reThought Flood

Forecasting Hurricanes Part 1

It’s that time again: forecasters are raising the alarm for the coming hurricane season. Why do we care at reThought Flood? Because the floods caused by storm surges and heavy rains that accompany hurricanes can cost ten times more than the cost of the wind damage. 

The forecasts begin every year in Spring, on the cusp of the half-year hurricanes season. It runs from June 1 to November 30, when historically about 97% of named storms occur. (Which is no comfort for sure to the coastal property owners who suffered storms earlier this year, or those in southern Texas areas that flooded earlier this month,  reminiscent of Ike.).

With the internet, everyone now becomes an expert at forecasting. Data on weather siblings El Niño and La Niña, Sea Surface Temperatures, the North Atlantic Oscillation, Ocean Heat Content, Coral Reef Bleaching, and many others are available ad nauseum for the amateur forecaster to make their own predictions. Yet most people still revert to a feeling: “we’re due” (think Miami, Wilmington), or “it just happened” (the west coast of Florida). In some areas popular opinion swings to “it won’t happen” (Savannah, New York), and others acknowledge hurricane landfalls as “what else is new?” (Houston, Mobile, Charleston).  

When it comes to the experts, two highly regarded teams that give actual predictive forecasts have come out with these stormy prognostications:

Colorado State University: 23 named storms in 2024, with a 62% chance of a major (CAT 3-5) hurricane making Continental US landfall.

University of Pennsylvania: 33 named storms. They don’t get more specific, but have an alternate model view which predicts 19 storms.

Note, that’s a range of between 19 and 33 named storms. If nothing else, this tells us that the amateurs perhaps have as good a chance of getting it right as the pros.

CSU is the granddaddy of all hurricane forecasters, but for  all professional and academic prognosticators it’s all about number-crunching the massive data now available. That means running sophisticated climatology models, and considering model  outputs against experience. Both CSU and UP recognize that the high uncertainties around any such forecast, and most update them monthly as the season progresses (expect a CSU update hitting the news on June 1.) 

For the armchair, Weather-Channel-watching forecaster, this year presents two very real anomalies:

  1. Sea water temperatures recorded by many stations, from Tampa to Miami, in July 2023 were the highest ever recorded (that’s since 1911). A buoy in Manatee Bay recorded 101oF. No swimming in that water!
  2. The Washington Post recently studied tidal gauges along the Gulf and East Coasts. More than a dozen, from Texas to North Carolina, show sea levels at six inches or more higher than they were in 2010. It’s less than a ½ inch per year, but over twenty years it’s almost a foot, which is huge when you think of how tightly we define flood zones – and the rise has been accelerating.  

But back to hurricanes. The experience obviously varies from year to year, but the actual thirty-year average is fourteen named storms. And note well: over twenty years the average is about sixteen, and over the past ten years it’s seventeen. That puts this year’s rather wide range of professional forecasts – between 19 and 33 named storms – into a different light.

The forecasts and the increasing averages both support the conclusion that we’re suffering more storms each year. True, high-frequency years are often followed by lesser numbers, which may indicate regression to the mean. But the mean is increasing.

Ed Haas

Ed Haas joined reThought Insurance in October 2019 as property risk and CAT modeling consultant. Ed was formerly with Marsh Risk Consulting for over twenty years as Senior VP, with a focus on natural hazard modeling for flood, wind, earthquake, and other perils. Work included managing data for modeiing, interpreting results, and applying them to insurance and risk management programs. The work included site assessments at a wide variety of unique properties throughout the US and internationally. Ed was the Risk Consulting leader in the Marsh Real Estate Practice and served as lead risk advisor to the largest clients. Responsibilities included insurance program design in regard to selection of limits and deductibles, as well as designing and managing appropriate risk control programs. Specification for new construction with regard to CAT perils was a key contribution to these clients, Ed’s career started with FM Global, and has held P&C broker’s and professional engineer licenses.

Leave a Reply

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