Tropical cyclone track forecasts have improved significantly over the past decades due to advances in numerical modeling and an increase in the amount of remote and in situ observational data. However, there has been limited improvement in tropical cyclone intensity forecasts, especially for storms that exhibit rapid intensification and/or weakening.

From a forecast perspective, a particularly problematic case was that of hurricane Lili, 2002. Operational models were able to predict the track of Lili. However, they failed to predict the rapid intensification of the storm over a 24-hour period from a category 2 hurricane to a category 4 hurricane with maximum winds of 125 kts. Furthermore, operational models failed to predict the even more rapid weakening of the system during its subsequent 13-hour traverse over the Gulf of Mexico to landfall on the Louisiana coast. During this time, the storm weakened to a category 1 hurricane with maximum winds below 80 kts.

A series of numerical modeling experiments using the Coupled Ocean/Atmosphere Mesoscale Prediction System (COAMPSTM) have been conducted in order to identify and understand the contributing factors and physical processes that lead to the rapid decay of hurricane Lili prior to landfall. The results of a control simulation indicate that the model successfully forecasts the rapid deepening of the cyclone but maintains the peak intensity of the storm until landfall, in contrast to the observed rapid weakening. Additional model simulations with different environmental conditions and physical parameterization methods are performed and evaluated using observational data collected for this storm. The details of this numerical study will be presented at the meeting.