Analysis of Five Types of Dropwindsonde Dynamic and Thermodynamic Profiles in Tropical Cyclones

Dynamic and thermodynamic profiles obtained by GPS dropwindsondes in tropical cyclones since 1997 are investigated using a k-means cluster analysis technique. Five types of profiles of wind speed are found, and the dynamic and thermodynamic clusters are investigated. The cluster analysis seems to be able to discriminate between clusters based on their location relative to the radius of maximum wind speed. The different wind speed profiles are also related to different temperature, humidity and stability profiles. The data also suggest weak (< 1 m/s) upward motion on average throughout the tropical cyclone below about 1500 m altitude.

The clusters are related to their location within the tropical cyclones relative to the radius of maximum wind speed. They can also be used to investigate whether high-resolution numerical models are accurately simulating these types of profiles within the tropical cyclone core.