Developing a Climate Conditioned View of Atlantic Hurricane Risk

For over 30 years, an entire industry has been built on quantifying the risks of extreme weather events to insured property, infrastructure and supply chains. Throughout that time, the natural hazards caused by landfalling Atlantic hurricanes have been a primary research focus, given they dominate the overall financial impacts amongst global climate perils. Since 1995, the ocean climate in which Atlantic hurricanes develop has become more favorable in terms of both genesis and development, and this is partly responsible for escalating losses. Insurance and reinsurance companies rely on stochastic simulation models to quantify the full range of impacts from potential future hurricanes, and in turn require such models to reflect changes in risk owing to a combination of natural climate oscillations and long-term climate trends.

This study provides a historical perspective on ‘climate conditioning’ of catastrophe risk models (“CAT Models”), the methodology used to develop a robust view of risk and recent innovations that apply a combination of historical data, modeling, physics-based sensitivity testing and AI/ML applications to ensure an up-to-date view of the risk in light of the warming global climate. The analysis deconstructs the hurricane life-cycle from genesis to development to landfall and examines the sensitivity to a warming ocean climate. From a genesis perspective, the findings reveal a number of zones within the Main Development Region (MDR) that have preferentially leveraged warmer sea-surface temperatures (SSTs) to produce anomalously high counts of named storms. The analysis goes on to point out that each of these zones has a unique landfall climatology in terms of how often, where and at what intensity hurricanes from these zones go on to strike the U.S. East and Gulf coasts.

After genesis, the development cycle leading to hurricane and in some cases major hurricane intensity sets up the maximum winds that could be experienced at landfall. The study examines development from two standpoints – intensification and tracking – and finds that in both cases, a warming ocean environment imparts a statistically significant signal. In terms of severity, Rapid Intensification Cycles (RICs) are becoming more frequent, but are not distributed uniformly across the MDR. This has implications on the regional risk profile of storms reaching the coastline. In terms of tracking, a large-scale warming of the Atlantic can alter the strength and position of sub-tropical high pressure over the Atlantic, known as the Bermuda-Azores High, and in doing so has a material influence on the propensity of hurricanes to make a U.S. landfall.

Other aspects of the global ocean climate such as the El Nino Southern Oscillation (ENSO) and the Loop Current are examined as well. While local Atlantic SSTs have the strongest and most direct influence on Atlantic tropical cyclone risk, it is well known that these other factors can play a role in the risk at the scale of an individual storm over days to weeks to a full season of storms over months to years. The study addresses how these and other climate factors can be weighted appropriately as part of a holistic view of hurricane risk.

While the analysis conclusively finds that Atlantic Ocean warming has increased the risk of U.S. hurricane landfalls, it is also clear that the relationship of genesis-to-landfall is complex and non-uniform along the U.S. coastline. Sensitivity testing provides further insight on the changing risk while motivating further investigation.