The 23rd Conference on Hurricanes and Tropical Meteorology

P7B.27
MONTE CARLO SIMULATIONS FOR THE UNITED STATES GULF AND ATLANTIC COAST

Gary D. Skwira, Texas Tech University, Lubbock, TX; and D. A. Smith and R. E. Peterson

Hurricane landfalls in the United States have historically caused billions of dollars of property damage and claimed many lives. After a hurricane makes landfall all of society bears the burden of the economic recovery. For this reason it is especially critical that hurricane wind speeds, their associated probabilities of occurrence, and the uncertainties associated with these wind speeds be studied in order for engineers to design for these events. With an increased understanding of these quantities, appropriate structural design provisions can be developed which will result in reduced property damage and fewer fatalities.

Hurricanes are major events, but they happen relatively infrequently at any one location, thus meaningful statistics which can be used for structural design purposes can not be compiled at a single site from the historical record. Thus, an indirect method of obtaining meaningful hurricane wind statistics is required. One such method is known as a Monte Carlo simulation.

A Monte Carlo simulation is a procedure whereby many tropical storms and hurricanes at a selected site are generated using a computer simulation. In this procedure, important meteorological parameters such as the radius of maximum winds, the central pressure, and the speed and direction of the hurricanes motion are randomly selected from probability density functions developed from historic data. The information selected is then used to initialize the wind field model. The hurricane is then propagated and the maximum wind speed and wind direction at the site of interest is captured. This process of initialization of the hurricane wind field, propagation of the hurricane, and capturing the maximum wind speed and direction at the site is performed repeatedly. The resulting database of wind speeds and directions can be used by engineers to establish load factors for the design of structures.

This paper investigates hurricane wind speeds generated using the Monte Carlo simulation method for a number of sites along the United States Gulf and Atlantic Coast. The resulting wind statistics are then presented and a number of interpretations, conclusions, and possible suggestions are discussed.




The 23rd Conference on Hurricanes and Tropical Meteorology