261
Satellite image characteristics of intense tropical cyclones

- Indicates paper has been withdrawn from meeting
- Indicates an Award Winner
Monday, 18 January 2010
Exhibit Hall B2 (GWCC)
Alan Schwartz, NOAA/NESDIS, Camp Springs, MD; and R. Weldon

A set of 400 tropical cyclones was collected from 1998 to 2006 for a large research study on the Changes of Tropical Cyclone Behavior related to Changes of the Upper Air Environment. From that set 26 strong tropical cyclones that reached intensities of 140 knot or greater were examined for image characteristics. A special type of storm with very small eye diameters, very cold CDO top temperatures and very rapid intensification was found. Four of these “Wilma Type” storms were identified, Wilma in the Caribbean, Dianmu in the western Pacific, Heta in the Southern Hemisphere, and Sidr in the Bay of Bengal. CDO top temperatures were found to be related to storm latitude, although the “Wilma Type” storms were cold compared to other storms at their respective latitudes.

Water vapor eye temperatures were found to increase as the storm intensity increased. Although no special rate of intensification per degree of warming was found, water vapor eye temperatures themselves were found to correlate with storm intensities. R Squared values for various storm sets ranged from .60 to .80 for the water vapor eye temperatures vs. the storm intensities. IR eye temperatures for the same sets ranged from .49 to .64. There were problems with using the water vapor eye temperatures. If the eye was too small in diameter, the satellite was unable to view the lower parts of the eye because of the angle of view. Different satellites have narrow and wide band water vapor instruments. The wide band satellites yield warmer eye temperatures. And, there appeared to be major problems after a storm reached maximum intensity.

Finally, tables converting estimated storm intensities from water vapor eye temperatures were derived from trend lines of wide and narrow band storm sets. These were applied to respective band width storms for the full periods of the storms before and after maximum intensity. Average error for the period prior to reaching maximum