So they set out to understand the people on the receiving end of the forecasts. It wasn’t enough to farm the problem out to others. They needed people in NOAA studying the way Americans responded to warnings, and to risk. NOAA was an agency staffed by hard scientists facing a problem that cried out for psychologists and behavioral economists. “The odd group, whatever the odd group is, needs to be in the room,” she said. “There’s all sorts of inclinations not to do that. The existing powers say, ‘Leave me alone, and let me do what I want to do.’” She wanted to start a conversation inside the agency, with the understanding that they couldn’t predict exactly where it might lead.
It reminded her of something that had happened just after the Challenger explosion. American cities were planning to name streets and schools for the astronauts, but that had felt inadequate to her—and to the astronauts’ spouses. Everyone who’d been close to the astronauts wanted the meaning of their lives to be better understood through their deaths. “They all had this shared joy of bringing science and technology education to lots of people,” said Kathy. “We asked, how do we continue that?”
By the end of 1986, the astronauts’ families had decided to create a science education program—though of what sort they did not yet know. The spouses asked Kathy to figure it out. She started by bringing them all together, to explain how uncomfortable it was going to be to create an entirely new thing when they didn’t know exactly what it would be. They’d need to invite many odd groups into the room and give them the power to influence the project. “I told them,‘It’s your legacy to the crew. But to do it you need to create a network of people who feel they can shape it. The conversation really matters. Converse means exchange with. It does not mean transmit at. That’s how you get new thinking.’” She’d heard a line once that still resonated with her: The only thing any of us can do completely on our own is to have the start of a good idea.
She found all sorts of odd groups, outsiders to the space project, unknown to the astronauts’ families, who might be relevant to the new mission: teachers, museum professionals, curriculum supervisors, textbook publishers, exhibition designers, video-tech types, and so on. Plus, an architect. She gathered all these people in Biosphere 2, in Oracle, Arizona, “to get everyone out of their ruts.” Pretty quickly the architect turned the event into a presentation of his plan for the building. Kathy and the others could see that he hadn’t listened to a word anyone had said. She let him go the next day. In the end, the group discussion led to a course aimed at middle-school students. There are now fifty-two Challenger Centers around the world, and they have taught four and a half million students.
In the aftermath of the Joplin tornado, the odd group—the new kids in school—were the psychologists and behavioral economists. In 2014 Kathy helped to persuade Congress to write into law the idea that social science was part of NOAA’s mission. The agency could now hire people to collect a different kind of data—data that would enable them to figure out what exactly was going on inside the minds of the American people, so that it might save their lives.
The funny thing about tornadoes is that no one knows how powerful they are until they’ve hit something. The National Weather Service can tell you days in advance what to make of a hurricane—the strength of its winds, and the size of its storm surge, along with the likelihood of its hitting your city instead of someone else’s. As you sit on your porch in New Orleans deciding whether you should get in your car and drive to Memphis to avoid a hurricane, you have a pretty good idea what you are in for if you don’t. Tornadoes aren’t like that. Like the rest of the weather in the continental United States, they move from west to east, but the paths they take are random. Their force can be judged only after the fact, by the damage they’ve done. If a hurricane is another night in a bad marriage, a tornado is a blind date.
The scale for judging tornadoes, after the fact, runs from 0 to 5. It’s called the Fujita scale. What makes it different from most scales is that it is consistently terrifying from beginning to end. An F1 tornado merely peels roof surfaces off houses and knocks cars off the road. By F2, mobile homes are being destroyed and cows are flying through the air.
Kim Klockow was seven years old, playing in a field in Naperville, Illinois, when she caught sight of her first tornado. She didn’t know what she was seeing. “I saw the booby clouds,” she recalled—the breast-shaped mammatus clouds that accompany big storms. “I was looking at the anvil of the storm.” No one ever actually saw the tornado until it wiped out some of Plainfield, Illinois, on August 28, 1990. It had eluded radar and, wrapped in a rainstorm, had been invisible to the naked eye. The National Weather Service didn’t even issue a warning until an hour after the event. Afterward it would go down as the only F5 tornado ever recorded in the Chicago suburbs. In an F5, cars become missiles and big, well-constructed houses simply vanish. Kim’s parents had driven her through Plainfield two days later, and she’d seen buildings she’d been inside of reduced to rubble or entirely gone, like in The Wizard of Oz. “You don’t think of buildings as being dangerous,” she said. “You think of buildings as being a place you were safe.”
That tornado had killed twenty-nine people, injured hundreds more, and traumatized the region. The following year, as another storm approached, people were on edge. When the wind kicked up and the hail began to ricochet off the pavement, Kim was in the neighboring city of Joliet, with her mother and two-year-old sister, registering for French lessons. Her mother grabbed them and fled. As they sped toward home, Kim could see her mother watching behind them. “We were actually being chased by the storm,” she recalled. “The hail sounded like bullets hitting the car.” For some reason her mother insisted that the windows remain down: hail fell onto Kim’s lap. “My mother was saying the same thing over and over, but I didn’t know what it was. She was saying Hail Marys.” They peeled into the driveway and her mother screamed at her, “Get into the house, and get downstairs!” She’d run and hid—and came away with the feeling that it was only by luck that her house had not been blown away. “After that,” said Kim, “any weather information we got, I wanted to know. This is actually the story of every meteorologist. We are a whole field of people who are child trauma cases.”
One hot May morning I picked Kim Klockow up from her office at the NWS Storm Prediction Center, in the National Weather Center Building, in Norman, Oklahoma. The center is a joint venture between the University of Oklahoma and NOAA, and about as perfectly situated as an institution can be. The south-central United States is the planet’s convective sweet spot: here the warm air from the Gulf of Mexico collides with the cool air tumbling down over the Rocky Mountains and creates storms with more energy than nuclear bombs. Texas has twice as many tornadoes as Oklahoma, but Oklahoma has them in about a fourth the space. Kansas has about a third more tornadoes each year than Oklahoma, but Kansas is a third again bigger than Oklahoma and has a third fewer people. If you have some need or desire to witness dramatic collisions between people and weather, Oklahoma is your place. “Being here during a serious tornado event is better than football,” says Hank Jenkins-Smith, who runs the University of Oklahoma’s National Institute for Risk and Resilience—which is as aptly sited as the Storm Prediction Center. At the top of the National Weather Center Building is a skybox, facing west, and equipped with special blast-proof glass, to watch the approaching tornadoes.
Kim came to the University of Oklahoma in 2006 as a graduate student to study . . . well, she hadn’t been sure what she was going to study. She’d received her undergraduate degree in both meteorology and economics and, up to that point, focused on the economic impact of storms. What happens to the finances of a community hit by a tornado, for instance. The work interested her, but she also felt something was missing. “I just felt that classical economics wasn’t really hitting on the questions that meteorologists were asking,” she said.
Her frustration led her first into behavioral economics, which was no more than psychology made res
pectable to the sort of people who tended to think psychology was all bullshit. She set out to investigate a problem: How do people respond to risk? How might you influence that response, to their benefit? If you told someone that a tornado might be headed his way in a week, he’d give you a funny look and go about his business. If you pointed out to that same person the tornado bearing down on his house, he’d dive for cover. She wanted to figure out when and why complacency turned to alarm and when and why alarm turned into action.
In December 2010 she was finishing up her thesis when an adviser suggested that what she really needed to do was some fieldwork. Go out and interview real-live Americans who had responded to the news that their lives might be at risk. “They said,‘If anything happens in 2011, we want you to do a case study,’” said Kim. “Then Joplin happens.”
For complicated reasons, she set out to survey people not in Joplin but in Alabama and Mississippi. A few weeks before the catastrophe in Missouri, tornadoes had wreaked havoc in those states, despite excellent warnings from the National Weather Service. What became known as the 2011 Super Outbreak spawned 360 tornadoes that killed 324 and injured thousands more.
In its wake a pair of ideas sprang up and gained traction—both inside and outside the Weather Service. The first was that the twenty-minute warnings that had been issued had not given people enough time to escape. Powerful congressmen from tornado-prone states insisted that the National Weather Service needed to improve its ability to predict tornadoes to the point where they could warn people an hour in advance. And the National Weather Service had simply nodded and accepted the challenge. “Everyone in the Weather Service is so drawn to the mission of helping other people,” said Kim. “That’s what was so crushing about 2011. Oh, I may have just spent my entire career possibly doing nothing.”
But Kim wondered about the wisdom of their new ambition. “It’s hard to talk to dead people about the decisions they made,” she said. “It’s one of the challenges we have. But I was trying to ask what they would do if they’d had more time.” She interviewed survivors in Alabama and Mississippi and came away with a startling insight: time might be beside the point. It wasn’t that people who had apparently ignored the government’s alerts had been oblivious to them. “They were all aware of the warnings,” she said. “It isn’t that people wantonly disregard warnings. It’s that they think it won’t hit them.” The paper Kim subsequently coauthored pointed out that people associate “home” with “safety.” This feeling was reinforced each and every day that nothing horrible happened inside of it. People acquired a “false confidence that they would not be hit.” Some inner calculation led them to believe that, if it’s never happened here, it never will.
The people who had failed to seek shelter in the way that, say, a meteorologist thinks they should have done had one thing in common: they lived in homes that had never been struck by a tornado. They inhabited a region prone to tornadoes; they had lived through many tornado warnings; but right up until 2011 they themselves had been spared a direct hit. They offered Kim lots of explanations for their immunity to catastrophic risk. They claimed that tornadoes never crossed the river they lived on, for instance. Or that tornadoes always split as they approached their town. Or that tornadoes always followed the highway. Or that tornadoes never struck the old Indian burial grounds. People who lived on the west side of a big city felt more exposed than people on the east side: they believed buildings offered protection. A lot of people seemed to believe that hills did, too. “Where tornadoes go is totally random,” Kim said. “The steering winds are in the upper atmosphere. But people are not thinking of the forces of the atmosphere. They are thinking of their place on the ground.” Psychologists have long known that people see patterns where none exist. Londoners during the Blitz felt they’d deduced the targets of German bombers by where the bombs had fallen, when the bombs had been dropped randomly over the city. Americans routinely made the same mistake with the weather.
Soon we were driving west together, Kim Klockow and I. A few minutes after leaving the Storm Prediction Center, we passed from Norman into Moore, and from one wan row of shopping malls and car dealerships to another. Here was another curious example of man’s attitude toward the things that might kill him—and another illustration of Kim’s point. The people in Norman think that tornadoes don’t hit them; the people in Moore believe they are especially prone to being hit by tornadoes. Moore’s sense of doom dates back to May 3, 1999, when a tornado crossed the freeway and cut through the town. It was a mile wide and generated wind speeds of 302 miles per hour, the highest ever recorded on earth. It killed thirty-six people, including a woman who had sheltered exactly as experts had instructed, by lying in a bathtub and covering herself with a mattress. (A car crashed through her roof and landed on her.)
On May 20, 2013, another F5 tornado struck Moore and killed twenty-four people, including seven children in a school, after an interior wall collapsed on them. Between those two events, Moore had been hit by two F4 tornadoes and been dealt glancing blows by several small ones. By 2013 its reputation as a magnet for tornadoes was sealed. “The perception of risk of the people in Moore is about twice that of people living in Norman,” said Kim. Moore is the only town in Oklahoma to have adopted building codes to defend itself against the wind; it has even devised a scheme that allows worried parents to bus their children to schools that have storm shelters. “The people in Norman are less likely to start preparing during a tornado watch than the people in Moore,” said Kim. “The people in Norman think that Moore is more likely to be hit than Norman. And this might be the most educated population, about tornado risk, in the world. Hundreds of meteorologists live in Norman.”
The road to the weather of the future is straight and hot. It leads after an hour or so to the city of El Reno. “You can still see this one,” said Kim. “In the trees.” Eleven days after the 2013 Moore tornado, there had been another spin-up, right here. Within minutes, what became known as the El Reno tornado was 2.6 miles wide, the widest tornado ever seen, and headed for Oklahoma City. “Tornadoes leave scars that are visible from space, when they are big enough,” Kim says.
The second idea that gained traction after the 2011 tornadoes was that people simply failed to appreciate what happened when a tornado hit a mobile home, or a car, or really anything that wasn’t bolted to the ground. If the warnings highlighted the potential destruction, the thinking went, people might pay them more attention. “Impact based warnings,” the new warnings were called, though the differences between them and the old warnings were fairly subtle. The Weather Service did not generally communicate directly with the public. It issued warnings to local emergency managers and the TV meteorologists, who then passed on what they’d been told. But the Weather Service now encouraged the weather media to help people to imagine what might happen if they did not seek shelter. “The idea was the people just don’t know how bad it is,” said Kim. “If they knew how bad it is, they’d take action.”
COMPLETE DESTRUCTION OF ENTIRE NEIGHBORHOODS IS LIKELY. MANY WELL-BUILT HOMES AND BUSINESSES WILL BE COMPLETELY SWEPT FROM THEIR FOUNDATION. DEBRIS WILL BLOCK MOST ROADWAYS. MASS DEVASTATION IS HIGHLY LIKELY, MAKING THE AREA UNRECOGNIZABLE TO SURVIVORS.
And so on.
The market for weather news in Oklahoma is fiercely competitive. The local TV weather anchors already felt pressure to make the reality more interesting than it was. “They glom onto the worst-case scenario days before we can have any confidence,” says Kim. “A government agency does not have an incentive to hype. Private companies have an incentive to hype. The problem when you hype is that you reduce confidence in all weather forecasts, because no one knows the source of the information.” About thirty minutes before the El Reno tornado reached Oklahoma City, a TV weatherman named Mike Morgan told his viewers that anyone who wasn’t underground was doomed. Most people had no underground place to go. The soil in Oklahoma is a sandy clay floating on a high water table: the place on the planet
where people most desperately need to dig a hole to hide happens also to be a place in which it is expensive to dig. Though a car might be the single worst place to be in a tornado, tens of thousands of Oklahomans fled by car. Instantly the southbound lanes of the interstate became a parking lot. The El Reno tornado bore down on what amounted to a miles-long traffic jam. . . .
And then it lifted. By sheer luck the El Reno tornado killed only eight people—most of whom had been fleeing it. What didn’t happen did not get nearly as much attention as it deserved, in Kim’s view. “If it hadn’t lifted, if it had continued on its path, the estimate of the fatalities would have been Katrina-level. It’s the worst catastrophe that almost happened. In the most tornado-savvy population in the world. It was really jarring.”
El Reno had been her turning point. “It struck me: How could we think we could help people without understanding people?” she said. “The way we have approached things is by learning about the threat. We’ve ignored the people being threatened.” She thought that impact based warnings were intellectually dishonest: How could you warn about the impact of a storm whose force you would only be able to discern after the fact? She was also pretty sure that people knew what a tornado could do to them. The people in Alabama and Mississippi knew. So did the people in Joplin. Their problem, as she saw it, was a different sort of failure of the imagination. People could not imagine that all those tornadoes that had wound up hitting other people could instead have hit them. The sirens had become fake news. The government needed to find ways to make the news feel real.
The Fifth Risk Page 15