An investigation of the relationship between ENSO and U.S. tornado climatology

The goal of this thesis is to examine the tornado frequency during different seasons in different zones of the eastern two-thirds of the continental United States, during different phases of ENSO. Recently, ENSO has also been examined in correlation with tornadic activity by Bove (1997). Bove (1997) found that during the La Nina phase of ENSO, there was a rise in tornadic activity in the Ohio and Tennessee River Valleys in the spring and early summer. He also found that in "Tornado Alley," which is the area of the central U. S. that annually receives the highest number of tornadoes, the tornadic activity decreased during the LN phase of ENSO. This approach to studying tornadic activity is more climate-oriented than previous research.

The data in this thesis will be studied by season and by geographic area. This is because different parts of the country have different peaks in tornadic activity. For example, Anthony (1988) found that the peak of tornadic activity for the majority of the southeastern states is in April and May. In other parts of the country, such as the northern high plains, the peak tornadic activity is in June (Tecson et al., 1982). Changnon (1982) found that the peak tornadic activity in the state of Illinois is spread out over June, July and August. Also, it is important to know if the largest deviation from normal takes place during the most active part of the year for a particular area, especially if the departure from normal is an increase in the number of tornadoes.

Using this information, possible relationships between ENSO and tornadic frequency of the eastern half of the CONUS will be exposed. The manner in which this subject is addressed is statistical. That is to say, conclusions about physical mechanisms will not be drawn from the results.

After examining the results, it is hoped that performing statistical analyses on the data will substantiate any changes in the tornadic activity that were reflected in the data by revealing that the changes were not chance occurrences. This would be helpful since the ultimate goal is to be able to predict when and where tornadoes will occur.

In this paper, a correlation between tornado climatology and the varying phases of ENSO will be investigated. In Chapter II, the data will be presented along with how it was processed. Chapter III will discuss the results of the data processing. Chapter IV will examine the statistics involved in this study. Conclusions and possible suggestions will be presented in Chapter V.