Open terrain gust factors in landfalling hurricanes based on mobile tower observations

The past decade has seen the deployment of mobile instrumented towers by teams from both Texas Tech University (TTU) and the Florida Coastal Monitoring Program (FCMP) in landfalling hurricanes in the United States. These towers have been designed to withstand hurricane-force winds and make high-resolution measurements of wind speeds in the surface layer in an effort to improve our understanding of the hurricane boundary layer. One of the key parameters of interest is the gust factor, which is the ratio of the maximum gust wind speed measured over some time period relative to the mean wind speed measured over the same time period. Current theory for the prediction of gust factors states that the gust factor is a function of a peak factor that depends only on the gust duration, the time over which the mean wind speed is averaged and the turbulence intensity, the latter being a function of the underlying surface roughness. This implies that for a given combination of gust and mean wind speed averaging times the gust factors measured at different sites with the same turbulence intensity or surface roughness will be the same. This thinking has guided previously published analyses of the gust factors measured by both the TTU and FCMP mobile tower programs, where the gust factors have been analyzed by considering all sites simultaneously and classifying individual gust factor measurements by roughness length. The gust factors have then been grouped by roughness length and the results of the analysis presented on this basis.

In this paper we take a different approach and first consider the gust factors measured at individual FCMP tower sites over the period 1999-2010 by mean wind direction. In general, for most sites there are two primary wind directions associated with the passage of a hurricane which allows two sets of gust factors to be identified, one for each primary wind direction, along with the mean turbulence intensity associated with each wind direction. We then group the gust factors by wind direction and mean turbulence intensity to allow comparisons between sites with the same turbulence intensity. The reasoning behind choosing this method of analysis is that gust factors measured for a given wind direction should reflect the upstream terrain exposure in that direction. Classifying the gust factors by turbulence intensity also removes the need to make an a priori assumption about the relationship between the turbulence intensity and the shear stress within the surface layer to calculate a roughness length as has been done in previous analyses. Preliminary results show that there can be quite significant differences between the observed gust factors for different sites with the same mean turbulence intensity, contrary to what current models for gust factors in hurricanes would suggest. Although the reasons for these differences are still being examined, one logical conclusion would be that non-equilibrium effects due to upstream changes of roughness are important when determining gust factors at individual sites.