Skip to content

If you feel it, chase it: using seismic signals to detect tornadoes

Tags: seismic

In the 2024 sequel to the 1996 film “Twister”, attention is once again being called to the severity and mystery associated with tornadoes. “Twisters” follows tornado chasers in Oklahoma, and as more towns get destroyed and lives are lost from twisters day after day, many chasers try to find ways to model and “tame” to prevent casualty. The threat of destruction and injury we see in the movie is real, so what can be done currently to warn against tornadoes, and how do we detect and track them using geophysics?

In most cases, twister and tornado are interchangeable terms, with the former being more casual and the latter a more technical term. Warm air rises while cool air falls, and that happens inside of thunderclouds during a very powerful thunderstorm called supercells. As the warm air is flowing upwards, the condensation from it creates moisture for the tornado to feed off of later. Because of all the winds and updraft of warm air combined with the downdraft of cooler air, a vortex can be created because of the temperature difference inside and outside of the cyclone. This instability and pressure is just what a tornado needs to develop, and the instant that the tube of spinning air hits the ground, it is considered a tornado. The tornado ends once the temperature difference outside and inside the tornado gets smaller, the conditions get more stable, or the moisture in the air dries up. The Enhanced Fujita Scale (or EF Scale) is a common way to rate storms and communicate the wind speeds and level of destruction from tornadoes, going from EF0 to EF5.

EF scale
The Enhanced Fujita Scale (EF Scale). (Credit: weather.gov)

Both the original film and the sequel take place in Oklahoma, in an area part of what is considered Tornado Alley. This is a loosely defined area in the central United States where tornadoes tend to be most frequent. About 1,000 tornadoes are reported in the US each year, and while Tornado Alley covers 15% of the area, it accounts for 30% of the tornadoes. Annually, there are around 80 deaths and at least 1,500 injuries due to tornadoes, and an incredible amount of damage to environments and infrastructure. In 2022, approximately $708 million worth of damage occurred in the US due to tornado damage.

A map of the US with Tornado Alley
A diagram of Tornado Alley. (Credit: Wikipedia)

Tornadoes cause incredible damage and destruction to people and towns, so researchers actively are looking for ways to mitigate damage and find early warning signals. While the new film is a bit of a stretch for what can realistically be done currently, the motivation for better early-warning systems and loss prevention is definitely a reality.

It’s also difficult to determine exactly when the cyclone hits the ground, making it for sure a tornado. The best method currently is simply using human observation and passing along the word by sounding alarms. Doppler radar can be used to detect rotation in storms, but it can’t distinguish when tornadoes make contact with the ground except at extremely close range. In fact, only about one in five actually touch down on the surface, leading to many false alarms. Other tornado alarms are designed to sense the rapid drop in atmospheric pressure (which is generally a good indicator of a tornado) or the traditional roaring sound that a tornado makes, but they don’t allow for a large window of time for warnings. Tornadoes can also change in intensity over time, so something presumed to be small can become very deadly very quickly, and ways of detecting them earlier can be life-saving.

Many researchers over the years have looked into seismic signals that are produced when a tornado makes contact with the ground—at this point there is a large amount of friction as wind interacts with the ground, and the energy used to overcome this friction in order to make contact and stay in contact is transferred into the ground in the form of thermal and seismic energy. The characteristics of this seismic energy can tell us about the tornado that created it. Many other hazards are already detected through seismic signals, like earthquakes, volcanic eruptions, coastal wave action, and atmospheric disturbances. The US Navy has used seismometers to track typhoons and hurricanes, so perhaps the next expansion of this is tornado tracking. 

Supercells turned into an EF4 tornado that ran through southeastern Missouri and southern Illinois on February 29, 2012, and happened to pass through an area with an array of over 100 deployed seismometers that were part of NSF’s EarthScope program. The research group didn’t intend on picking up seismic data from the tornado, but they did. The data recorded the effects of a pressure change in the atmosphere. This could represent a correlation between the tornado activity and the atmospheric pressure, helping us to better understand the atmosphere right before tornadoes touches down, allowing for more early warning signs.

Another earthquake monitoring station in Kentucky recorded strong ground vibrations on December 10-11, 2021. During this time, a devastating tornado was passing through the area, providing another example of seismic data recording tornado passage.

In 2011, the Joplin tornado, a devastating multi-vortex EF5, struck Joplin, Missouri. It killed 158 people and injured about 1,150 others—one of the deadliest tornadoes in the US, causing $2.8 billion in damage at the time. On its route, it passed within 2 km of a seismometer that was part of the EarthScope Transportable Array, which picked up seismic signals from it. One study modeled the waves to see the amount of force tracked on the ground with the changing strength of the tornado. The model also was able to clearly show the time that the Joplin tornado was in contact with the ground. Despite the high correlation between the seismic data and the strength of the tornado, the amplitude of that signal was still small. In order to get a decent reading, the tornadoes must pass very close to the seismic stations, so a larger network of seismic stations would be helpful—especially in places like Tornado Alley. In their models, they had to account for other sources of seismic noises, isolating the tornado in order to see a correlation.

Bar graph of Economic damage caused by tornadoes in the United States from 1995 to 2022 (in million U.S. dollars)
Economic damage caused by tornadoes in the United States from 1995 to 2022 (in million U.S. dollars). (Credit: Statista)

Seismic tornado detectors are generally conceptually possible, but the biggest issue with the idea is the amount of additional noise within the data. Whether the tornado occurs in an area with civilization or just natural ecosystems, damage and movement of structures generates some seismic noise in the tornado’s path. The process of sorting out all these unwanted signals in order to leave just the touchdown of the cyclone is complex, but is well-developed and has gained attention in seismic analysis.

The process usually doesn’t require any new equipment, just looking at data in a different way. Seismic stations are already abundant throughout the US, but are usually much more dense in earthquake-prone areas. Placing additional ones in the central US may allow us to see more correlations between tornadoes and seismic signals.

In both theory and a few instances of practice, seismic stations have measured the intensity of passing tornadoes. Currently, the National Weather Service uses radar and human sightings to issue tornado warnings, but these can sometimes not allow a lot of time for people to seek safety, and the classification of tornadoes can change very rapidly. “Twisters” shows the intensity of tornadoes, and how they can become catastrophic or disappear in an instant. If more evidence was collected indicating a strong link between seismic signals and incoming tornado touchdown, it could increase the certainty that a tornado has touched down and give more time to people in the area, saving lives and preventing injury.

Cover image by Niccolò Ubalducci