How to Prepare for a Tornado

Tornado warnings average roughly thirteen minutes of lead time. Written down that sounds workable. Lived through, it is not much at all, and the reason is that almost nobody spends those minutes the way they imagine they would.

Play it out honestly. The alert goes off and the first instinct is not to move, it is to check: open a window blind, look at the sky, find the local station, ask someone else in the house whether they heard it too. Perfectly reasonable behaviour, and it eats five of the thirteen. Then comes gathering, the dog, the phone charger, shoes, a child from an upstairs bedroom, and that is most of the rest. The sheltering itself is the shortest part of the whole sequence.

This is why almost all tornado deaths happen in places that were already under a warning. The warning is rarely the missing piece. What is missing is a decision made in advance, so that the alert triggers movement rather than a conversation about whether the alert is serious.

The practical version: decide today where you go, and decide it for every building you spend real time in. Home is the one people think about. Work, school, the gym and a relative’s house are the ones that catch people out, because the answer is different in each and none of them are obvious under pressure.

These two words decide what you do, and people mix them up:

• Tornado Watch: Conditions are favorable for tornadoes. Review your plan, check supplies, and stay alert.
• Tornado Warning: A tornado is happening or imminent; take cover immediately.
• Lead time is short: Warnings average about 13 minutes; a warning is not the time to start planning.

Tornado winds are almost unimaginable. The strongest ever measured reached about 318 mph in the May 3, 1999 tornado near Bridge Creek and Moore, Oklahoma, and even the top of the damage scale, EF5, begins at just over 200 mph.3

Here is the part that surprises people: those EF numbers are not wind measurements. Almost no tornado is ever measured directly, because that would require an instrument to survive inside one. The rating is assigned afterwards by surveyors who walk the damage path and work backwards from what the wind did to buildings and trees.

That approach comes out of Texas Tech University, whose National Wind Institute developed the Enhanced Fujita Scale to fix three specific weaknesses in Ted Fujita’s original 1971 scale: a lack of damage indicators, no account of construction quality and variability, and no definitive correlation between damage and wind speed. The EF scale replaced a single judgement call with a catalogue of things a tornado might hit and the degrees of damage each one shows.4

Follow that logic and you land somewhere counterintuitive. A tornado is rated by what it happened to encounter, so an extremely violent one that crosses nothing but open pasture cannot earn a high rating, because there is nothing standing in it to record the damage. Construction quality matters too: a poorly anchored house destroyed at 120 mph tells you less than a well built one destroyed at the same speed. The rating describes the damage, not the storm.

Above-ground safe rooms exist because of a tornado and a piece of artillery. Wind research at Texas Tech began in 1970, after an F5 tornado struck Lubbock, killed 26 people and caused more than $100 million in damage. The institute that grew out of it runs a pneumatic cannon capable of producing simulated wind speeds over 250 mph, firing simulated wind-borne debris at storm shelters, shelter components and building materials.5

That cannon is why a modern safe room can be trusted above ground at all. The testing was used to develop the standards for safe above-ground shelters, and it is still in demand for qualifying new shelter materials and construction. When a shelter is described as rated, this is broadly what it survived: not steady wind, but a projectile arriving at highway speeds, which is what actually kills people in tornadoes.5

That also explains the ordinary advice. Put mass between you and flying debris, and put yourself below the level it travels at. Every rule below is a version of that one idea.

• Best: A basement, storm cellar, or dedicated safe room.
• Otherwise: A small interior room (bathroom, closet, or hallway) on the lowest floor, away from windows.
• Protect yourself: Cover your head and neck; use a mattress, blankets, or a helmet if you have one.
• Mobile homes and vehicles: Are not safe; get to a substantial building before the storm arrives.

Expand

Tornado folklore is unusually persistent, partly because some of it was official advice decades ago and partly because dramatic footage taught a generation the wrong lesson. The Storm Prediction Center is unusually direct about all four of these.

Sheltering under a highway overpass is the worst of them. The SPC calls it "a very dangerous idea": the space under a bridge accelerates the wind rather than blocking it, funnels debris through at speed, offers nothing to hold on to, and leaves you stopped in a traffic lane during an emergency. The famous video of people surviving under one is the exception that taught everybody the wrong rule.6

Opening the windows to "equalise pressure" is the next. In the SPC’s words it is "absolutely useless, a waste of precious time, and can be very dangerous. Don’t do it." Houses do not explode from pressure difference; they come apart because wind and debris take them apart, and you have just spent some of your thirteen minutes helping the wind get inside.6

The southwest corner of the basement is an old myth too. It survives from an era that assumed tornadoes always arrive from the southwest, but the winds in a tornado rotate, so debris can come from any direction and no corner is inherently safer. What you actually want is the middle, with as many walls between you and the outside as possible.6

Outrunning one in a car is the most tempting and the most situational. The SPC’s guidance is to either leave the vehicle for sturdy shelter or drive out of the tornado’s path, and if the tornado is far off and traffic is light you may have room to move at right angles to its track. What you should never do is try to outpace it along its own path in heavy traffic, at night, or in rain that hides where it actually is.6

Most people carry a mental map of tornado risk that runs up the middle of the country, Texas through Oklahoma and Kansas into Nebraska. That map is out of date, and if you live east of it the error runs in the dangerous direction.

Peer-reviewed work on the tornado record has found the pattern shifting. Analysis covering 1980 to 2022 reports negative tornado frequency trends in parts of the Great Plains while the Southeast United States shows positive trends, building on earlier findings from Northern Illinois University and NOAA that the frequency of both tornado reports and tornado-favourable environments has been climbing across the Midwest and Southeast and falling across portions of Texas, Oklahoma and northeast Colorado.7

This matters more than a line on a map. The classic Plains tornado arrives in the afternoon, over open country, where you can see it coming for twenty miles. A Southeast tornado is more likely to be wrapped in rain, hidden behind trees and hills, and moving through country where the visual warning people rely on simply does not arrive. If your plan is built on the idea that you will see it, and you live where you cannot, the alert on your phone is not a backup. It is the entire warning.

The image everyone carries is the mile-wide wedge grinding across a county for an hour. Real tornadoes are mostly nothing like that. Based on path-length data since 1950, the average tornado has travelled about three and a half miles, and most last less than ten minutes. They also tend to move in a consistent direction, most often southwest to northeast or west to east.6

That should change how you think about the risk rather than reduce it. A tornado is a narrow, short-lived event, which is exactly why sheltering works: you are not trying to survive a storm, you are trying to not be in a specific small place for a few minutes. It is also why "it will probably miss us" is technically good odds and terrible reasoning, because the cost of being wrong once is total.

The other half of the picture is what you are sheltering in. Mobile homes do not attract tornadoes, but they are, in the SPC’s words, "much easier for a tornado to damage and destroy than well-built houses," and most tornado deaths occur in them. Manufactured homes built after 1976 are somewhat stronger, but the SPC is careful to add the qualifier that matters: only if strongly anchored down.6

If you live in one, the plan is different in kind, not degree. There is no interior room that helps and no closet that counts. The whole plan is where you go instead, which means knowing the nearest community shelter or sturdy building now, knowing whether it is actually open at two in the morning, and leaving on the watch rather than on the warning. Thirteen minutes is enough time to walk down a hallway. It is not always enough time to drive somewhere.

Nearly all tornado deaths happen in areas that were already under a warning; the gap is usually getting the message and acting on it in time. ZoneWatch delivers real-time severe-weather alerts for your exact locations, so a warning reaches you and your family directly.

Set it up for home, school, and work, and pair it with your shelter plan so the alert triggers an action, not a scramble.