The 2024 Atlantic Hurricane season was expected to be one for the record books. Early forecasts had a record number of storms developing in the Atlantic basin because of favorable environmental conditions.


What You Need To Know

  • The GFS model did well with several tropical systems this season from formation to dissipation

  • Hurricane Hunters happen to be at the right place to survey Hurricane Oscar

  • Models don't do well with precursor small-scale tropical systems

But no one could have predicted how the actual season would pan out, with five storms making landfall in the United States, three of them in Florida, with these storms causing fatalities and massive destruction.

The forecast in May predicted “the highest number of named storms NOAA had ever issued” in its pre-season outlook." Thanks to a waning El Niño transitioning to a La Niña during peak season, environmental conditions were expected to become conducive for development, and that it did.

First half of the season

The season officially began on June 19 with Tropical Storm Alberto. It was followed by Beryl, Chris, Debby and Ernesto.

Beryl and Debby became the most notable storms during the first half of the season, as both made landfall in the U.S. Beryl made the history books, reaching Category 5 status earlier in the season than any other storm.

Late summer lull

After Ernesto dissipated on Aug. 20, there seemed to be a quiet period. With hurricane activity in the Atlantic subsiding near the peak of hurricane season, the validity of seasonal forecasts was in doubt. 

It’s important to note that it’s not that there wasn’t any activity. Hurricane Francine developed in the Caribbean Sea and moved into the Gulf of Mexico before landfall in Louisiana on Sept. 11.

“As is often the case with seasonal hurricane activity, there isn’t usually just one cause for it being active or quiet," says Meteorologist Craig Setzer, a Certified Consulting Meteorologist and Hurricane Preparedness Specialist.

He adds, "This year, it appeared to be a combination of very dry air and an increase in Saharan dust outbreaks along with warmer than normal air temperatures in the middle-and-upper parts of the atmosphere that created the mid-seasonal lull." Less unstable conditions existed and tropical cyclones need instability for development.

"The long range forecast tended to be right for all of the wrong reasons," says Spectrum Bay News 9's Chief Meteorologist Mike Clay. "A very active Cape Verde season was expected, which didn't happen. Instead, the Central American Gyre was very active late in the season giving us several significant hurricanes."

Mid-September activity increases

As soon as the climatological peak of hurricane season was reached (Sept. 10), tropical activity seemed to awaken. Tropical Storm Gordon formed on Sept. 13 in the central-eastern Atlantic, but remained over the open Atlantic and never surpassed tropical strom strength.

Mid-Sept. also brought the formation of Helene, one of the costliest and deadliest storms this season. Helene would make landfall near Perry, Florida on Sept. 26.

Models and forecast track cones

From tropical storm to landfall was only a matter of days for Hurricane Helene, and knowing the track of the storm was vital to protect lives and property. Various data models assist meteorologists in forecasting, and the National Hurricane Center (NHC) uses these models to create a forecast track cone.

“The National Hurricane Center relies heavily on global and regional (hurricane) models, which they call ‘guidance’ or ‘aides’ to help create a forecast track,” says Setzer.

National Hurricane Center

One specific model did relatively well with several storms this season, from development to landfall. “The Global Forecast System or GFS is a global model which has shown improved skill in predicting tropical cyclone development well ahead of other models.” 

However, just because it does well with one storm, doesn't mean it will accurately predict all storms. Setzer says, “It has a high false alarm bias, so it also tends to overpredict development that ends up not happening.”

In this age of social media, where folks might be heightened aware of tropical development, this bias could create a panic with the model showing a land-falling hurricane ten days out only for the storm to not develop or dissipate well before land.

Regional forecasting models

Another type of model the NHC uses to refine its forecast tracks is the regional hurricane model. Referred to as regional because, as Setzer says, “They don’t cover the whole globe, just the region of a hurricane.” This model became vital during Hurricane Milton.

Adding, “The Hurricane Analysis Forecast System (HAFS) version B did a good job not only predicting Milton’s intensity but also predicting when it would strengthen and weaken based on internal storm dynamics.”

However, that was after Milton had already formed. Before development, the models struggle. This was the case with both Hurricanes Milton and Oscar.

Models aren't exact

The NHC releases outlooks every six hours during the season for the tropical basins, including the Atlantic. Those include both a two-day graphical tropical weather outlook and a seven-day graphical tropical weather outlook. This year, both Hurricanes Milton and Oscar precursor storms were given low odds of development, only to then form within 12 to 24 hours.

An area of disturbed weather that would eventually become Milton was given low odds to develop two days before Milton formed. (NHC)

“Milton was fairly well predicted after the area of convection started to consolidate in the western Gulf of Mexico. Prior to that, there were several areas the models seemed to focus on for development, but it didn’t occur,” explains Setzer.

Adding, “Oscar was such a small system, it likely was underrepresented in the data going into the model, and then the model has difficulty in the handling of a very small atmospheric feature.”

Oscar ended up being the smallest hurricane on record, with a hurricane wind field of five or six miles. The Hurricane Hunters, scientists and pilots who fly into tropical systems to investigate conditions happened to be at the right place at the right time. Due to its Oscar’s proximity to land, they flew into the storm.  

Would the same designation have been given if the storm was in the far eastern Atlantic?

“It’s a good question.”

He explains. “While our satellites are very good, they currently cannot resolve small-scale storm intensity. It is likely Oscar would have been only designated a tropical storm had it remained in the open ocean away from aircraft reconnaissance. “

Overall, the models did well with all 18 named storms this season. “I’m hopeful this trend in model improvement will bring us better predictions, resulting in more specific warnings and storm preparation areas in the near future," says Setzer.

Our team of meteorologists dives deep into the science of weather and breaks down timely weather data and information. To view more weather and climate stories, check out our weather blogs section.