The tropical cyclone-ocean interaction was investigated using a high-resolution tropical cyclone ocean coupled model. The model design consisted of the NOAA Geophysical Fluid Dynamics Laboratory tropical cyclone prediction model which was coupled with a multilayer primitive equation ocean model. Coupling between the hurricane and the ocean models was carried out by passing into the ocean model the wind stress, heat, and moisture fluxes computed in the hurricane model. The new sea surface temperature (SST) calculated by the ocean model was then used in the tropical cyclone model. A set of idealized numerical experiments were performed in which a tropical cyclone vortex was embedded in both easterly and westerly basic flows of 2.5, 5, and 7.5 m s-1 with a fourth experiment run with no basic flow specified initially. The profile of the tangential wind for Hurricane Gloria at 1200 UTC 22, September 1985 was used as the initial condition of the tropical cyclone for each of the experiments. The model ocean was initially horizontally homogeneous and quiescent. To clarify the impact of the ocean response to the hurricane's behavior, analogous experiments were also carried out with the SST kept constant (control cases).

The experiments indicated that the cooling of the sea surface induced by the tropical cyclone resulted in a significant impact on the ultimate storm intensity due to the reduction of total heat flux directed into the tropical cyclone above the regions of decreased SST. The sea surface cooling produced by the tropical cyclones was found to be larger when the storms moved slower. In the experiments run without an initial basic flow, the maximum SST anomaly was about -5.6 ¿C with a resulting difference in the minimum sea level pressure and maximum surface winds of 16.4 hPa and -7 m s-1, respectively. In contrast, in the experiments run with the 7.5 m s-1 basic flow, the maximum SST anomalies ranged from about 2.6¿ to 3.0 ¿C with a difference in the minimum sea level pressure and maximum surface winds of about 7.3 hPa and -2.7 m s-1. The tropical cyclone-ocean coupling significantly influenced the storm track only for the case with no basic flow and the 2.5 m s-1 easterly flow. In these cases the storm with the ocean interaction turned more to the north and east (no basic flow) or the north (2.5 m s-1 easterly flow) of the experiments with constant SST. In the first case, the storm by 72 h was located over 70 km to the east-southeast of the control case. A possible explanation for this track deviation is related to a systematic weakening of the mean tangential flow at all radii of the storm due to the interaction with the ocean and resulting alteration of the beta drift. ¿ American Geophysical Union 1993