Enhancements to the SHIPS Rapid Intensification Index

Predicting episodes of tropical cyclone rapid intensification (RI) remains one of the highest operational forecasting priorities of the National Hurricane Center (NHC). In recent years, a statistically based rapid intensification index (RII) that employs predictors from the SHIPS model has been developed utilizing linear discriminant analysis. The SHIPS-RII provides estimates of the likelihood of RI over the succeeding 24-h for both the Atlantic and eastern North Pacific basins and is currently utilized as an operational forecasting tool by the NHC. Although the current operational version of the SHIPS-RII has generally been skillful, its utility has been somewhat limited since the original version was developed exclusively for the 24-h lead-time and its skill has tended to be somewhat limited particularly in the Atlantic basin. Thus, in an effort to improve the overall forecasting utility of the current operational SHIPS-RII, a number of model enhancements have been developed. First, ensemble-based versions of the RII that employ both the current SHIPS-discrimant RII as well as newly developed Bayesian and Logistic versions have been derived for the 24-h operational forecast lead-time as well as the added lead times of 12-h, 36-h and 48-h to provide additional guidance during the critical watch and warning period that has recently been extended to 48-h by the NHC. In addition, new versions of the deterministic rapid intensity aid guidance that provides intensity forecasts utilizing both existing operational deterministic intensity forecast models and the probabilistic RII have been developed utilizing the new multi lead-time ensemble RII. Lastly, microwave imagery-based versions of the RII that have been shown to be capable of providing a more accurate measure of the overall inner-core tropical cyclone structure have also been developed. In this poster, the results of independent tests of the aforementioned multi-lead time Atlantic and E. Pacific basin version of the SHIPS that were conducted by re-running the RII using real-time model forecast data for the period from 2004-2013 will be presented. In addition, a verification of the performance of a revised versions of the recently derived deterministic rapid intensity aid guidance that has been developed utilizing the new multiple lead-time ensemble-based RII models will be shown for the period 2008-2013 when all of the guidance required to run the rapid aid guidance were available. Finally, an evaluation of the performance of a real- time version of the microwave-based RII for the 2013 Atlantic and Eastern Pacific hurricane seasons as well as for retrospective tests that were performed using real-time data from the 2008-2012 hurricane seasons will be presented in a separate poster presentation by Rozoff et al. at the upcoming meeting.