Statement from Dr. Mark Powell in response to NHC Best Track Committee assigning a 145 kt wind speed to Hurricane Andrew at landfall in South Florida.

1. It is very important to accurately and objectively assess the intensity of hurricanes during landfall. The extreme wind climate is based on this information and has an influence on wind load standards used in building codes as well as the risk associated with insurance rates. External scientific oversight and assistance from the atmospheric science, oceanographic, and wind engineering communities is needed to help drive a well-balanced assessment.

2. I am in agreement with the assessment of Hurricane Andrew as a Category 5 storm during periods when it was over the open ocean in deep water . HRD research on boundary layer structure determined from GPS sondes indicates that the sea surface becomes covered with foam at very high wind speeds and that the roughness of the sea surface actually decreases. The HRD hurricane wind analysis system has a method (based on GPS sonde research) to estimate surface winds from flight level measurements at 10,000 ft. Applying this method to the Andrew reconnaissance aircraft data results in sustained maximum surface winds over open sea of 150 kts (172 mph), consistent with the methods used by NHC. The uncertainty of this estimate is ~ +/- 22 kts (25 mph).

3. However, once Andrew reached shallow waters, waves generated by the strong winds became closer together and steeper, causing a much rougher surface than over the deeper ocean further offshore. Limited GPS sonde data near land are consistent with this process. Therefore the winds over areas where waves are breaking on the outer reefs, the bays, and shorelines, would tend to be flowing over a rougher surface than over the open ocean, leading to a wind speed decrease.

4. The highest official wind measurements from Fowey Rocks (about 4 miles southeast of Key Biscayne) are consistent with this condition but the instrument failed while in Andrew's north eyewall. An analysis of aircraft winds adjusted to the surface based on the new GPS research (for open ocean conditions) shows overestimates of ~ 29 % at Fowey Rocks and ~ 20 % in Perrine.

5. I disagree with the estimate of Andrew as a Cat 5 storm during any point of its history when over shallow water or land. I believe that Andrew's wind speeds were consistent with a strong Cat 4 storm at landfall in south Florida, or ~ 132 kts (152 mph). However the uncertainty of this estimate is high, +/- 26 kts (30 mph), since we know very little about sea surface roughness in extreme winds.

6. Research is needed to investigate how shoaling and breaking waves influence the wind over shallow water. Additional GPS sonde measurements are needed in hurricanes with strong winds over deep and shallow water. New remote sensing instrumentation must be evaluated and calibrated for use in this type of condition.

7. Additional research is needed to document: a) the transition of the flow from open sea to rough sea, and finally to variable terrain over land, b) the turbulent wind structure in landfalling hurricanes, and c) to determine how the wind changes with height in coastal areas.

8. The HRD Hurricane Field program in cooperation with our federal agency and university partners will attempt to answer these questions in the coming years.

More information on Dr. Powell's hurricane research and PDF files of his peer-reviewed publications