Synthesis of common structure features at landfall in Hurricanes Katrina and Wilma using real-time airborne and ground-based observations

Three new airborne observing technologies were synthesized in near-real time to obtain the time evolution of Hurricane Katrina's structure and intensity as it approached the Louisiana and Mississippi coasts. In situ GPS dropsondes, the new Stepped Frequency Microwave Radiometer and airborne Tail Doppler radar on NOAA WP-3D aircraft in addition to flight level observations at 3 km from a pair of WC-130J Air Force Reserve hurricane reconnaissance aircraft are used to synthesize storm structure prior to landfall. The observations show a remarkable transformation from a compact CAT4/5 storm late on Sept 28 to a broad, considerably weaker storm during the landfall process on Sept 29. A very similar structure is shown to exist as Wilma approached landfall over southern Florida, using high density surface observations in addition to the aircraft data. Wilma's transformation from the most intense CAT 5 system on record to the large, broad system that struck South Florida extended over a much longer time span than Katrina, and was aided by nearly 3 days over the northeast tip of Yucatan. In both cases, flight level winds at 700 mb in excess of 65 m/s coexisted with surface winds nearly 40% weaker. The TA Doppler radar showed the existence of a 700 mb jet, the SFMR showed the relatively low surface winds, and the GPS sondes measured the highly sheared boundary layer that accompanied this structure.

WSR 88D Doppler radar data within 20 km of the Slidell and Miami radars demonstrated the existence of boundary layer wind streaks, possible ‘roll vortices', in both Katrina and Wilma. These features were aligned with the wind with cross-wind scales that varied from 500 to 1500 m and along-wind scales of several km. These features had trough to crest velocity amplitudes of 10-15 m/s. At the surface these perturbations would have had periods of 4-6 minutes in Katrina and 1-2 minutes in Wilma, determined by the factor of two difference in storm speed between the two storms. Implications of these structure features for estimation of peak surface sustained winds at landfall will be discussed.