Application of Ensemble Forecasts to the Study of Tropical Cyclone Intensity: A Case Study of Hurricane Katia (2011)

Although tropical cyclone (TC) intensity change under idealized conditions is fairly well understood, accurate forecasts of intensity change for real TCs remain elusive. Having recognized this limitation, the goal of this study is to determine the processes responsible for different TC intensification rates using a large set of equally likely forecasts. A five-day, 96-member ensemble forecast for Hurricane Katia (2011) was produced using the Advanced Hurricane Weather Research and Forecasting (AHW) model. Hurricane Katia posed great challenges for numerical forecasts during its early development, with some operational models predicting a weak storm versus other models predicting a major hurricane. Ensemble forecasts from the AHW model were no exception, with some members predicting that Katia would remain a 35-knot tropical storm, with others predicting a category 2 hurricane. In order to understand what processes lead these different forecast scenarios, two distinct subgroups within the ensemble were identified: 1) 10 members that predicted the lowest time-averaged minimum sea level pressure (named strongest members) and 2) 10 members that predicted a weak storm (named weakest members). Results indicate that all members were initialized with a moderately strong environmental vertical wind shear, but the weakest members had a drier environment than the strongest members. In turn, this moisture difference can subsequently impact the development and maintenance of the convection near the inner-core, which ultimately leads to differences in the intensity forecasts. This hypothesis is evaluated by comparing shear-relative forecast fields and moisture budget terms within the strongest and weakest members. The implications of those results on TC intensity predictability and observation strategies will be discussed.