Rapid tropical cyclone transitions to major hurricane intensity: structural evolution of infrared imagery

A subset of rapidly intensifying Atlantic tropical cyclones that begin as strong tropical storms or weak hurricanes and transition to major hurricanes in twenty-four hours (h) are investigated using an archive of GOES Infrared (IR) imagery. The twenty-four cases begin with a mean intensity of approximately 60 kt (31 ms-1) and in 24h ends with an intensity of approximately 108 kt (56 ms-1) – 2 T-numbers on the Dvorak intensity scale. The subset was purposely designed to represent the real forecast problem where a relatively benign tropical cyclone, which requires little preparation and mitigation, rapidly becomes a life and property threatening event.

To examine the evolution of IR imagery a 48-h period of imagery is collected encompassing the 24h before and during the rapid transition to major hurricane. The storm-centered IR imagery is filtered spatially and gridded on a 10o x 4 km cylindrical grid. To account for the satellite eclipse periods and the occasional missing images the cylindrically gridded data are also interpolated to equal half-hourly intervals. A method called “Complex Empirical Orthogonal Functions” (CEOF) is then applied to the resulting equally temporally spaced and spatially filtered dataset. The CEOF analysis produces spatial and temporal amplitude and phase functions that describe the mean evolution of these rapidly evolving events. The details of this evolution along with ideas as how one may utilize this type of information to make better intensity forecasts will be presented.