NEWS
Lessons from Hurricane Ian
The 2022 Atlantic Hurricane season could unveil the shortcomings of the seasonal forecasting the insurance industry leans on: BMS.
After a slow start and then a big blow from Hurricane Ian, the 2022 Atlantic Hurricane season could yet go down as “in line” with seasonal forecasts for above-normal activity levels, but will more likely unveil the shortcomings of the seasonal forecasting the insurance industry leans on when placing North American nat cat bets.
That is the view of Andrew Siffert, SVP and senior meteorologist at BMS Group. He is a hurricane tracker and forecaster who has seen beyond what he considers the industry’s rather narrow seasonal views thanks to his prior role in the construction of the live-cat programme at Flagstone Re.
“Right now, the insurance industry looks at the number of named storms expected in the basin and that increases the overall probability of a US named storm landfall,” Siffert said. A calm 2022 interrupted by the whiplash of Hurricane Ian shows “we need to move away from this notion that above-normal or below-normal is what we ought to look at”.
The insurance industry had been primed for action in 2022, with all of the key meteorological institutes warning of just such “above-normal” activity levels in the Atlantic basin.
Going into September, the bulk of those seasonal forecasts, just like the hurricane season itself, had looked like duds. The Atlantic had whipped up only three named storms, none of which had managed a run of longer than two days. July 4 through September 1 was a complete shut-out. Hurricanes Fiona and Ian added rather sudden and violent balance to the picture.
“Overall, I think you’re going to look back and say the seasonal forecast panned out,” Siffert said.
In fact, most seasonal forecasts from leading institutions such as the National Hurricane Center (NHC), Colorado State University’s Tropical Weather & Climate Research department or the Tropical Storm Risk Consortium of the University College of London have winning track records, Siffert can demonstrate.
“The science needs to push beyond seasonal forecasting.”
Andrew Siffert, BMS Group
The question becomes, is 2022’s accuracy—a three-month massive undershoot balanced by a one-month massive overshoot—the kind of accuracy the insurance industry most needs?
Such seasonal forecasts for total named storm count or related metric “should really mean nothing to the insurance industry” compared to insights into when and where storms will track under the age-old industry logic: “it only takes one”.
“The science needs to push beyond seasonal forecasting,” Siffert said. A step-up in understanding seasonal steering currents that determine where tropical convection should track would beat the dice-rolling that follows even the most accurate prediction of named storm counts.
As it is, steering currents emerge only from the shadows when danger is imminent. Not long after tropical depression #9 became Tropical Storm Ian, NHC maps were pointing to Cuba and Florida. Could that have been said long before tropical depression #9?
New priorities
Climate science, for all the additional attention and resources it has brought to the game, seems somewhat bent on taking matters the other way: enabling better forecasts for the number of named storms in 2050 rather than the most likely track and timing of named storms in 2022.
2022 could be the poster child for such a realignment of the scientific priorities. The slow start and the late-season rebound both appear to have been predictable.
“Tropical convection needs a lot to come together to be perfect.”
The full-year seasonal forecasts for above-normal activity levels were fully justified by the La Niña conditions that reduce Atlantic basin windshear and the high Atlantic surface temperatures that support convection, Siffert noted. “Everything pointed to having a higher named storm count.
“But tropical convection needs a lot to come together to be perfect,” Siffert added. He suspects two misses in the 2022 forecast ingredient pool: drier than usual air coming from Africa and the kick-on effect of having had multiple La Niña years in succession.
To explain why the uptick in late-season activity proved so drastic, researchers may end up looking deeper at what happens when La Niña seasons start stacking up on one another. “The back half of multi-consecutive La Niña seasons—and we are in our third La Niña season in a row—tend to be heavy with activity,” Siffert said.
Such insights leave Siffert imagining a point down the road when the science might tell the industry, for example, that we’re all clear through August, that Texas is clear through September or Louisiana-Mississippi should be safe after August, or the Florida west panhandle is x percent safer than the east panhandle for the season.
To a meteorologist who honed his skills in the live-cat industry, increased accuracy in sub-seasonal forecasting could open the door for solutions to the “major capital allocation problem” in reinsurance, chiefly holding expensive coverage for a given peril through the catastrophe off-season.
“There is a tremendous amount of capital that has to sit in place for a full year where it is not needed.” Siffert said. “The capital could move elsewhere in the off-season.”
Main image: Shutterstock / Bilanol
