Predicting the formation of tornadoes using association rule mining by studying a real life tornado event : Georgia, USA January, 2013

Abstract

Tornadoes form in violent thunderstorms due to instability and wind shear present in the lower atmosphere. The spinning of a tornado is the result of the updrafts and downdrafts caused due to unstable air. The mystery that how and why tornadoes are formed are far away from a satisfactory explanation. In this paper, data is extracted from real time tornado event occurred at Georgia, USA in January, 2013. Then in-depth analysis has been done on each variable responsible to bring tornado and finally association rule mining has been applied to find association among all those weather variables. Our study produced interesting rules to predict non tornadic and tornadic weather conditions.