The authors provide a statistical and physical basis for understanding regi
onal variations in major hurricane activity along the U.S. coastline on lon
g timescales. Current statistical models of hurricane activity are focused
on the frequency of events over the entire North Atlantic basin. The except
ion is the lead author's previous work, which models the occurrence of hurr
icanes over the Caribbean Sea, Gulf of Mexico, and the southeast U.S. coast
separately. Here the authors use statistics to analyze data from historica
l and paleoclimatic records to expand this work. In particular, an inverse
correlation in major hurricane activity across latitudes at various timesca
les is articulated. When activity is above normal at high latitudes it tend
s to be below normal at low latitudes and vice versa. Past research, paleoc
limatic records, and historical data hint at the potential of using the Nor
th Atlantic oscillation (NAO) as an indicator of where storms will likely t
rack over long timescales. An excited (relaxed) NAO is associated with high
er (lower) latitude recurving (nonrecurving) storms. The Gulf (East) Coast
is more susceptible to a major hurricane strike during a relaxed(excited) N
AG.