Examination of atmospheric and oceanic circulations may explain interannual climate variability in the Northern Hemisphere on a seasonal scale. It is crucial to develop more accurate seasonal climate forecasts using both global circulation and sea surface temperature (SST) indices to aid in long-range weather forecasts. These global circulation and SST indices are becoming increasingly available to worldwide users and using them for seasonal prediction has spread not only to scientists, but also to brokerage firms, utilities, and the Department of Defense (DoD). DoD is extremely interested in long-range seasonal forecasts of severe weather for asset protection, mission planning, and worldwide operations. The goal of this research was to create a predictive algorithm for locations in the southeastern and south-central portion of the United States in support of the Air Force Combat Climatology Center (AFCCC) to use in predicting the intensity of the spring and summer severe weather seasons.

The most significant predictor of the intensity of the severe weather season in the southeast and south-central regions of the U.S. was identified as the proximity of the indices to the respective region. Beginning with multiple linear regression, this study found there were relationships between several severe weather parameters, such as thunderstorm and heavy precipitation events, and these known global circulation and SST indices. However, R² values showed that SST indices had more significance with severe weather since they appeared more often in the multiple linear regression models. In addition, analysis of variance provided valuable incite into the development of classification and regression tree (CART) analysis. After little predictive value was found using traditional statistics, CART analyses were developed to create an algorithm for DoD forecasters to use for seasonal severe weather prediction. Results confirmed that algorithms with reasonable predictability can be produced for forecasting the intensity of the severe weather season.