An objective algorithm for the identification of convective tropical cloud clusters in geostationary infrared imagery

Tropical cyclones trace their origins to disorganized groups of thunderstorms over tropical waters called "cloud clusters." The transformation from cloud clusters into mature tropical cyclones is not well understood due to the remote location of these clusters (away from the main observation network) and the complexities of the smaller-scale physics that occur during genesis. Several studies have used a catalog of developing and non-developing cloud clusters to search for distinguishing features between the two. The creation of these datasets is cumbersome since they must be identified visually through the application of arbitrary and subjective rules.

An automated algorithm is developed to identify tropical cloud clusters in geostationary infrared satellite imagery. The algorithm proposed here significantly reduces the time and effort needed to identify and catalog cloud clusters by applying objective search criteria in an automated way. This algorithm will enable the creation of a comprehensive, historical cloud cluster database that will have both research and operational applications. The database will be made available to the scientific community and will be updated in real-time.