Interpreting dropsonde measurements of turbulence in the tropical cyclone boundary layer

Some of the most dangerous winds on earth occur in tropical cyclones. Their destructive power depends as much on their gustiness, or turbulence, as on their mean strength. Further, the efficiency with which energy is extracted from the ocean to power the tropical cyclone depends strongly on the boundary layer turbulence. Thus understanding turbulence in tropical cyclones is crucial. At one level, this understanding should be a simple problem because stability effects become small as the wind speed increases. At another, the problem is complex because the strong winds severely modify the air-sea interface. At a third, testing theoretical predictions of turbulence, or extrapolations of parameterisations from lower wind speeds, is problematic due to the severe difficulty in taking observations where it matters, in close to the interface.

Over the last decade, U.S. hurricane reconnaissance aircraft have deployed thousands of GPS dropsondes in and near the eyewall of tropical cyclones. These instruments parachute towards the surface, reporting back wind, temperature, humidity and pressure at ½-second intervals, or about every 6 m vertically. These measurements have had a considerable impact on our knowledge of the mean structure of the tropical cyclone boundary layer. Their high sampling rate implies that there is a considerable potential to illuminate the turbulent structure as well. However, this potential has as yet not received much attention.

This talk will present a theoretical framework for interpreting the turbulence signal within these measurements. Proper interpretation requires close attention to the response characteristics of the instrument, careful quality control, and a suitable theoretical description of 3-dimensional surface layer turbulence. The spectral velocity tensor formulations of both Kristensen et al. (1989) and Mann (1994) will be used for the latter.

Several useful results can be inferred from the comparison of theory and measurement. The marine drag coefficient does not increase indefinitely with wind speed, thus confirming recent direct measurements and extending those results to much higher wind speeds. A method for estimating gust characteristics from the dropsonde data also follows fairly directly. The theory helps answer the question of what averaging period is represented by the dropsonde measurements. Finally, the scope for using similar calculations to estimate the heat and moisture fluxes might also be discussed.