Relationship of ice content to hurricane intensity examined using TRMM/TMI data sets

Summary ?In this study, TRMM Microwave Imager (TMI) data were used to investigate a possible relationship between the total ice contents of Atlantic basin hurricanes to their intensities. Ice content data from the TMI was gathered for all available TRMM swaths for the ten hurricanes. These data were then correlated to the observed maximum wind speeds and minimum central pressures from the storms at the respective times of the TRMM swaths to examine the relationship. Several methods of correlating the ice content data to the observed data were tried and the most successful method involved the use of multiple linear regression to compute correlation coefficients using the observed intensity and TMI ice content data as well as storm size. This method allowed for the inclusion of storm size, which had been found to be an important factor in total ice content. These correlation coefficients were then used to produce regressed intensities for the hurricanes and these regressed intensities were then correlated to the observed intensities. The result was a correlation coefficient of 0.437 for the relationship between the regressed maximum wind speeds and observed maximum wind speeds and 0.483 for the regressed minimum central pressures and the observed minimum central pressures. Though not entirely conclusive, the results do appear to substantiate the possibility of an ice content-intensity relationship within Atlantic basin hurricanes. With further investigation, it may be possible to use ice content as another determinant in league with the Dvorak method and other methods of tropical cyclone intensity in regions that are out of the range of surface or aircraft observations. Received April 30, 2002; accepted July 18, 2002 Published online: March 20, 2003