In this presentation we introduce preliminary results from a climatology of Great Plains thunderstorms developed using the ThOR algorithm (Thunderstorm Observation by Radar) which synthesizes Level II data from the NEXRAD network of surveillance Doppler radars, cloud-to-ground lightning recorded by the National Lightning Detection Network, and storm motion estimates from the NCEP North American Regional Reanalysis database. Key components of ThOR include 1) radar reflectivity quality control using a neural network to remove radar anomalies including ground clutter from anomalous propagation, 2) reflectivity filtering using a fuzzy logic technique to remove stratiform precipitation, 3) reflectivity cluster identification through image segmentation, 4) cluster tracking to develop candidate thunderstorm tracks, and 5) cloud-to-ground lightning attribution to classify candidate tracks as thunderstorms.
While the period of record for the data sets used to construct this climatology (2005-2007) is too short to reveal the manifestation of climate change in the number, distribution, or internal characteristics of thunderstorms, valuable insight into the distribution, structure, and behavior of thunderstorms can be identified. To this end, several diagnostic studies have been performed to interrogate the climatology. Results from an examination of the environments associated with thunderstorm initiation and a survey of the spatiotemporal patterns of the ratio of supercells to non-supercells will be highlighted in this presentation. Directions for future work will also be addressed.