Real-Time, Ensemble-Based Probabilities of Tropical Cyclone Rapid Intensification

Tropical cyclone rapid intensification (RI) is a significant forecast challenge despite the varied statistical and dynamical guidance available to forecasters. The growing availability of real-time quasi-operational ensemble prediction systems could provide a new source of probabilistic guidance for TC intensity change. The purpose of this talk is to describe the RI probabilities that were computed from HWRF, HMON, COAMPS-TC, and HWRF analog ensemble output computed from retrospective cases from prior seasons and in real-time during 2017. Verification of the ensemble-based guidance indicates that output from some systems, including the HWRF and COAMPS-TC systems, do not produce as skillful of probabilistic RI predictions as statistical models (i.e., SHIPS, DTOP), but do provide reliable probabilities of other intensity change categories. Individual case studies, such as Hurricane Harvey, indicate that dynamical model probabilities can have good ability to forecast RI. The presentation will also include our plans for future seasons, which include expanding the intensity change probabilities beyond just RI and multi-model ensemble-based probabilities.