In support of the CSTAR tropical cyclone wind project, David Glenn from WFO Newport, NC collected observations from marine locations off the coast of the mid-Atlantic and Southeast during ten tropical cyclones that impacted the Carolinas and Virginia. From this dataset we examined the sustained winds, wind gusts, gust factors, and wave heights for these locations. A map of the tracks of the tropical cyclones examined and the final set of buoys used in the analysis is shown to the right. We need to acknowledge the great contribution that student volunteer Dan Brown made in completing much of the statistical work.
Hourly observations of winds, wind gusts, and waves from more than two dozen buoys and offshore observational platforms were examined. The final dataset includes only hourly observations with wind speeds of 10 knots of more. In addition, only observations from buoys that have an anemometer height of 5 meters were included to remove any of the variability introduced by different observational heights (the initial database included observations with heights ranging from 2.8 to 44.2 meters). Finally, only observations in which wave data was available with the wave heights less than 5m were included. This was done to remove any uncertainty in the quality of the wind observations in large waves. High sea states associated with high surface winds can shelter the buoy and reduce the buoy’s wind speed observation (Skey et al. 1995). Also worth noting is that large waves may change the vertical wind profile in the boundary layer. In addition, in strong winds, tethered buoys may be shifted out of the vertical and tilted downwind likely making the wind data unrepresentative.
For the ten tropical cyclones examined, the hourly wind gust factor for each buoy location was computed as the ratio of the wind gust to the sustained wind speed. The gust factor is defined as the ratio between the peak wind gust of a specific duration to the mean wind speed for a period of time. More information on gust factors in hurricanes can be found in Krayer and Marshall (1992), Vickery and Skerlj (2005), and Powell et al. (1996).
From this dataset, a total of 1,720 gust factors were computed for all of the storms. The average gust factor was 1.23 with a median of 1.22. The standard deviation was just 0.055 around the mean of 1.23. The wind speeds ranged from 10 to 45 knots and the wind gusts ranged from 11 to 65 knots.
The chart below is a scatter plot of the sustained winds in knots versus gust factors for the 1,720 observations included in the study along with a best fit regression curve (y = 0.0371ln(x) + 1.1148). This chart demonstrates the close relationship between the wind speed and gust factor with a decrease in the number of observations as the wind speed increases and a slight upward trend of the gust factor with increasing wind. This chart is similar to the results of Vickery and Skerlj (2005) which used 10m stations but shows a similar pattern in their scatter plot. The chart below for marine observations differs considerably with a similar study of land based locations for the same ten tropical cyclones.
A histogram of the frequency of gust factors for the ten tropical cyclones examined is shown below. Note the very small range in gust factors with 1,582 of the total 1,720 gust factors (92% of the total) ranging between 1.1 and 1.3 resulting in standard deviation of 0.055 around the mean of 1.23. The marine gust factors are much smaller and less varied than the land locations noted in a similar study of the same ten tropical cyclones for land locations. These results are similar to those offered by Paulsen and Schroeder (2005) which showed that tropical systems in a lower (smoother) roughness regime tended to have lower and less varied gust factors.
It should be noted that this dataset contains a limited number of observations of strong winds with only 9 observations of sustained winds of 40 knots or more, only 15 observations of wind gusts of 50 knots or more, and just one observation of a hurricane force wind gust (65 knots).
We also generated regression equations for each of the storms individually and then combined them into one equation. A plot of the curves for each storm and the combined curve for all storms is shown below. The curve in our data which shows a very slight upward trend in gust factor with wind speed is similar to other studies including Vickery and Skerlj (2005). The increase in gust factors with wind speed may result from increased roughness associated with larger waves in the stronger winds.
Some key points from this analysis include:
1) The marine gust factors which averaged 1.23 are much less than the land gust factors which averaged 1.47.
2) The marine gust factors are much less varied (standard deviation of 0.05) than the land gust factors (standard deviation of 0.30) and with the marine gust factor showing a very slight increase with increasing winds.
3) Given the differences in the land and marine gust factors, different gust factors are needed for land and marine locations.
An experimental Gridded Forecast Editor (GFE) methodology is being tested by nearly a half dozen WFOs in the Southeast uses grids of wind gust factors (you can learn more from a previous blog post – Tropical Storm Andrea Provided an Opportunity to Test New CSTAR Based Tropical Cyclone Wind and Wind Gust Tools). The results from this examination of marine gust factors complements the previous work and will be used to further refine the tools used in GFE.
Krayer, William R., Richard D. Marshall, 1992: Gust factors applied to hurricane winds. Bull. Amer. Meteor. Soc., 73, 613–618. [Available online at http://dx.doi.org/10.1175/1520-0477(1992)073<0613:GFATHW>2.0.CO;2]
Paulsen, B. M., J. L. Schroeder, 2005: An Examination of Tropical and Extratropical Gust Factors and the Associated Wind Speed Histograms. J. Appl. Meteor., 44, 270–280.
[Available online at http://journals.ametsoc.org/doi/pdf/10.1175/JAM2199.1]
Powell, Mark D., Samuel H. Houston, Timothy A. Reinhold, 1996: Hurricane Andrew’s Landfall in South Florida. Part I: Standardizing Measurements for Documentation of Surface Wind Fields. Wea. Forecasting, 11, 304–328. [Available online at
Schroeder, J. L., M. R. Conder, and J. R. Howard, 2002: “Additional Insights into Hurricane Gust Factors,” Preprints, Twenty-Fifth Conference on Hurricanes and Tropical Meteorology, San Diego, California, 39-40.
Skey, S. G. P., K. Berger-North, and V. R. Swail, 1995: Detailed measurements of winds and waves in high sea states from a moored NOMAD water buoy. Proc. Fourth Int. Workshop on Wave Hindcasting and Forecasting, Banff, AB, Canada, 213–223.
Vickery, P.J., and P.F. Skerlj, 2005,: Hurricane gust factors revisited, J. Struct. Eng., 131, 825-832. [Available online at http://www.asce.org/uploadedFiles/Communications-NEW/Hurricane/Hurricane_Gust_Factors_Revisited.pdf]