Real-time Verification of Passive Microwave Imagery-Based Statistical Models of Tropical Cyclone Rapid Intensification

Recent studies show promise in including passive microwave data from low-earth orbiting satellites in probabilistic forecasts of tropical cyclone (TC) rapid intensification (RI). As part of a Joint Hurricane Testbed (JHT) project, a real-time suite of these RI models have been developed, including microwave imagery-enhanced probabilistic RI forecast schemes for the North Atlantic and Eastern Pacific Ocean basins based on a logistic regression model. These models predict the probability of RI occurring in a 24-h period and include various RI thresholds (i.e., 25, 30, and 35 kt per 24 h period). The models were derived from climatological data describing the storm's intensity trend, its environment, and structure. The structure predictors are obtained from global reanalysis data, satellite infrared imagery, and passive microwave data. The microwave predictors were developed from 19, 37, and 85 GHz passive microwave brightness temperatures from the Advanced Microwave Scanning Radiometer – Earth Observing System (AMSR-E), the Tropical Rainfall Measuring Mission (TRMM) Microwave Imager (TMI), and the DMSP Special Sensor Microwave/Imagers (SSM/I and SSMI/S) for the period of 1998-2012.

Independent testing of the microwave imagery-enhanced logistic regression model in both the Atlantic and East Pacific Ocean basins shows that adding selected microwave-based predictors improves both the Brier skill score and reliability of RI forecasts significantly. In this poster presentation, we will show how a real-time version of this model performed in the 2013 Atlantic and Eastern Pacific hurricane seasons. This evaluation is augmented with retrospective tests using real-time data from the 2008-2012 hurricane seasons.