A Preliminary Analysis of the RIPA and SPICE Models for the 2019 Hurricane Season

Statistical-dynamical techniques for predicting the intensity change of tropical cyclones have long been a component of the forecast guidance suite at the National Hurricane Center (NHC), Central Pacific Hurricane Center (CPHC), and Joint Typhoon Warning Center (JTWC). This presentation focuses on two experimental statistical-dynamical products run in real-time for the 2019 hurricane season as part of the Hurricane Forecast Improvement Project (HFIP): the Rapid Intensification Prediction Aid (RIPA) and the Statistical Prediction of Intensity from a Consensus Ensemble (SPICE).

RIPA, which is run operationally at JTWC for the western North Pacific, northern Indian Ocean, and Southern Hemisphere hurricane basins, was run experimentally in real-time for the Atlantic, eastern North Pacific and central North Pacific hurricane basins for the 2019 hurricane season. RIPA combines forecasts of the probabilities of rapid intensification from linear discriminant analysis and linear regression techniques into a consensus probability for several different rapid intensification thresholds. Any of these thresholds reaching a probability of 40% produces a deterministic forecast of linear intensification to that threshold (e.g. 30 knots in 24 hours), with the highest threshold that triggers the deterministic forecast being labelled the RIPA forecast, and included in an operational intensity consensus at JTWC.

SPICE was run experimentally in real-time for all global basins for the 2019 hurricane season. SPICE combines forecasts of the Statistical Hurricane Intensity Prediction Scheme (SHIPS) run from the Global Forecast System (GFS), Hurricane Weather Research and Forecasting (HWRF) model, and Hurricanes in a Multi-scale Ocean-coupled Non-hydrostatic (HMON) model into an unweighted ensemble mean. SPICE also produces an unweighted ensemble mean of runs of the Logistic Growth Equation Model (LGEM) from GFS, HWRF, and HMON. These two ensemble means are then combined into a weighted consensus forecast which favors SHIPS at shorter forecast lead times and LGEM at longer lead times, to produce a forecast for intensity change out to five days.

Preliminary verification using the working best tracks for the 2019 hurricane season will be presented. Additionally, the environmental and storm-based predictors that served as input to the models will be assessed.

Disclaimer: The views, opinions, and findings contained in this article are those of the authors and should not be construed as an official National Oceanic and Atmospheric Administration (NOAA) or U.S. Government position, policy, or decision.