Automated Real-time Extraction of Storm Properties from Gridded Fields

A multiscale segmentation algorithm has been developed to identify storms at different scales through contiguityenhanced K-Means clustering of texture features. The identified storms are used as a template for tracking storms through time and for identifying storm properties. A variety of scalar features are computed from the gridded values falling within the geographic and temporal extent of the identified clusters. The resulting algorithm is illustrated in real-time operation and the utility of the extracted properties is demonstrated through a decision tree that is capable of identifying the storm type based on just the automatically extracted properties.