ON THE FRIDAY OF MEMORIAL DAY WEEKEND, I stood on the outskirts of Pampa, high in the Texas Panhandle, feeling like I was in the middle of a battlefield. I was with Martin Lisius and several other storm chasers, the kind of people whose lives inspired the characters in the movie Twister. But on this late afternoon, seeing the storm they had tracked down was far more frightening than watching monster twisters surrounded by Dolby sound. All hell was breaking loose as an ominous black cloud hovered overhead, throwing off bolts of lightning that sparked grass fires, which sent billows of smoke hundreds of feet high. We had parked on the road’s shoulder, and the chasers were setting up their camera equipment to capture the storm’s onslaught of hail, intense downpours, gusts of violent winds, and window-rattling thunderclaps. Close by, I spotted black swirls rising above one of the fires. “Tornado!” I thought and held my breath. But no, I was told, it was a gustnado, created by a downdraft from within the storm, and it dissipated as quickly as it had appeared.

Amid the fire and rain, the chasers kept their eyes fixed on a wispy dark cloud that emerged from the base of the storm. They were waiting for it to rotate, confirmation that the storm was a mesocyclone, or supercell. As I nervously surveyed the roiling mass, I was again certain a tornado would soon drop. But the storm chasers saw things in the explosive weather that I could not. The edges of the cloud base were not sharply defined, indicating that the storm was weakening. When cold blasts of wind shot out from the storm cloud, Lisius, a storm video producer and the founder of the Texas Severe Storms Association (TESSA), jumped back into his Jeep. There wasn’t the right mix of warm and cold air needed to make a tornado. As the grass fires glowed in the dusk beneath a flashing, angry sky, we futilely chased clouds south and west for another hundred miles before calling it a day.

That night, at the Travelodge East in Amarillo, a group of chasers gathered in the coffee shop, making plans for the next chase. Around midnight, the waitress asked them if they needed anything else. “Yeah,” somebody cracked. “An upper-level disturbance.”

FOR THE PAST MONTH I HAD BEEN LIVING AND BREATHING tornadoes, tuning in the weather radio, watching the Weather Channel for hours, and chasing phantom supercells. One day I’d driven from Arlington to north central Kansas in pursuit of a big storm but had seen only blue skies. Conditions, however, were improving. At noon the day after my Pampa experience, about twenty people were gathered in a conference room of the National Weather Service office in Amarillo, studying meteorological data.

“What do you think? Is this a good day for tornadoes?” asked Alan Moller, a lead forecaster for the weather service’s Fort Worth office. Many of the mostly male group nodded their heads in agreement. “It’s not a sure thing, but it’s never a sure thing,” answered Chuck Doswell, a researcher for the National Severe Storms Laboratory in Norman, Oklahoma. With Doswell and Moller were three of Moller’s colleagues, meteorology students from Texas A&M and Michigan Technological University, a climatologist from California, and a few meteorologists from the Weather Channel in Atlanta. The rest were plain old weather nuts. Moller and Doswell obviously had rank. When they talked, everyone listened.

The 46-year-old Moller and 50-year-old Doswell were chasing storms long before Doppler radar, computer modeling, cellular phones, laptops, and the Weather Channel became tools of the trade. They met in 1972, when they were meteorology students at the University of Oklahoma, and have been storm chasing ever since. They schedule their vacations for late May and early June, the best time for tornado outbreaks on the Great Plains, which is the site of more violent weather than anywhere else on the continent.

But this spring, Tornado Alley—roughly extending from Texas to Nebraska—had been quiet. Tornado Alley is the ideal point of convergence—it is where cool winds in the upper atmosphere collide with warm, moist air from the Gulf of Mexico to cook up the storms that spin off tornadoes. Yet the big storms, born from supercells whose cloud tops reach 60,000 feet, were hammering the Ohio Valley, not Tornado Alley, where the flat, treeless, lightly populated landscape and gridlike roads provide excellent conditions for storm tracking.

On this day the prognosis was good for the area around Amarillo: The polar jet stream had dipped down to provide the shot of cool upper-level winds crucial to break up the dome of warm air, also known as the capping inversion, which hovered at five thousand feet and prevented thunderstorms from forming. Also, a cool front from the north had stalled near the Texas-Oklahoma border and a low-pressure system was moving up from central New Mexico. The Lubbock office of the National Weather Service called it “the most favorable severe-weather scenario for West Texas so far this spring”—good news, if you happen to be a storm chaser. Moller and Doswell and some of the others determined that they had to start in Clovis, New Mexico, to get a piece of the action.

The hour-and-a-half drive southwest on U.S. 60 was uneventful. As the cloud cover broke into scattered “cues,” as cumulus clouds are called, Moller’s tape deck blasted weather-related blues—“That Mean Ol’ Twister,” by Lightnin’ Hopkins; “Texas Tornado,” by Tracy Lawrence; and “Lightnin’,” by Johnny Winter. Near Hereford, two dark-colored Ford Explorers with satellite dishes, whip antennas, anemometers on the roof (to measure wind speed), and signs identifying the vehicles as Severe Storm Spotters zipped past doing 90 miles per hour. The lead vehicle, according to the banner on the side, belonged to Warren Faidley, an adviser to Twister and the jut-jawed self-proclaimed “World’s Only Full-time Professional Storm Chaser.” “I used to think Faidley was an okay guy until he started taking himself so seriously,” Doswell said matter-of-factly.

The agreeably grumpy Doswell laid out his ground rules: “Don’t ask me how many tornadoes I’ve seen, don’t ask me how close I’ve been to a tornado, and don’t ask me if I get scared.”

“That’s because you’re too chicken to get close to one,” Moller said, laughing.

Doswell preferred to talk about the four decades of benefits brought by the work of storm chasers, the National Severe Storms Laboratory, and the National Weather Service despite constant budget cuts from Congress. “In the fifties, tornadoes killed hundreds of people every year. Now it’s a few tens a year. But the weather service is suffering from its success. We’ve done such a good job, no one thinks tornadoes are a threat. To get their attention, tornadoes have to kill people. The history of funding for tornado research follows the history of deaths from tornado disasters.”

Although Doswell and Moller rack up more than 10,000 miles on their chasing vacations each spring—shooting storm photographs and video—they’re lucky if they spot one or two big tubes a season. For this trip, Moller had added a two-way radio that he bought from Sam Barricklow, a fellow chaser from Dallas and a member of the Dallas-area Skywarn radio storm spotter network. Barricklow was following us, as were several others. But the conversation coming over the radio was irritating Doswell. “Why do I have a bad feeling about today?” Doswell said. “Because everyone’s talking about everything but the weather.” Just then he was interrupted by a report from the National Weather Service in Amarillo. The entire Panhandle and South Plains were under a tornado watch.

At a convenience store in Clovis the woman behind the counter couldn’t help but notice the T-shirts and gimme caps pro-moting Storm Track magazine, Twister, and TESSA on the guys in shorts milling around the store and standing out in the parking lot, heads tilted toward the sky. “Y’all aren’t tornado chasers, are you?” she asked. She’d just seen Twister with her husband. “Don’t believe the movie,” a chaser told her.

In the parking lot, Moller and Doswell identified three cloud structures with supercell potential, watching the cloud towers build, collapse, and reform. “They’re all explosive,” Moller said, pointing out the well-defined sides of the towers and the glaciation, or icing, of the cloud tops. The storm cloud forming to the north would be hard to track because its projected path was over ranchland with few roads. The two to the south were toss-ups. On a hunch, the chasers decided on the one blossoming closest to us. “These are starting a little early,” Doswell said. “Sometimes you can go after one and stay with it. Other times, they’re just sucker storms.” The storm clouds were moving fast. Moller hollered for the chasers to get into their vehicles and head east in hopes of intercepting the potential supercell along U.S. 84.

“It’s prime West Texas tornado country,” Moller whooped as we recrossed the state line. “The edge of that anvil is already straight-edged and hard. That sonofabitch is gonna be a supercell!” Four miles later, he suddenly veered off the road to a spot near a tractor tilling the bare red dirt. Five vehicles followed. Five more cars and trucks parked about a quarter mile farther up the highway. The chasers began unpacking cameras and setting up tripods. The storm cloud was a few miles away, and it was producing lightning, thunder, and downpours. Doswell observed a cut in the base, a sign of downdraft that often leads to mesocyclone rotation, but Moller noticed a stratoform, or scud cloud, pushing out from under the base. “That means there is outflow, which is not good for tornadoes,” Moller told me. From where we stood, humid wind continued blowing toward the storm, which sucked the energy up. The storm was strengthening. “This is just gorgeous,” Doswell said, awed by the spectacle. But the base continued expanding, then contracting. “I’m thinking we ought to go south,” Doswell said, but no sooner had we gotten into the car than Sam Barricklow, who was watching the storm from his van nearby, came on the radio. “This storm is starting to intensify and there is slight rotation.”

Martin Lisius broke in on the radio. He was north of us and was observing the same storm. He reported that a wall had dropped from the base. Moller pushed the speed limit on the two-lane blacktop, trying to catch up with the storm. Moller saw a banded collar above the base, like a barber’s pole, as the wall cloud began rotating, meaning a mesocyclone had formed. The storm was a supercell. “Chuck,” he said, his voice rising, “call the National Weather Service. They may have to upgrade this from a watch to a warning from what we can see.”

A funnel tail briefly dropped from the low cloud that was rotating violently now. Moller stopped the car. Then the tail retreated, and we resumed toward Bovina, speeding past fields just drenched by a downpour and past toppled farm machinery, indicating wind damage. The radio reported that a tornado warning had been issued. Our car now led a trail of headlights up the road in a storm-induced darkness. Moller madly steered the car through a thick fog rising from just-fallen hail, using his two-way radio to warn others behind us of the road hazard. East and north of Bovina, the wall cloud finally gave the chasers what they wanted, as an imperfect tornado dropped from the mesocyclone’s menacing tentacles onto the ground. But no sooner had it touched down than it began to dissipate. “It’s getting eaten up by the rear flank downdraft,” Doswell said.

Outside Friona, we whipped past two vehicles with flashing lights parked on the shoulder—storm spotters monitoring the situation. Danger was clear and imminent. Violent gusts shook the car and knocked off the radio antenna. In Friona the scene was unnerving, as people dashed from their homes to storm cellars. Emergency vehicles were stationed at the town’s main intersection. We pulled over at the north edge of town, and Moller and Doswell scrambled for their cameras. Another funnel was dropping from the rotating wall cloud.

“Yes!” Doswell cried as the tornado warning siren in Friona began to wail. “It’s well developed . . .” And then a minute later, “Damn! It broke up.”

“Let’s go get it,” Moller yelled, jumping back into the car. “It’s gonna do it again.”

Sam broke in on the radio. “I see some teeth around the old meso. It may be intensifying.” With Moller and Doswell in the lead, the caravan of chasers followed the storm north, then west. Suddenly Moller hit the brakes. “There’s a tornado on the ground to the left,” he shouted. The cameras focused on the funnel, which was indeed a tornado, though it was practically transparent. Except for the sweet song of a bird on a phone wire, there was an eerie silence. Then Moller yelled again: “We need to back up. It’s coming right at us.” The translucent cone was only a few hundred yards away and headed in our direction. Moller shifted into reverse, stopping at a safe distance where other chaser vehicles were pulling up.

After about a minute, the funnel fell apart. Up the road, wisps of a low black cloud circulated on the ground near a farmhouse. “That was your quintessential low-level mesocyclone,” Doswell said, getting back into the car as other vehicles were still arriving.

“Let’s try to think ahead of everyone else,” Moller said, speaking into the radio. “Sam, you made a great call on this storm early on, when I thought it was a dead duck.” We raced north into Deaf Smith County, passing a truck with flashing lights and a huge radar on the trailer. The National Severe Storms Lab’s portable Doppler radar was taking the pulse of the supercell.

A new mesocyclone was developing out of the cloud base. Trying to get ahead of the storm, we drove into Hereford, then zigzagged north and west about 25 miles to intercept the storm. It was two miles ahead of us, but we could clearly make out the funnel’s silver-gray elephant’s trunk dancing across the range, flashing its power for several minutes before retreating skyward. Watching it filled me with awe and, bizarrely, a desire to get closer. Though nowhere near an F5, which is the most dangerous twister, it was a lethal beauty.

We continued our chase for another forty miles until Moller and Doswell reluctantly agreed it was time to give up and start heading south in search of another storm. “Well, we started out this year getting a tornado,” Doswell said. “Now we’ve got to get a photogenic tornado.”

“This is where the reality is far better than the fantasy,” Moller said.

“You were pretty lucky,” Doswell told me, “seeing all that on your first day with us.” 

At dusk we watched one last storm near Littlefield, vainly trying to make out the silhouette of a rotating meso. Then we headed to the twinkling lights of Lubbock for the night. At the entrance of the Hub City Brewery, Moller and Doswell ran into several other chasers, including Betsy Abrams and Matt Crowther, husband-and-wife meteorologists on holiday from the Weather Channel in Atlanta. They too had started the day in Amarillo, but they had headed down south and west of Lubbock instead of to Clovis. The inevitable question was quickly asked: “See anything?” Abrams reported that they had spotted a gustnado, which they videotaped, and had run into baseball-size hail. That was it.

In the middle of the late-night meal, a wild thunderstorm passed over the city. Moller couldn’t restrain himself from going out to watch the action. As if on cue, a lightning bolt hit a pole across the street, briefly turning it a glowing red. A waitress squealed with fright. “Sorry,” Moller said. “We probably brought that with us.”

LUBBOCK WAS THE PERFECT PLACE for me to end the chase, because with the storm comes storm debris, the subject of study at the Institute for Disaster Research and the Wind Engineering Research Center on the campus of Texas Tech University. The research centers were created after the 1970 Lubbock Tornado caused 26 deaths and $135 million in damage and wrecked the city’s tallest building. “We realized that Mother Nature had given us a $135 million laboratory in our own back yard,” said James McDonald, the director of the Institute for Disaster Research.

With funding from the National Science Foundation, and later from the Nuclear Regulatory Commission, which was trying to establish baselines for tornado-proof nuclear reactors, the institute has studied wind, tornado, hail, wind shear, and other severe-weather-related events throughout the region, as well as hurricanes on the Gulf Coast. Texas Tech also has a field lab that features a metal building, loaded with sensors, that circles on a railroad track to test wind pressure and wind stress. There is an indoor wind lab with air compressor cannons that fire fifteen-pound two-by-fours, metal pipes, and chunks of hail at speeds of about 100 miles per hour at various wall materials.

Documenting damage from sixty storm events, the centers’ studies have yielded some interesting results. Most wind damage is done by weak tornadoes, with winds from 125 to 150 miles per hour. Opening a window in advance of a tornado is no longer recommended because wind damage is caused by the Bernoulli effect, the tendency of wind to speed up as it moves around and over objects. Using reinforced masonry in buildings is essential to avoiding disasters like the 1987 Saragosa Tornado. “It wasn’t wind that killed people. It was the falling concrete blocks,” said Kishor Mehta, who oversees the wind lab. Since a tornado-proof home is economically unfeasible, Texas Tech researchers came up with the in-home storm shelter, essentially reinforcing the walls and ceiling on an inside room, for less than $2,500. “The mobile home is still very susceptible to tornado damage,” McDonald said. “Every mobile-home community should have enough shelters so that no one has to go more than a hundred and fifty feet to be safe.”

Even disaster specialists must sometimes answer the call to chase, as Richard Peterson, a meteorologist who works with Mehta and McDonald, did when a farmer near Matador called Peterson after hail the size of grapefruit fell on his land the night before. “He says he’s got a three- by five-inch specimen in his freezer, but I’ve got to come get it today or he is going to let it melt.”