Evaluation of Upper Ocean Mixing Parameterizations for use in Coupled Models

The mutual response of hurricane-ocean presents one of the more extreme air-sea interaction events. As the underlying boundary of hurricanes, upper ocean plays a crucial role in the formation and maintenance of these storms. In particular, pre-existing oceanic mesoscale features such as energetic western boundary currents and warm core eddies have been shown to provide the necessary heat source for rapid intensification of hurricanes. The objective of this paper is to understand the relative importance of key physical mechanisms and identify more appropriate mixing schemes within the context of predicting the upper ocean heat content in the ocean component of a coupled prediction system.

Using the Hybrid Coordinate Ocean Model (HYCOM), upper ocean response during three major hurricanes (Gilbert, Isidore and Lili) is investigated in the Gulf of Mexico and Caribbean Sea in this paper. Gilbert (1988) was a category 3 storm in the Gulf of Mexico. Hurricane Isidore (2002) intensified to a Category 3 storm over the high heat content waters of the Caribbean sea and Lili (2002) intensified rapidly over the Loop Current region where there is also strong current variability. In addition, high resolution atmospheric and oceanic measurements were acquired during these storms as part of a NSF/NOAA sponsored research. Therefore, these storms are ideal cases for evaluating the upper ocean mixing schemes and quantifying the ocean response and associated feedback effects to the atmosphere.

Realistic pre-storm initial conditions for HYCOM in this study are derived from in situ observations and the 0.08 North Atlantic HYCOM that routinely assimilates altimetric height anomalies and high resolution sea surface temperatures. Surface wind fields derived using buoy and aircraft measurements are blended with the large scale atmospheric model fields to provide boundary layer forcing every 3 hours for the two storms. Simulated ocean response and resulting fluxes for realistic and quiescent initial conditions for the same forcing will delineate the effects of pre-storm ocean variability. HYCOM configured with 5 upper ocean mixing parameterizations is used to simulate the upper ocean response. Performance of these schemes is evaluated by comparing the simulated fields to observations. Clearly, identifying a more accurate entrainment closure scheme is crucial for the prediction of hurricane intensity in coupled prediction models.