The Storm Prediction Center's Severe Thunderstorm Data Base, that extends back through 1950, is used to generate time series of the monthly occurrence of tornadoes and severe thunderstorms. These series are then smoothed using the "shape preserving" filter proposed by Tukey (1977). The plots of the number of monthly occurrences are analyzed to see if there are any consistent relationships between the number of events in any given month or sequence of months to the number of events in subsequent months. Both the actual number of reported events and the number of days that the events occurred are considered. Also the data are categorized by type and intensity. This analysis sheds light on the credibility of the various permutations of the stem sentence "there were a lot of/few tornadoes in (select month or season), so this will be an active/sub-normal tornado year" that appear in the popular media each spring.

Reference: Tukey, J. W., 1977. Exploratory Data Analysis, Reading, MA: Addison-Wesley Publishing Co.,506pp.