Statistical Relations between Summer Thunderstorm Patterns and Continental Midtropospheric Heights

Abstract

An orthogonal principal component analysis of National Climatic Center summer thunderstorm frequency data over Virginia revealed five statistically significant modes of variation. In some, power was spread statewide, while in others it was concentrated over a few small regions. A similar analysis was performed on the NCAR record of monthly 500 mb heights over North America, and four significant height patterns were determined. Multiple regression was used to relate the individual thunderstorm components to the upper-air patterns. The fit of each time series of the first four thunderstorm components by the 500 mb patterns was statistically significant, with approximately one-third of the variance of the thunderstorm pattern explained. A subsequent canonical correlation revealed one significant multivariate relationship. The results provide statistically significant evidence for the linkage between macroscale flow patterns and mesoscale convective storms.