It is well-known that forecasting the amount and distribution of rainfall that accompanies a tropical cyclone (quantitative precipitation forecasting, or QPF) is one of the most difficult tasks in predicting the impact of a landfalling tropical cyclone. Factors that affect rainfall distributions include the translational speed of the storm, topography and orientation of the coast, and interactions with the environmental wind and thermodynamic fields. These effects can combine to produce vastly different rainfall distributions for different storms, even for the same storm at different periods in its lifetime. The purpose of this study is to focus on one factor, the vertical shear of the environmental wind, and investigate its role in governing tropical cyclone rainfall distributions.

This talk presents results from high-resolution (grid length ~ 1 km) multi-day numerical simulations of various tropical cyclones that have been conducted during the past two years using the Penn State/NCAR mesoscale model MM5. These simulations have produced a dataset encompassing a variety of tropical cyclone strengths, motions, and rainfall distributions, as well as a variety of environmental shear profiles. Three storms have been simulated with a reasonable amount of success: Hurricanes Bonnie and Georges from 1998 and Hurricane Floyd from 1999. A movable-mesh technique has been developed for MM5 that enables a high-resolution domain to follow the storm for considerable periods of time, providing the ability to resolve changes in the inner core over time periods that allow for potentially significant changes in the environment of the storm. Relationships between the magnitude and direction of the vertical shear vector, the storm motion vector, and the rainfall distributions produced from the tropical cyclones have been noted in the simulations and will be presented here.