'Screwball,' 'crazy,' 'strange': As hurricane season ends, researchers note its surprises

A man looks for salvageable items in the aftermath of Hurricane Melissa in Black River, Jamaica, Thursday, Oct. 30, 2025. (Matias Delacroix / AP)

Three Category 5 storms, one of the most powerful hurricanes ever recorded, zero U.S. landfalls and a mystifying lull at the usual peak of activity: Together, these and other factors made for a "screwball" hurricane season this year.

That's how atmospheric scientist Phil Klotzbach put it, anyway.

"It was just a strange year," said Klotzbach, who studies hurricanes at Colorado State University. "Kind of a hard year to characterize."

Hurricane season comes to its official close on Nov. 30. In some ways, 2025 fits what researchers expect to see more often as the climate warms: Hurricanes continued forming late into the season and severalintensified at extreme rates to produce some of the most intense storms in history.

But in other ways, it was simply odd. Fewer hurricanes formed than experts predicted, but almost all of them became major storms. And the continental U.S. was spared a landfall for the first time in a decade. The surprises were a reminder of hurricane season's unpredictability — particularly in a warming world — even asforecasting gets more accurate.

Fewer hurricanes, higher intensity

Forecasters at the National Oceanic and Atmospheric Administration in Maypredicted an above-average season with six to 10 hurricanes. Of those, at least three were expected to be major storms, meaning Category 3 or above, with sustained winds at or above 111 mph.

Klotzbach came up with the same forecast independently, and other hurricane-tracking groupswere in the same ballpark.

In the end, fewer hurricanes formed, but of the five that did — Erin, Gabrielle, Humberto, Imelda and Melissa — four were considered major.

Hurricane Imelda over Bermuda on Oct. 1. (NOAA)

"That's the highest ratio there's been in the past 50 years," said Brian McNoldy, a hurricane researcher at the University of Miami's Rosenstiel School of Marine, Atmospheric, and Earth Science.

What's more, three of those major storms were Category 5, the highest level of intensity.

Forecasters' predictions of an above-average season still proved accurate despite the lower number of storms because of a metric called accumulated cyclone energy — essentially a calculation of the overall intensity and duration of all tropical storms in a season.

Klotzbach predicted the accumulated energy would be 125% of the 30-year average. The season ended up at 108%, which, given the low number of hurricanes, means each packed a punch.

"It was a quality season, not a quantity season," he said.

Nine of the past 10 Atlantic hurricane seasons have been above normal, according to Klotzbach, who attributes the trend to high ocean temperatures and La Niña, a seasonal circulation pattern that tends to weaken the high-altitude winds that discourage hurricane formation.

McNoldy, who closely tracks Atlantic water temperatures, said 2025 was "anomalously warm."

"Whatever storms were out there definitely had a lot of fuel to tap into," McNoldy said. Ocean heat drives evaporation, causing warm, moist air to rise from the surface to create convection; hurricanes require ocean temperatures of at least 79 degrees Fahrenheit to form.

A man walks in the rain before the arrival of Hurricane Melissa in Santiago de Cuba, Tuesday, Oct. 28, 2025.  (Ramon Espinosa / AP)

Elevated sea temperatures allowed several hurricanes to build intensity at a blistering pace. Erin's maximum sustained wind speedincreased by around 75 mph in just 24 hours. Melissa went from atropical storm to a Category 4 hurricanein the same amount of time.

Both are "really, really exceptional" rates of intensification, McNoldy said.

Melissa also tied the 1935 Labor Day hurricane as the strongest storm to make landfall on record. Both had sustained winds of 185 mph, and Melissasaw a 252 mphgust.

A strange lull at peak time

Despite the ocean's hurricane fuel, this season at times lacked a spark. Storm activity typically peaks in late August and early September, but not this year, McNoldy said: "For about three weeks during the peak of the season, it was completely dead."

Nothing of significance formed in the Atlantic between Aug. 24 and Sept. 16, according to Klotzbach. That hadn't happened since 1992.

Wrightsville Beach in North Carolina as Hurricane Erin approaches (Peter Zay / Anadolu via Getty Images)

Although the lull wassimilar to a slowdown observed last year, researchers do not think it will become a trend.

"That we had two quiet peak seasons in a row is certainly strange, but I think it's just a coincidence," McNoldy said.

The reason for last year's lull was that tropical storms developed too far north to make their way across the Atlantic and pick up steam, according to Klotzbach. During this year's pause, the eastern Atlantic was relatively stable and dry, conditions not conducive to the formation of powerful storms.

One big near miss

No hurricanes made landfall in the U.S. this year for the first time in a decade, according to Klotzbach.

But that would not have been true if Hurricane Imelda had not made a sharp turn.

As Imelda tracked toward the U.S. as a tropical storm in September, forecasters thought the Carolinas could receive buckets of rain. But Hurricane Humberto, a Category 5 monster churning several hundred miles away in the open ocean, steered Imelda away from the coast.

Homes Collapse Hurricane (Heather Jennette / AP)

Thephenomenon is known as the Fujiwhara effect: When two storms rotate around a shared midpoint, the weaker storm is often subsumed by the stronger.

"Had Humberto not been there, Imelda was probably a big flooding story," Klotzbach said.

As McNoldy put it: "Another crazy aspect of the season."

AI forecasting shows promise

In written forecasts from the National Hurricane Center during Hurricane Melissa, a new term popped up again and again: "Google DeepMind."

Federal forecasters are increasingly leaning on the novel artificial intelligence hurricane forecasting tool, which was instrumental in an earlyprediction of Hurricane Melissa's rapid intensification.

Google DeepMind built an impressive resume over the course of hurricane season, McNoldy said. Recently, he assessed the levels offorecasting error in more than 10 hurricane modelsand found that DeepMind was among the best.

"It was a model that was just put out in the public eye in June, and it outperformed the other conventional models for track and intensity, which is unheard of, to have something right off the shelf new be that good," McNoldy said. "It's hard to not be optimistic about it."

 

COSMO MAG © 2015 | Distributed By My Blogger Themes | Designed By Templateism.com