Florida Guard’s 159th Weather Flight Wins Weather Challenge

  • Published
  • By Tech. Sgt. Chelsea Smith,
  • 125th Fighter Wing Public Affairs

Airmen assigned to the 159th Weather Flight bested 66 teams to win Thor’s Forecast Challenge, a demanding two-week competition that tested weather flights’ ability to develop hour-by-hour forecasts across the globe.

The 2021 reigning champions, a geographically separated unit of the 125th Fighter Wing, defeated teams from across the three components of the Air Force to continue their winning streak.

The competition required teams to forecast 24 hours of weather in locations from Taiwan to Nigeria to Poland. Every day, weather forecasters had to predict wind directions and speeds, ceiling heights, dew points and temperatures using computer models, observation and satellite data. At the end of each 24 hours, the forecasts were scored for accuracy.

“The competition originated in 2017 as a way to build camaraderie and challenge the weather community to test what they can bring to the fight,” said Senior Master Sgt. Jason Ramos, deputy 1W career field manager, USAF Headquarters Weather Force Management. “We are a broad career field with units all over the globe supporting a wide range of Army, Air Force and Space Force operations. This is a fun way to get everyone on the same page and test one of our core functions as weather professionals, which is to predict the environment.”

The team included Maj. Matthew Tanner, 159th WF commander, Tech Sgts. Donald Knopps and Jonathan Miller, and Airman 1st Class Elizabeth Durban — all part-time Florida Air National Guard Airmen. They competed against their active-duty and reserve counterparts remotely – collaborating by e-mail and text to produce forecasts from Florida and Georgia.

“What’s remarkable is that we had to juggle full-time civilian priorities, like work and school, in between working on a forecast,” said Tanner. “We really had to be efficient with our time to keep up with the demands of the challenge and other outside priorities we had going on.”

Durban, assigned to the unit for three years, is a full-time student who works night shifts outside the Guard. 

“I’d come home and stay up until 1 or 2 in the morning trying to pour into the forecast and then hand it off to people working the night shift,” she said. “Time management was really difficult because I was trying to juggle so many things but also give my all to the challenge.”

Durban said she was energized by the versatility of the challenge and looked forward to disentangling data for a new weather event each day.

“I’m a huge nerd when it comes to weather,” she said. “I absolutely loved all the different locations in the world and I’m a really competitive person, so learning who was on top every day was really fun and gratifying for me.”

The team ranked among the top teams throughout the challenge. They eliminated several teams in the first week and outperformed the best 16 teams in the second week’s bracket challenge. The final assignment was to forecast weather for remote and tiny St. George Island in the Bering Sea off Alaska.

“The location is so austere that there’s not a lot of data available,” said Tanner. “There was snow, gale-force winds and a lot of changing weather patterns happening over the 96-hour period. I felt pretty certain that we would not come away with a win again due to the level of difficulty.”

But in the end, a victory for the team was indeed in the forecast.

“I felt a surge of joy and massive elation when they announced we won,” he said. “For us to be able to balance our civilian careers and compete as a secondary duty is a testament to the way we train and interpret data.”

Tanner credits their win to an emphasis on human interpretation, expertise, and discernment over raw data output by computer models. 

“We are going against the grain because we are not holding steady to this idea that we should take what the model gives us and run with it,” he said. “I truly believe that the next big fight is going to happen in degraded operations, and more than likely, computer models won’t be available. That means we need people to analyze raw data, observe with our own eyes and use prior experience to develop a great product. That’s how we’re going to win the next fight.”