A statistical study of extreme nor'easter snowstorms

• Published in: Involve, a Journal of Mathematics
• Main Authors: Karvetski, Christopher, Lund, Robert, Parisi, Francis
• Format: Text
• Language: English
• Published: MSP 2009
• Subjects: extreme values; North Atlantic oscillation; peaks over threshold; Poisson processes; snowstorms; 60G55; 62G07; 62G32; 62M99; North Atlantic
• Online Access: https://projecteuclid.org/euclid.involve/1513799167; https://doi.org/10.2140/involve.2009.2.341

This short paper studies the statistical characteristics of extreme snowstorms striking the eastern seaboard of the United States — the so-called nor’easters. Poisson regression techniques and extreme value methods are used to estimate return periods of storms of various snow volumes. Return periods of several memorable events are estimated, including the superstorm of 1993, the North American blizzard of 1996, and the blizzard of 1888. While nor’easters are found to occur more frequently in late winter than early winter, no evidence of increasing/decreasing storm frequencies in time or dependencies on the North Atlantic oscillation is found.