Eavesdropping on a twister

Each spring hundreds of tornadoes rip across the American heartland — so many that a belt from Texas to southern Canada is often referred to as Tornado Alley.

Scientists know surprisingly little about these small but deadly storms, whose wind speeds can exceed 320 kilometres per hour. Much of what they do know comes from a mix of Doppler radar and storm chasers who, in vehicles fitted out like mobile weather stations, pursue them overland, trying to get close-up measurements while remaining safely out of the storm’s path.

But this isn’t the only way tornadoes can be monitored. They also produce infrasound, says Brian Elbing, an experimental fluid mechanics researcher at Oklahoma State University, Stillwater, Oklahoma. {%recommended 5052%}

Not to be confused with ultrasound (sound pitched above the limits of human hearing), infrasound is a low rumble at frequencies below the limits of human hearing. It came to prominence when scientists realized that it travels so far in the atmosphere that could be used to monitor nuclear bomb tests from half a globe away. More recently, scientists have realized that it is also produced by a wide array of natural events, ranging from earthquakes to meteors breaking up in the upper atmosphere. 

One of these is tornadoes. Nobody is quite sure why they produce ultrasound, Elbing says, though his favorite theory is that it comes from oscillations in the radius of the tornado vortex. The frequency of these oscillations, he says, appears to be correlate to the size of the funnel cloud, which in turn is related to its destructive power.

In a paper presented at a meeting of the Acoustical Society of America, in Minneapolis, Minnesota, Elbing eavesdropped on an 11 May 2017 tornado and estimated, based on its ultrasound frequency, that it had a diameter of 46 meters. Later, he says, the official report from the US National Oceanic and Atmospheric Administration also put the diameter at 46 meters—a figure so close to his own group’s calculation that he jokes his group should have inserted some kind of error into their figures just to ensure people would believe them.

Despite his finding’s unexpected precision, Elbing notes that it’s still a pilot study. “We really need a lot more observations,” he says. Right now, he says, the scientific literature shows only a handful of detailed infrasound observations of tornadoes. “I’d say the number is less than ten.” 

That said, he thinks that infrasound holds tremendous promise for tornado detection and monitoring. To begin with, tornadoes can show up on infrasound earlier than they do on Doppler radar. “We started detecting this one ten minutes before,” he says, adding that other studies have suggested that they might show up as much as an hour earlier. 

Infrasound also works at longer ranges than radar because infrasound wraps around the curvature of the Earth, while radar travels in a straight line. That’s a big enough effect that tornadoes don’t have to be all that far way to hide from Doppler radar. Also the infrasound may be better at detecting storm systems with multiple tornadoes. In the case of his 11 May measurements, Elbing says, conventional measurements showed the presence of only a single tornado, while his data revealed two.

That said, Elbing may face an uphill battle in convincing other tornado researchers that he really has invented a better mousetrap. One of the sceptics is Howard Bluestein, a storm chaser and research meteorologist from the University of Oklahoma, Norman. {%recommended 5583%}

“We currently rely on Doppler radar,” he says. “[Elbing] has presented evidence that there may be a sonic signal. If this is true, he would have to show why it is better than Doppler radar.”

Elbing counters that conventional tornado warnings aren’t actually all that accurate. Seventy-five percent, he says, turn out to be false alarms. Reducing that percentage is an obvious step toward getting people to take them more seriously.

Furthermore, he says, the greatest benefits of infrasound detection and tracking might not be in the prairie regions of Tornado Alley. “Oklahoma is a big, flat state,” he says. “Storm systems move in a line, so we predict them quite well.”

Of greater concern, he says, is a region he calls Dixie Alley, which encompasses much of the American South. Tornadoes here tend to be smaller than those on the Great Plains, he says, but they are also deadlier. One problem is mountainous terrain, which blocks Doppler radar signals. Another is a maze of twisty roads in which it’s easy to be trapped by an approaching storm. “Storm chasers will not go there,” Elbing says.