Development of An Automated Approach For Identifying Convective Storm Type Using Reflectivity-derived and Near-storm Environment Data

Abstract

Extending storm classification into a separation of convective storm types has applications to nowcasting, severe weather warning and quantitative precipitation estimation and forecasting. This work presents an exploration of an automated classification scheme that uses a combination of radar reflectivity products and near-storm environmental parameters derived from Rapid Update Cycle (RUC) model analyses to assign one of eight classes to each storm observed. An assessment is made of the classification accuracy, and it is found that the scheme that uses the combination of radar and model data outperforms the scheme that uses only radar products. This is particularly notable in the identification of storms that have rotation, and therefore improves the assessment of those storms' potential longevity and severe weather threat.