The Hurricane Research Division's (HRD) Tropical Cyclone Windfields at Landfall experiment collects data in landfalling tropical cyclones from NOAA aircraft to improve real-time and post-storm surface wind analyses that are prepared by HRD. Analyses of flight-level data are provided in real time to forecasters at the Tropical Prediction Center. After the storm, airborne Doppler radar data are combined with National Weather Service WSR-88D radar data in three-dimensional analyses to document the evolution of tropical cyclones at landfall, and to provide corroborating data for testing WSR-88D tropical cyclone algorithms.
The experiment was flown in Hurricane Danny on 18 October 1997, as the Category 1 storm moved slowly towards Mobile Bay. Because parts of the hurricane were overland, the aircraft remained at 4.3 km (14,000') for the entire flight. The pattern included several legs along radials from the New Orleans (KLIX) and Mobile (KMOB) Doppler radars. Twelve GPS sondes were launched near marine and coastal surface platforms, to provide additional details of the storm windfield at low levels. During the flight, Hurricane Danny remained in the KLIX-KMOB dual Doppler coverage.
Preliminary analyses of the airborne Doppler radar data collected from 1400 to 1500 UTC during the first two passes through the storm show winds > 30 m/s at 1 km altitude, with embedded regions > 35 m/s and a 17 km radius of maximum winds (RMW). A multiple Doppler analysis incorporating data from KLIX, KMOB,and the aircraft, centered at 1848 UTC, shows that the maximum winds at 1 km remained > 35 m/s, but the RMW may have increased. Independent confirmation of
the low-level wind maximum was provided by a GPS sonde that was dropped in the southern eyewall at 1928 UTC; winds were 30 m/s below 1 km, and there was a maximum wind of 37 m/s only 200 m above the surface. The multiple Doppler analysis agrees well with single Doppler analyses from KLIX and KMOB provided by the Ground-Based Velocity Track Display (GBVTD).
GBVTD analyses can be generated 10-12 times per hour, filling in times when the aircraft was not in the storm. At the conference we will show time-series of windfields synthesized from the airborne and WSR-88D radars. We will discuss the problems of combining airborne and ground-based Doppler radar data, and consider how the Doppler derived winds are combined with HRD's surface wind analyses.