A Recalculation of MPI Using Upper—Ocean Depth—Averaged temperatures: climatology and Case Studies

Tropical cyclone track forecasts have improved greatly in recent years. However, intensity forecasts still pose a large problem in tropical meteorology. Several theories have been developed over the past fifty years which attempt to arrive at an upper bound or Maximum Potential Intensity (MPI) of tropical cyclones. Emanuel (1986, 1988a,b, 1991, 1995a, 1997), in particular, uses the SST and atmospheric sounding to arrive at a maximum intensity for the hurricane using a Carnot Engine cycle approach to the energetics of the storm. While this approach has captured the upper bound of a hurricanes intensity reasonably well, the use of SST for MPI calculations may be overestimating the maximum intensity. Hurricanes draw their energy from a significant depth of the upper ocean, the average temperature of which is almost always less than the SST. Consequently, there may be more cases of tropical cyclones approaching or exceeding their MPI than an SST-based MPI would dictate alone. Presumably, a purely SST-based MPI calculation would only be valid for intense storms if the entire layer affected by hurricane mixing was isothermal.

A recalculation of the MPI climatology acknowledging this oceanic variability is performed here in an attempt to better isolate those storms that have approached or exceeded a more realistic MPI. Using the NODC (Levitus) World Ocean Atlas Data (1994) oceanic temperatures over five layers to a depth of 50m have been weighted producing a climatology of mean upper ocean temperature over the globe. This new SST is then used in a new calculation of the Emanuel MPI. The weighting of oceanic layers chosen is supported by observations taken during the passage of Felix over the Bermuda testbed mooring on 15 August 1995 (Dickey, et al. 1998) and will be compared to simulations in a one-dimensional ocean model (Kantha and Clayson, 1994) driven by fluxes typical of a tropical cyclone.

A revised MPI climatology for the named tropical systems in the North Atlantic basin from the years 1982-2003 is produced. Results have shown that there is a dramatic impact upon the MPI climatology when the mean upper oceanic temperature is used instead of just SST. With only the skin SST being used for MPI, the only storms to actually exceed their MPI are recurving, poleward moving systems, a consequence of the storm accelerating rapidly and moving over cooler waters while weakening more slowly than the timescale required for the storm to come into thermodynamic balance with the decreasing SST. With the new MPI calculation using a mean oceanic temperature, several storms in the Atlantic basin actually exceed their MPI over waters south of 35° N. Several examples of these will be examined (including Isabel 2003) in an attempt to determine whether the ocean, the atmospheric environment (trough interaction), the internal dynamics (eyewall contraction or vortex Rossby wave forcing) were the cause of intensification.

Further, it was discovered that for strong storms of category three or larger, the most intense estimate of the MPI was not found under the storm at analysis time, but rather 10-11 days before the storm actually reached the location. This shows that strong tropical cyclones begin to affect the ambient environment well before the TC center arrives at the location, and that the thermodynamic stabilization process of the environment begins a week or more prior to a strong TC's passage. It is unclear whether this stabilization is a result of the TC outflow itself, or of the modification of the intensity of the Hadley and Walker circulations in the process of TC generation and intensification.