Infrasound signals from nontornadic and pre-tornadic storms indistinguishable

As very intense, long-lasting storms, supercells are different from the average afternoon thunderstorm. They rotate, whether they host a tornado or not, due to the wind in the atmosphere changing speed and direction significantly with height.

Almost all significant tornadoes form in supercells, but most supercells remain nontornadic. Distinguishing between these two categories is difficult because they are so similar, and current radar techniques have about a 75 percent false alarm rate for tornadoes. This motivates other tornado detection methods, like infrasound signals.

Coffer and Parker simulated supercells to see if infrasound signals could detect pre-tornadic storms. All geophysical phenomena produce infrasound signals, and they are often used for avalanche, tsunami, and nuclear testing detection.

“You can imagine yourself on a mountain. You could hear snow falling down the mountain in an avalanche,” said author Brice Coffer. “That’s the audible component of what you’re hearing, but it’s actually producing pressure waves in a spectrum, just like the sun produces visible and other light.”

The team used observations of nontornadic and pre-tornadic storms to initialize their models, then changed aspects of the supercell to see how the infrasound signals responded. They found the detection method may not be suitable for predicting tornadoes.

“While we were able to document differences that eventually lead to one storm producing a tornado or not, there weren’t any unique pressure signals associated with those processes,” said Coffer.

As this is the first study to explore infrasound signals in simulated supercells, the researchers are interested in comparing their simulations to other groups gathering infrasound data from storms directly.

Source: “Infrasound signals in simulated nontornadic and pre-tornadic supercells,” by Brice E. Coffer and Matthew D. Parker, Journal of the Acoustical Society of America (2022). The article can be accessed at https://doi.org/10.1121/10.0009400 .