Max Mayfield wrote in the Guillermo report, “Profiles of wind speed versus altitude showed considerable variations among the individual ‘drops’. Figure 4 [reproduced below] shows a profile from one of the GPS sondes dropped within the southwest quadrant of the eyewall at 2342 UTC. In this example, it is noted that the winds are strongest in the low levels, and in fact, are a little stronger [italics in the original] at the surface than at flight level. These data will continue to be studied to better understand the relationship between flight-level winds and surface winds. In-depth analyses from HRD scientists are in progress, and publication of these analyses are eagerly awaited.”
Before this, the main technique to estimate tropical cyclone intensity from aircraft reconnaissance was to reduce flight-level winds by a certain percentage, depending on the altitude of the aircraft. A year later, dropwindsonde observations in the eyewall of Atlantic Hurricane Georges were used to operationally assess that the hurricane had a maximum sustained wind speed of 135 kt, just below category 5, and specialists noted that the surface wind speed was equal to or greater than those at the 3-km flight level. When Georges made landfall on the Mississippi coast a week later, NHC forecasters noted that the surface wind speed was 20-30% below those at the 3-km flight level.
The great variability eventually led to the Franklin et al. (2003) paper that quantified the mean wind speed profile in the hurricane eyewall and the expected variability about that mean. It also led to a reassessment of the best tracks of multiple tropical cyclones, including the famous upgrade of Hurricane Andrew (1992) to category 5 at landfall near Homestead, FL. This study garnered the authors a Department of Commerce Gold Medal.
Franklin, J. L., M. L. Black, and K. Valde, 2003: GPS dropwindsonde wind profiles in hurricanes and their operational implications. Wea. Forecasting, 18, 32–44.