How Catastrophe Experts Model Hurricane-Induced Storm Surge

By Aarti Dinesh | July 1, 2013

The 2004 and 2005 North Atlantic hurricane seasons were a turning point in the way the property/casualty insurance industry viewed hurricane risk. The needs of the insurance industry required catastrophe models to explicitly model storm surge risk. Until this point, storm surge risk was not explicitly modeled.

A slew of hurricane landfalls in 2004-2005 led to massive levels of destruction to property and life. A trigger point was Hurricane Katrina (2005), which caused unprecedented levels of losses (an estimated $47 billion in total losses), the largest storm surge and flooding damage to date in the United States. Katrina caused a fundamental shift in the way the insurance industry looked at risks along the U.S. eastern coastline.

The lack of focus on storm surge was not due to cognitive dissonance within the risk management community. Hurricanes have always arrived with storm surge – Hurricane Andrew (1992) hit South Florida with tides reaching as high as 16.9 feet above normal; Hurricane Hugo (1989) had a peak surge of 20 feet near Seewee Bay, S.C.; and the Galveston Hurricane of 1900 (a category 4 hurricane at landfall) brought a tremendous surge that caused thousands of fatalities.

Recent demographic shifts in population and wealth toward more-exposed coastal locations greatly increase the risk of losses from storm surge related to hurricanes. Hurricane Katrina (2005) and Superstorm Sandy (2012) introduced an associated phenomena – flooding that persisted for days as opposed to hours.

Recent demographic shifts in population and wealth toward moreexposed coastal locations produce an enormous increase in the risk of losses from storm surge related to hurricanes.

A Known-Known Risk

Storm surge, an omnipresent factor of hurricanes that has received attention in recent decades as losses attributed to it have risen, is caused primarily by high winds pushing on the ocean’s surface, the result in water piling up higher than the ordinary sea level. Additional factors include the low-pressure core of the translating storm, and the offshore bathymetry (the underwater equivalent of topography) and tidal surge. The resulting storm surge (water depths, velocity and on-shore run-up) are influenced by the speed, intensity, radius of maximum winds (RMW), radius of the wind fields, angle of the track relative to the coastline, the physical characteristics of the coastline and the bathymetry of the water offshore.

Storm surge causes damage to properties on shore from many aspects of the surge – static water depth, the kinetic energy in the moving water (velocity), and from the kinetic energy of waves impacting partially submerged structures.

Rainfall associated with the hurricane can exacerbate the flooding associated with surge. Hurricane Irene (2011) demonstrated that the water flowing down inundated watersheds can trigger flooding far upstream of the coast as rivers are prevented from flowing to the ocean.

Inundation from storm surge can also slow down recovery efforts. Storm surge damage can also increase building cost inflation post-storm (demand surge) due to pent-up demand for labor and materials.

The risk of storm surge damage along the U.S. coastline varies.

The immediate coastal regions are within the high velocity zones, which during a hurricane are where wave action or high velocity water can cause significant structural damage. At more than 60 pounds per cubic foot, a 3-foot wave traveling at 5 mph can deliver a lateral force far exceeding building standards. Further inland, rising waters could cause significant damage to building interiors.

The interactions of storm characteristics upon the inconsistently populated U.S. coastline produce a varying risk of catastrophic surge potential.

For example, Miami has high coastal exposures where storm surge losses would be expected to be high, but the risk is in fact lower due to deeper bathymetry characteristics, while Tampa has very shallow sea floors where surge losses can be much greater.

Central Louisiana has shallow sea floors resulting in significant storm surge risk, but its low coastal exposures lessen potential damage. The west by northwest land falling azimuth of Super Storm Sandy greatly contributed to the severe and persistent flooding along the New Jersey shoreline.

Most flood losses are uninsured, but the biggest insurance losses from flood have been from the National Flood Insurance Program. Another take away from these recent events is that private insurers are accepting a significant amount of flood exposure either through the acceptance of mortgage-originated forced insurance placement measures, as excess layers on large residential risks and as sub-limited layers on commercial multi-peril and difference in conditions policies.

The continued treatment of flooding as a separately named sub-peril of a hurricane makes loss settlements difficult, especially in the velocity zone where there may not be much evidence available to make a forensic determination of the cause of loss.

The lasting images of a flooded New York subway system and the damage the New Jersey coastline are searing reminders of this sub-peril’s powers.

But Sandy was unique, arising from the confluence of an Atlantic tropical storm that merged with an onshore cold front to transition to an extra-tropical storm that was steered that all happened during an uncharacteristically extreme high tide.

Extensive measurements of flooding have provided opportunities for the validation of storm surge model components.

The hazard data was available shortly after Sandy made landfall, and was incorporated within the model to create the wind footprint and storm surge footprint for the event.

The modeled peak gust wind speeds and storm surge heights for Sandy were compared and validated against observed data to ensure that the model is emulating the characteristics of the actual event.

Claims data are now becoming available for use.

What’s next?

Lessons from past storms impact insurers’ view of risk. Over the years, the risk footprint for storm surge risk has changed, attributable to climate change. Additionally, population shifts are putting more of society’s assets at risk. The U.S. government is reconsidering the role of NFIP and private insurers may be expected to accept more insured flood risk. Consequently, risk evaluation strategies must constantly evolve to provide the best value in a competitive marketplace.

About Aarti Dinesh

Dinesh is product manager at EQECAT Inc. She joined EQECAT in 2007 and is responsible for the management of EQECAT's global tropical storm models.

From This Issue

Insurance Journal West July 1, 2013
July 1, 2013
Insurance Journal West Magazine

The Disaster Issue: Flood, Earthquake, Windstorm, Terrorism, Cyber Attack & More; Commercial Auto (Including Taxis, Limos & Fleets); Digital Product Guide; Bonus: The Florida Issue (Special Supplement)