Wednesday, 24 May 2000: 2:45 PM
We describe a simple hurricane intensity prediction model, consisting of an axisymmetric,
balanced hurricane model coupled to a series of one-dimensional ocean columns. Storm winds
cool the surface temperature of each column, according to a bulk Richardson number-based
mixing closure; the reduced sea surface temperature is then fed back to the hurricane model. The
model is initialized with a synthetic vortex and the subsequent track is specified from the official
NOAA/TPC forecast. Environmental atmospheric and initial ocean conditions are specified from
monthly mean climatology, and landfall is simulated by turning off evaporation when the storm
center crosses the coast.
While the model performs quite well in hindcasts of many storms, with the actual track specified,
real-time application during the 1999 Atlantic hurricane season showed mixed results and
interesting limitations of the model. In most cases, the model overpredicted the storm intensity,
with the degree of overprediction apparently related mostly to the magnitude of the vertical shear,
though in one case (Floyd) the overprediction was at least partially owing to the cold wake left by
a pervious storm (Dennis), not accounted for in the present model set up. Implications of both the
hindcasts and the forecasts for general prediction of hurricane intensity will be discussed.
- Indicates paper has been withdrawn from meeting
- Indicates an Award Winner