Publication Date


Document Type


First Advisor

Gensini, Victor A.

Degree Name

M.S. (Master of Science)

Legacy Department

Department of Earth, Atmosphere and Environment


Over the last 50 years, the United States has experienced an increase in severe storm events that produced $1 billion in damages or greater. Much of this loss is attributed to significant tornadoes and hail associated with deep, moist convection. Improving forecasts for these significant events assist in mitigating the impacts of these events. Previous work has identified statistically significant environmental parameters associated with severe thunderstorms, but more research is needed in identifying statistically significant ingredients associated with environments that produce significant tornadoes and hail.

This thesis aims to answer the following question: “Can diagnostics commonly used to forecast for severe convective storms be used as discriminators between severe and significant-severe tornadoes and hail?” Observational and reanalysis-derived proximity soundings were examined for the 1996–2018 period. Severe and significant-severe local storm reports gathered by National Weather Service offices and the Storm Prediction Center were used as a proxy for sounding location and associated hazard magnitude.

For each sounding, a total of 45 different atmospheric diagnostics were analyzed. Results from statistical testing and Support Vector Classification machine-learning showed that 10–500m bulk wind difference, 10-m–1-km average mixing ratio, and 10-m–1-km SRH were able to discriminate between significant and non-significant tornadoes with a maximum of approximately 36% more skill over a random forecast. Overall skill for tornadic discrimination was primarily associated with kinematic diagnostics, while discrimination for hail showed higher amounts of variability and generally lower skill scores across all diagnostics tested with the SVC algorithm.


83 pages




Northern Illinois University

Rights Statement

In Copyright

Rights Statement 2

NIU theses are protected by copyright. They may be viewed from Huskie Commons for any purpose, but reproduction or distribution in any format is prohibited without the written permission of the authors.

Media Type