Date of Award
Doctor of Philosophy
There is a poor understanding of severe thunderstorm variability over the Southeast U.S. due to the lack of a reliable, homogeneous, long-term observational records. A potential avenue to address severe thunderstorm variability involves computing environmental indices derived from reanalysis datasets and use those as proxies for severe thunderstorm days. Unfortunately, reanalysis datasets have known issues with inhomogeneity that are not related to climatology. Thus, this research first sought to reconstruct a climatology of severe thunderstorm days back to 1925 from the Twentieth Century Reanalysis (20CR) dataset, while developing methods that potentially minimize non-climatic inhomogeneities. Severe thunderstorm day proxies were reconstructed from 20CR by computing two separate time series using two thresholds of the significant severe parameter. Analysis of each time series included change point analysis, two-proportion z-tests, autocorrelated conditional Poisson models, and verification of a random sample of days before and after 1950 using archived sources. Results suggested an overcount in severe thunderstorm days prior to 1950, and a bias correction factor was determined and applied to the time series that seems to improve the estimate of severe thunderstorm days. The Southeast also has a high climatological risk for tornado outbreaks, and several notable tornado outbreaks impacted the region in the 1930s. A particularly deadly outbreak occurred 21-22 March 1932 across the region, but no formal study has examined this historic event. Alabama suffered the greatest loss with approximately 268 fatalities and 1,874 injuries. This research reconstructed a series of surface maps using archived meteorological observations that depict the progression and synoptic patterns of this outbreak. A comparison of these surface maps to 20CR derived surface maps suggest that 20CR generally reproduced the broad synoptic patterns of this outbreak. Further, archived newspapers and tornado databases assisted in describing the societal impacts of three violent F4 tornadoes that impacted north and central Alabama. The findings of this doctoral research contribute to the understanding of severe thunderstorm variability and tornado outbreak vulnerability in the Southeast U.S. during the 1930s and 1940s. These results could assist weather forecasters and emergency managers better address vulnerability to severe thunderstorms and tornado outbreaks in the region.
Dissertation or thesis originally submitted to ProQuest
Baker, Bradley Wayne, "Southeast U.S. Severe Thunderstorm Climatology and 1932 Tornado Outbreak" (2019). Electronic Theses and Dissertations. 2437.