A “drought” of major hurricane landfalls? Science weighs in.

Hurricane season is here. Tropical Storm Colin is the third named storm in the Atlantic basin this year, and has brought rough seas, heavy rain and flooding to Florida and parts of the east coast of the U.S. So far all three storms have been fairly weak without much impact. It’s been a fairly calm start to the season in terms of losses. It’s also worth noting that two of the three named storms (Tropical Storms Bonnie and Colin) have affected the U.S. coastline. Is this a sign of things to come for the 2016 Atlantic Hurricane season?

There is an air of anticipation this year. Recent seasons have been over-shadowed by the presence, or expectation, of El Niño conditions, which normally suppress hurricane activity in the North Atlantic. This year however; a La Niña looms.

A hurricane doesn’t have to be a category 3 or more to cause fatalities and huge financial losses – it just has to make landfall in a highly exposed area.

Forecasters around the world have been issuing their best estimates of hurricane activity over the last few weeks. Will we see the first major hurricane to make landfall in the U.S. in over a decade? And if so, where is most likely to be hit? Have we learnt the lessons from big storms of the past, and are we more resilient and risk aware? Sobering questions, which are raised at the start of every hurricane season (June 1st to November 30th), but with the ten-year anniversary of Katrina marked last season, these questions are refreshed in our minds.

Without a major hurricane (category 3 or above on the Saffir-Simpson hurricane intensity scale) making landfall in the U.S. since Wilma hit the Florida coastline in the busy 2005 hurricane season (which also included Katrina and Rita), should we be on high alert or complacently rely on the apparent major hurricane landfall “drought” which has lasted over a decade?

Major hurricane landfall “drought”?

Last year Dr. Robert Hart at Florida State University and colleagues pointed out that the term “drought” is somewhat arbitrary anyway. Using hurricane landfall statistics to infer reliable climatological trends (drought or otherwise) is a process fraught with difficulties.

Taking into account the uncertainty in the wind estimates that decide whether a hurricane is major or not, it’s easy to suggest that “droughts” have occurred in recent decades; the current respite from U.S. landfalling storms is not so unusual. And if the atmospheric pressure at the centre of the storm is considered as a measure of storm intensity instead of wind speed, then hurricanes Irene (2011) and Sandy (2012) should also be classed as major hurricanes, arguably debunking the so-called “drought.”

Ultimately, storm impacts are what really count. A hurricane doesn’t have to be a category 3 or more, to cause fatalities and huge financial losses – it just has to make landfall in a highly exposed area.

Cutting edge of climate modelling

If the atmospheric pressure at the centre of the storm is considered as a measure of storm intensity instead of wind speed, then hurricanes Irene (2011) and Sandy (2012) should also be classed as major hurricanes.

Long-range seasonal forecasting (beyond normal weather forecasts of around a week or so) is becoming more advanced and more useful. A range of hurricane activity forecasts are produced ahead of and during the hurricane season by the major weather and climate modelling centres around the world, which show a degree of skill but come with large uncertainty—especially when considering whether or not any storms will make landfall. Physical and statistical models are run, and together these give a broad view of what to expect.

At the cutting edge of physical modelling, new offerings such as the HiFLOR at the Geophysical Fluid Dynamics Laboratory (GFDL) are starting to show skill in pointing out where there may be more chance of hurricanes hitting the coastline in a given season. Dr. Gabriel Vecchi of the GFDL, gave a Willis Research Network (WRN) webinar just last month to discuss the latest advancements.

Impact research

In terms of impact research, WRN partners at the National Center for Atmospheric Research (NCAR) and Wharton, have also been developing a Cyclone Damage Potential (CDP) index, which takes into account more than simply wind speed or central pressure when assessing the potential for a storm to induce loss. Factors such as a storm’s size and forward translation speed (how quickly it approaches the coast) have a large bearing on storm surge and duration of impact in any particular area. The CDP can be calculated over a historical record or for a single storm, and has recently been adapted to be applicable to seasonal forecasts.

Is the industry ready?

Read WRN's 2016 Atlantic Hurricane Season Outlook

Read WRN’s 2016 Atlantic Hurricane Season Outlook

The insurance industry has a strong capital base to absorb losses and catastrophe modelling, along with scenario-based stress tests that are now more sophisticated than ever before. But is this enough?

Specialist insurers offering homeowner insurance in Florida have been largely untested by hurricane losses, and have developed in a market conducive to showing strong returns and market share growth, according to Reuters.

But forewarned is forearmed, and as some forecasts indicate the onset of La Niña conditions during the middle of hurricane season, seasonal hurricane forecasts should be considered in preparation for this year’s storms. It is not a question of “if” a storm hits the coastline, rather simply “when” is it going to happen.

Hurricane season outlooks

WRN partners at NCAR have provided the first in our yearly series of hurricane season outlooks. There is significant degree of uncertainty in hurricane forecasts this early in the season, but the indications from most of the models that either ENSO (El Niño Southern Oscillation)-neutral or La Niña conditions are expected to develop, at least giving us confidence that El Niño conditions are not expected. Most of the forecasts are currently predicting a close to average season, though some are indicating higher than average hurricane activity, which further indicates uncertainty.

As we progress through the season and we learn more on the potential for La Niña, the certainty should increase as the spread of the model outputs reduces. The next WRN forecast round up will be released in late-June. In the meantime, the hurricane-prone communities and financially-exposed insurance industry alike, will be keeping an eye on how the season develops.

About Geoffrey Saville

Geoffrey Saville is a member of Willis Towers Watson's Analytics Technology Team, having joined the company in 2013…
Categories: Natural Catastrophe, Reinsurance | Tags: ,

Leave a Reply

Your email address will not be published. Required fields are marked *