Forecasting Post-Extratropical Transition Outcomes for Tropical Cyclones Using Support Vector Machine Classifiers

Abstract

ntensity changes following the multistage process of extratropical transition have proven to be especially difficult to forecast because of the extremely similar storm evolutions prior to and during the first stages of the transformation from a warm-cored axisymmetric tropical storm to a cold-cored asymmetrical extratropical low pressure system. In this study, differences in surrounding synoptic environments between dissipating and reintensifying extratropical transitioning tropical cyclones are used to develop a predictive technique for extratropical transition intensity change that can be used to enhance the standard numerical guidance. Using a set of all historical transitioning storms between 2000 and 2008 in the western North Pacific, common differences between 850-hPa potential temperature fields surrounding extratropical transition intensifiers and extratropical transition dissipaters, respectively, were identified. These features were then used as inputs into a support vector machine classification system in the hopes of creating a robust prediction system. Once the system was trained on a random subset of the data (80%), performance was tested on the remaining test set (20%). Overall, it was found that the prediction system was able to correctly predict extratropical transition intensity outcome in >75% of the test cases at 72 h prior to extratropical transition. This paper discusses the feature selection and classification system used, as well as the performance results, in detail.