For as long Bill Ramstrom can remember he’s been fascinated with hurricanes.
“There was this hurricane when I was about 8 and it was just so interesting to see the power of nature and how adults and parents and the school and everybody else just had to get out of the way of it,” he said.
The experience left a big impression, opening the door to reading anything he could on meteorology.
“…And then as I got into my teen years, and there were computers available, I started fiddling around with programming, says Ramstrom.
As the senior software engineer at NOAA’s hurricane research division, Bill has had a hand on several projects, but none as important as what he works on now. It’s called Hurricane Analysis and Forecast Systems, more commonly referred to as HAFS, a computer model that went fully operational last year with one goal in mind – to accurately predict hurricanes.
“It performed very well last year,” he said. “We were very happy with how the numbers looked with it.”
One of the fundamental building blocks of the model was the improved grid, or nest, it runs on. That nest provides more accurate and higher-resolution data into the model.
“Just like with your phone, if you get more pixels in your camera, you get a nicer looking photo,” Ramstrom explained. “We have the same sort of idea of simulating the hurricane.”
Until recently, most hurricane models have lacked the computer power and resolution necessary to make these forecasts possible. But with the implementation of NOAA’s newest weather and climate supercomputers in 2022, they are now able to create some of the highest resolution forecasts to date.
“It gives you some more confidence that the model is really capturing what’s going on in nature.”
And so far, it’s worked. Since it’s experimental debut in 2019, HAFS has shown an improvement in track predictions for storms by 10 to 15%. In 2022, Hurricane Ian rapidly intensified off Florida’s southwest coast to a powerful Category 5 storm. The only model to correctly predict that period of rapid intensification was HAFS – while it was still in its testing phase.
“I was actually also involved in flying the hurricane at the time, so it was a whirlwind,” Ramstrom said.
Dr. Andrew Hazelton is an associate scientist who works at NOAA’s Hurricane Research Division, collecting and evaluating data that flows into and out of the HAFS model.
“Ian was one that we were able to simulate really well because we had aircraft in there throughout, had a lot of data going into it,” he said.
But producing traditional model forecasts isn’t all HAFS can do. Hazelton and his fellow team also employed a new suite of graphics, visualizations of a developing storm that haven’t been reproduced by any models before. In one example, Hazelton showed how wind shear and dry air attempted to move in on Hurricane Ian before landfall but failed.
Hazelton explains, “So it’s coming in, you know, basically trying to get into the inner core. But then, we see because it doesn’t get all the way in there, you know the storm is able to get, become a Category four … five storm.”
It’s these simulations that really set HAFS apart from the rest. Helpful for forecasters from the hurricane center down to your weather experts to really dissect a storm and bring a more accurate forecast to those who need it most.
“HAFS has come a long way and continues to show promise. And with the success the team has seen so far, they hope it continues for another hurricane season. But there are still some problems the team is looking to solve, including the problem of multiple nests. But seeing the success this team has brought so far to the model, it’s safe to say the future of hurricane forecasting is in good hands.
Zach Covey is a meteorologist with Spectrum News 13, the Orlando Sentinel’s television partner.