Note this is a multi-part review of Arthur, with the focus in this post on the creation the wind gust forecasts.
To evaluate the performance of CSTAR related research to operations activities, we examined the sustained winds, wind gusts, and gust factors for Hurricane Arthur (2014) across coastal North Carolina. The map below is a subjective analysis of the maximum wind gusts observed during Hurricane Arthur. Many locations across and near the Outer Banks reported wind gusts in excess of 70 MPH with a few locations in the southern and central Outer Banks reporting wind gust greater than 90 MPH. The western edge of the enhanced wind gusts directly associated with Arthur extended to near Interstate 95 with values of around 25 MPH.
Hourly observations of winds and wind gusts from 16 regular ASOS or AWOS METAR locations impacted by the over land wind field associated with Hurricane Arthur were examined. The locations examined in this analysis include KECG, KEDE, KEWN, KFFA, KHSE, KMQI, KMRH, KNBT, KNCA, KNJM, KNKT, KOAJ, KOCW, KONX, KPGV, and KSUT. Only observations from routine hourly METARs were used (special observations and observations not at the top of the hour were excluded). In addition, gust factors were only calculated for sustained winds of 10 MPH or greater. For each observation, the hourly wind gust factor was computed. The gust factor is defined as the ratio between the wind gust of a specific duration to the mean (sustained) wind speed for a period of time. A total of 344 gust factors were computed for Arthur.
Note that a previous blog post highlighted the examination of gust factors for ten tropical cyclones across the Carolinas and Virginia – A Wind Gust Factor Database from 10 Tropical Cyclones for Use with GFE Tool Development. This previous post discusses the creation of the database used to develop a regression equation included within a GFE smart tool that is used to create a wind gust factor grid. The discussion below examines the winds, wind gusts, and gust factors associated with Arthur and compares them to the database of 10 tropical cyclones.
The chart to the right is a scatter plot of the sustained winds in MPH versus gust factors for the 344 observations included in the study along with a best fit regression curve (y = -0.274ln(x) + 2.3599). In general, the chart demonstrates an inverse relationship between the wind speed and gust factor as well as a decrease in the number and variability of observations as wind speeds increase. This chart is very similar to the scatter plot created from the database of the 10 tropical cyclones.
The gust factors with Arthur were rather variable at low sustained wind speeds but generally converged and decreased with increasing sustained wind speed. The maximum sustained wind contained in the Arthur data set was only 48 MPH and the maximum wind gust was 71 MPH. The data set contained a large number of lower end sustained winds and wind gusts. More than 69% or 239 out of the 344 observations contained in this data-set had sustained winds less than 20 MPH. Only 30 observations or less than 9% of all observations had sustained winds of 30 MPH or more.
A histogram of the frequency of gust factors for Arthur across the North Carolina is shown to the right. The average gust factor for Arthur was 1.58 which is somewhat higher than the average of 1.47 for the database of 10 tropical cyclones. The average gust factor for Arthur was identical to the average gust factor for Hurricane Sandy. Both Arthur and the 10 tropical cyclone database had gust factors that were very similar to several previous studies (see comparison chart from a previous blog post). Schroder (2002) found an average gust factor of 1.49 during tropical cyclones at several airport locations while Krayer and Marshall (1992) found an average gust factor of 1.55 in a data set standardized to open terrain. The histogram data noted that the gust factors were most frequently in the bin between 1.5 and 1.6. While the distribution of gust factors for Arthur is shifted to the right with higher gust factors when compared to the 10 storm database, the pattern and character of the distribution for Arthur is very similar to the 10 storm study.
Forecasters at NWS offices in Charleston, Newport, Raleigh, and Wilmington tested an experimental GFE methodology during Hurricane Arthur based on research activities associated with the CSTAR project. In this methodology, the forecaster initially populates a WindGustFactor grid based on the regression equation derived from the dataset of 10 tropical cyclones that uses the sustained wind speed as an input. The forecaster can then adjust the WindGustFactor grid to account for local effects such as boundary layer stability, prior to calculating the wind gust. The product of the wind forecast and the WindGustFactor is then computed as the wind gust forecast.
The example to the right is the 23-hour NDFD wind gust forecast from midnight EDT on 3 July valid at 11pm EDT on 3 July, 2014 which demonstrates a consistent and well collaborated wind gust forecast from the 4 WFOs using the new methodology. The area to the right or east of the thin yellow line encompasses the WFOs that used this experimental methodology. It is difficult to identify the CWA borders among the 4 offices within the yellow semi circle but the CWA borders among other WFOs or between participating and non-participating WFOs straddling the yellow line can be more easily identified. During Arthur, forecasters provided positive feedback on this methodology and noted the much improved consistency and an improved quality of the forecast wind gusts using this approach when compared to past experiences. This event demonstrated a notable CSTAR research to operation success of the new CSTAR motivated methodology.
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]
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.