Coupled Air-Sea Interactions During Hurricane Bonnie

The passage of hurricane Bonnie in 1998 in the western Atlantic Ocean basin underscored the uncertainties in predicting wind-field changes of tropical cyclones. Bonnie slowly drifted towards the north-northwest in this experimental domain as manifested by the marked upper ocean heat loss of about 16 Kcal cm^-2 (or four times the value required to sustain a tropical cyclone) where the storm had a maximum intensity of about 50 m s^-1 and central pressures in the 950-960 mb range. Prior to this intense heat loss, the heat content change was about 4 Kcal cm^-2, but as the storm slowed down, Ekman divergence in the upper ocean surface layer (i.e. upwelling) coupled with the enhanced air-sea fluxes caused a significant heat loss. As the storm started to accelerate, the winds decreased to about 45 m s^-1, and the maximum upper ocean heat loss was about 8 Kcal cm^-2. The track of Danielle actually intersected Bonnie's wake, which may have been one reason that Danielle did not significantly intensify in this region as the heat source was sufficiently depleted. Danielle's intensification began only after the inner core crossed one of the cold patches created by Bonnie and entered deeper, warmer water.

Of particular interest was the APBL structure detected by the GPS sondes dropped in Bonnie's eyewall. Here, the vertically averaged flow was removed from the observed wind profile to show the perturbation wind field. The perturbation wind vectors (typically 10 m s^-1) indicate a clear upward anticyclonic spiralling of the winds with height. There is also considerable structure in these profiles which perhaps implies baroclinicity in the wind perturbations since the height-averaged wind was removed. These wind profiles are markedly similar to oceanic current profiles where energy propagates downward from a wind-forced OPBL (Shay et al. 1998). Given this wind vector rotation theory predicts upward propagation of group velocity and hence energy flux from the near-surface layer through the APBL. The equivalent potential temperatures (i.e. theta_e) are about 360° K in the area of maximum heat loss in the upper ocean. Although the correspondence is not necessarily one-to-one, along-track variations of theta_e are apparently correlated to these upper ocean heat content changes. This result suggests more of a positive feedback from the ocean to the atmosphere rather that differs from previously held notions that the oceans only provide negative feedback. These profiles and their relationship to the atmospheric and oceanic environments represent a unique opportunity to examine their relative importance on both tropical cyclone structure and intensity variations.