can, in principle, be used within an advanced 4D data assimilation system (e.g., EnKF), it is demonstrated, via the selected HFIP cases, that it works well enough with a simple nudging of operational NCEP analysis, producing much improved intensity forecasts and competitive skill in tracks. We will present results using the GFDL global cloud-resolving prototype model, which contains physical parameterizations suitable for horizontal resolution ranging from 1 km to 100 km.