In the recent literature, considerable attention has been paid to the relationship between climate signals and tropical cyclone activity. Much of the research has focused on Atlantic basin activity while less attention has been given to landfall frequency and the geographic distribution of that potential risk to life and property. Moreover, little attention has been given to the main sources of uncertainty in predicting activity through the near-term horizon (multiple upcoming seasons). These are important questions for decision makers in strategic planning, risk management, and risk mitigation.

A statistical analysis of historical data back to 1900 can be used to examine where the relationship between sea surface temperatures (SSTs) and hurricane landfalls is statistically significant. This type of analysis can be sensitive to the choice of SST anomalies, the analysis period, and the definition of landfall. SST forecasts beyond one season are highly uncertain and sensitive to one's choice of statistical time series model. However, by examining where along the U.S. coastline the SST signal on hurricane landfall frequency is strongest, one can further study the physical characteristics of tropical cyclones that relate to landfall risk in those regions.

These characteristics include the origin of tropical cyclones (genesis), their track (steering), and their ability to develop into and maintain hurricane strength before making landfall (intensification). Results indicate that storms making landfall along the east coast of the U.S. have different genesis and development characteristics than those making landfall along the Gulf coast. As SSTs warm, genesis density shifts, influencing regional landfall risk at a statistically significant level. Further, the probability of hurricane landfall is sensitive not only to warm SST conditions, but also to the effect of ocean temperature anomalies on the atmosphere's general circulation.