In the United States, nothing captivates nor terrifies quite like a tornado.
Mythologised by Hollywood films and increasingly, across social media, the funnel-shaped ‘twisters’ have helped spawn multiple generations of meteorologists and storm chasers alike.
But this natural phenomenon remains a mystery that scientists are yet to fully explain.
While it’s known tornadoes develop from a specific type of thunderstorm, called a supercell storm, the exact triggers of a tornado are still not fully understood.
Research scientist Sean Waugh instructs the TORUS team about how to use windsonds.
“What we're hoping to do is to understand the supercell (storm) better, understand that cascade of processes that lead to tornado genesis and then take that understanding and put it in the forecast process,” Dr Adam Houston, one of the project’s lead investigators, tells SBS Dateline.
“Right now, we can have a sequence of supercells lined up in a particular area, and we do not know which one of those is going to become tornadic.”
“What we want to know is, 30 minutes in advance or an hour in advance, which of these storms is going to become tornadic because that's the lead time that people need to take shelter.”
More than a thousand tornadoes occur in the United States each year, causing more than a billion dollars in damage. So, the US government has invested heavily in the team’s research.
The project is made up of more than 50 weather researchers working at the cutting edge of tornado climatology, including teams from the National Oceanic and Atmospheric Administration’s (NOAA) National Severe Storms Laboratory, meteorologists from several universities, and their students.
In tow, is a state of the art fleet of more than 20 vehicles – including mobile radars, a P3 plane called a Hurricane Hunter and eight mobile weather stations known as mesonets.
Together, the team must spend one month storm chasing across the Central Great Plains of the United States. It’s an area stretching almost a million square kilometres, and they can travel anywhere from North Dakota to Texas, and Wyoming to Iowa, sometimes crossing multiple states in single day.
A member of the TORUS team on assignment.
They need to all work together in a tightly choreographed dance under deadly conditions, to capture data from multiple points of the storm - things like wind speeds, air pressure and temperatures, as well as getting radar observations.
“So [the teams] kind of move around and they work off of each other, so they can use their information together to form kind of a cohesive picture of what the storm environment's doing and how it's changing as we continue to operate on the storm,” says Sean Waugh, one of the team’s research scientists, who has been studying tornadoes for 12 years.
“So, the radar track, for example, that's going to stay a little bit further away. It gives us a broader scale picture of what's going on. But the [mesonets], their job is to get a little closer and, you know, more personal with the storm.”
The TORUS team in action monitoring a tornadic supercell storm.
“We definitely get up close and personal with things. A lot of times that means driving in and out of the hail core, a lot of heavy rain. So, it's a lot of staying on top of roads, traffic, you know, weather conditions, that sort of thing. It's a very complex moving picture and it's my job to keep the vehicle safe and as well as the occupants safe.”
Making the TORUS team’s work even more challenging, is the fact tornado behaviour in the Unites States is changing.
March to June each year is typically tornado season. They most frequently occur in the Central Plains states, giving it the nickname Tornado Alley, and most usually, between the hours of 4-9pm.
What we want to know is, 30 minutes in advance or an hour in advance, which of these storms is going to become tornadic because that's the lead time that people need to take shelter.Dr Adam Houston
But in recent years, the intensity and locations of tornadoes have become increasingly unusual and harder to predict.
“The biggest thing we've seen is an increase in the variability of occurrence,” says Dr Harold Brooks, Senior Scientist, at the National Severe Storms Laboratory, in Norman, Oklahoma.
“The number of days with 30 or more (tornadoes) has gone from being one day every other year to about two and a half to three days per year.”
According to Dr Brooks, there’s also been a rise of about 10 per cent in the number of tornadoes occurring east of the Central Plains, in more densely populated areas, like the state of Kentucky.
Tornadoes most frequently occur in the Central Plains states, giving it the nickname Tornado Alley.
"The population density, especially in rural areas of the Southeast, is a lot higher than it is in the Great Plains.
"As a result, you put a tornado through the Southeast, it’s almost guaranteed to hit houses and hit people, whereas in the Plains you can have a tornado on the ground for a long time and it just doesn't hit any people at all.”
In December last year, the town of Mayfield, Kentucky was decimated by an EF4 tornado almost a mile wide.
It stayed on the ground for more than 300 kilometres. Crossing through five states, it destroyed more than 15,000 buildings and left 88 people dead, more than the average annual death toll in just one day.
The strength, path and duration of the tornado that hit Mayfield were all unusual. Even more unusually, it happened at night, in the middle of winter.
The city of Mayfield in Kentucky was destroyed by a tornado in December 2021.
Currently, the average warning time for a tornado is just 9-15 minutes.
By revealing the hidden composition of severe storm and potentially the triggers for tornadoes, the TORUS team hope to improve tornado forecasts and ultimately, save lives.
Dr Houston says it will take a multi-faceted approach.
“You have to attack it from a better understanding. You have to attack it from better observations. You have to attack it from better numerical weather prediction and all of these things fit into that that larger process of forecasting to try to improve that lead time.”
