The relationship between flight level and 10-m winds in numerically simulated landfalling hurricanes

Because of safety considerations and instrument limitations, a hurricane's sustained surface (10 m) wind is difficult to measure (Franklin et al., 2003). Reconnaissance aircraft typically fly at 700 hPa and, therefore, the forecaster must estimate the surface winds from observations made at flight level using ‘reduction factors'. Franklin et al. (2003) used global positioning system (GPS) dropwinsondes to calculate reduction factors in the eyewall and outer core regions of real hurricanes. They used averages of 630 GPS dropsonde profiles from 17 different storms. A modeling study will allow calculation of averages of many soundings within a single storm. Different storms can then be compared. In this study, reduction factors in different model storms making landfall on different land surfaces are compared.

The storms all make landfall on a straight east-west oriented coastline with low, flat terrain. However, the land surface, forward speed, attack angle, and intensity of the storms are different.

In a first set of simulations, a weakening storm moves at around 4 m s-1 in a northeasterly direction in response to the evolving environmental steering flow. At t = 15 h into the simulation, the storm centers cross the coastline about 400 km east of the original location. The storms fill slowly as they approach land and continue to do so after the centers cross the coastline. The 10m windspeed drops off abruptly as the centers cross the coastline, while at 700 hPa decreased gradually. At 700 hPa, the maximum wind speed is consistently located in the southeast. At the surface, the maximum remains in the southeast until the hurricane makes landfall; then the maximum wind speed varies its location from the southwest to the southeast. This will have a significant impact on the reduction factors at various locations within the storm. Reduction factors will be explored spatially as will their evolution over time. An attempt will be made to obtain time-lagged adjustment factors that relate a given location at 700 hPa at sea to that same location at 10 meters some time later over land. Using statistical analyses, the optimal time lag for accurate prediction of the 10-m wind over land will be determined. Adjustment factors will be stratified by their radial and azimuthal location in the storm, storm characteristics, and land surface characteristics. This information will ultimately provide forecasters with useful guidance in the use of the appropriate reduction factors to forecast 10m winds over land from flight-level observations over water.