The deaths of three experienced storm chasers—Tim Samaras, his son and fellow researcher Paul, and Carl Young—in the EF5 tornado that struck El Reno, Okla. this past Friday has spurred critics to again attack the folly of those who put themselves in the way of twisters by choice. It should. Being around any tornado is insanely dangerous, and the one in El Reno on May 31 was a monster among monsters: its funnel was 2.6 miles across, the widest ever documented in the U.S., and radar clocked its wind speeds a few hundred feet above the ground at 296 miles per hour.
The Samarases and Young weren’t in El Reno for the sake of entertainment, though they had all been featured on the Discovery Channel program Storm Chasers. Tim Samara was a respected tornado investigator and the founder of TWISTEX (Tactical Weather Instrumented Sampling in Tornadoes Experiment), the field team to which Paul Samaras and Carl Young also belonged. (Examples of Tim’s published research can be found here, here, and here.) Other researchers seem to have regarded him as brave but cautious, and not a thrillseeker.
Unfortunately, no amount of experience might have made a difference in a direct confrontation with a storm this powerful and erratic. Other storm chasers on the scene who got too close saw their cars battered catastrophically. A car holding Mark Bettes of The Weather Channel and other passengers was thrown 200 yards (he and the other occupants were injured but not killed). The SRV Dominator 2, an 8,000-lb. storm-chasing vehicle armored against tornado damage by storm chaser Reed Timmer, had its hood torn off. Richard Charles Anderson, an amateur storm chaser, was also killed only minutes after snapping a photo of the El Reno twister.
Whether any of those people belonged in a vehicle near this tornado may be dubious. Then again, almost none of the thousands of people who jumped into cars and jammed interstates around Oklahoma City in an attempt to flee the approaching storm belonged on the road. Shamefully, they may have been led astray by at least one local meteorologist and a shelter builder interviewed on CNN, who suggested leaving the threatened area. The logic of escaping sounds compelling, but it ignores the more compelling consideration that the risks of being caught on the road, even in a seemingly solid vehicle, are overwhelming.
That fact was driven home to me while working on The Weather Channel series Hacking The Planet with Cara Santa Maria and Brian Malow, and on our later special The Truth About Twisters. It became clear that surviving an encounter with a tornado is not just a matter of doing the right things. It’s also about not making some lethal errors. Here are a few of them; we might as well start with the one that was the downfall of the Samarases and others at El Reno.
Don’t be in a car during a tornado. Yes, in theory, getting away from a tornado’s likely path would seem to be a smart move, and a car would typically be the fastest, handiest way to do it. But putting that theory into practice is harder than it seems. A Tornadoes can change course unpredictably, and it isn’t always easy to see their funnels in the murk of bad weather surrounding them, which puts storm chasers relying on visibility alone, without support from tracking radar, at a profound disadvantage.
Moreover, according to the National Oceanic and Atmospheric Administration, tornado warnings give on average only 13 minutes notice. That’s enough time to drive miles away—but not with throngs of cars all trying to abandon the area at once, as they did in around Oklahoma City last week. So trying to drive away during that small window of opportunity could easily lead to your being trapped in your car as the tornado bears down on you.
No car’s safety features offer much guarantee in that case. The greatest harm may come from flying debris, which can be of any size or composition—paperclips, loose lumber, mailboxes, shards of glass, pieces of other cars, anything and everything—moving on average about 100 mph. In The Truth About Twisters, Cara observed a demonstration of what an airborne two-by-four moving at that speed could do to a car: it punched a hole through the front windshield and out the back, and would have pulped any passenger in its way. Far smaller objects might not have as much penetrating power, but they could still pummel, slice, or kill.
Moreover, even if your car isn’t tossed by a tornado or shredded by debris, and even if it isn’t stuck in traffic, the danger isn’t over. Severe rain and hail can make driving hazardous, with cars losing control and smashing into one another. Torrential rain also leads to local flooding, so it’s not uncommon for cars driving away from tornadoes to end up underwater. According to the National Weather Service, most of the people killed by floods during tornadoes drown inside their cars.
But suppose you are unlucky enough to be in a car and find a tornado bearing down on you, what’s your most prudent choice?
Don’t hide under an overpass. In April 1991, a group of people successfully hid under an overpass for shelter against a tornado in Kansas. Videotape of their adventure helped to popularize the idea that overpasses offered a good refuge from twisters to highway drivers. The problem is, the Kansas incident seems to have been a fluke. As meteorologist Greg Forbes of The Weather Channel told Cara in The Truth About Twisters, several people have been killed trying the same trick. There is even one tragic case of a woman who died after leaving her perfectly solid house for what she thought was superior protection under an overpass.
Overpasses seem as though they would offer great protection from the wind—and sometimes they can, if the tornadic winds are coming primarily from certain directions. From other directions, however, the narrow confines of an overpass will focus and intensify the winds, and funnel flying debris toward anyone hiding inside. Winds during tornadoes are highly changeable, so even if conditions under an overpass seem benign at first, they can rapidly (and lethally) change for the worse.
So to repeat my previous question, if you’re in a car and need protection from a tornado, what should you do? The best option would be to leave the car and try to get inside any nearby house, store, office building or other sturdy structure that you can get into—ideally, one with a good basement or a proper tornado shelter. But short of that, counter intuitively, if you can’t find indoor refuge, your best bet may be to lie flat in a ditch, holding onto something heavy as an anchor, with cushions or blankets over you as a shield against debris.
Don’t open the windows and doors of your house. If you do have the option of sheltering indoors, you want to seek a spot that is underground, or in a windowless space toward the center of the structure so that you are shielded from debris that might blast through the outer walls. Be attentive to what might fall on top of you in the event of a building collapse; if you can get under a strong piece of furniture for additional protection, do it. But don’t try to protect the house as a whole by listening to advice I grew up with: to open all the windows and doors in the interest of making it easier for the pressures inside and outside to equalize. The received wisdom in those days was that because the pressures inside a tornado’s funnel were very low and the pressures inside the wall of spinning wind were very high, closed buildings would implode.
That is a myth. According to the experts we interviewed for Hacking The Planet, when a tornado hits a house, it subjects the structure to complex, fast-changing forces that push and pull in rapid succession. In effect, the tornado squeezes and tugs at your house as though it were playing the accordion, exposing any weakness. Opening the windows and doors only succeeds in letting the winds into the house so that internal supports can be shaken apart, too, which weakens the structure even more.
Don’t believe me? Watch this clip of a demonstration from The Truth About Twisters in which Brian observes how leaving the door of a house open makes a catastrophic difference in its stability.