Characterization of hurricane gust factors using observed and analytical data

Date

2009-05

Journal Title

Journal ISSN

Volume Title

Publisher

Texas Tech University

Abstract

The nature of turbulence in the hurricane boundary layer has been the subject of much discussion. Two questions in particular continue to be the source for debate and ongoing research. The first question is whether or not hurricane GFs exhibit the same behavior as GFs from winds generated by extratropical systems (thunderstorms excluded). The second question is whether the structure of the wind, and the resulting gust factors, change at high wind speeds. This study seeks to address those two questions using a variety of data sources and analysis techniques. Observational data were collected from both landfalling tropical cyclones and synoptically generated extratropical wind. Analytical data at a variety of wind speeds were created using an inverse fast Fourier Transform of the universal spectrum for wind in the perturbed terrain. Gust factors and other parameters were computed for both types of data and the results assimilated in a data base. Analysis of these data yielded interesting results. A strong dependence on surface roughness was noted for gust factors from both observed and analytical data. However, once efforts were made to control for this dependency by stratifying the data into roughness regimes using the roughness length, slight differences between the tropical and extratropical gust factor data remained. Analysis of the artificial data, suggest spectral differences between the tropical and extratropical regimes due to the presence of additional low-frequency energy in the tropical regime. A slight decrease of the gust factor with increasing wind speed was noted in the high-speed analytical data. A similar decrease was suggested in the tropical data. It was concluded that the low-frequency spectral differences between the two regimes have less of an effect on the resulting gust factors as the wind speed increases, resulting in better agreement between the two distributions.

Description

Citation