Using a K-Means clustering and watershed segmentation algorithm to automatically classify convective storm types

An automated approach to classifying convective storms is based upon both environmental and radar data using a K-Means clustering and watershed segmentation algorithm. Composite reflectivity was tracked and clustered according to three scales: 20 km, 200 km, and 2000 km. For each scale, certain storm types were chosen: short-lived convective cells, supercells, ordinary cells, and convective lines. Using data from five events between May 2008 and May 2009, storms were hand-classified into types. Initial analysis shows that these storms are distinct and can be automatically classified using a decision tree. Ultimate goals of this study are to predict hazards for each storm type using this automated decision tree.