Forecasters had seen it coming for days – an angry blob of bright pink blossoming on forecast maps over a nearly 175,000-square-mile area of the Great Plains. For only the second time in its history, the National Weather Service held a pre-outbreak press conference Friday, April 13, alerting the region – much as it does for hurricanes ahead of landfall – to brace itself. Conditions were ripe for a significant tornado outbreak.
By nightfall Saturday, the last in a parade of roiling supercells, the thunderstorms that spawn twisters, had marched northeast, and some TV meteorologists in Oklahoma City were saying the worst was over. So residents of this cattle, wheat, and oil town along the path the storms had taken northwest of Oklahoma City went to bed relieved. All that was left was the squall line that TV forecasters said would bring strong storms – but nothing like that day's supercells.
By midnight, grandchildren were asleep in the Lord home here, and the adults were getting ready to follow. Chad Lord, a storm-savvy contractor up from Florida to visit his parents, made one last check for tornado warnings. Nothing on the National Oceanic and Atmospheric Administration weather radio. Nothing on TV.
As Chad was about to turn in, warning sirens began to wail. The squall line had broken into supercells; a tornado was bearing down.
As he recalls it, about two minutes passed as he rallied three generations of Lords to run across the street to the neighbors' basement. His father, Paul, was halfway across the road, and Chad was in the front yard beckoning others to follow when the rising locomotive-like roar and clatter of debris hitting homes stopped them in their tracks. The tornado was perhaps 100 yards away, and closing.
Both men turned to sprint back into their house as Chad shouted for the rest of the family to take shelter in an interior room.
Paul tripped, but nearly caught up as Chad reached the front door. Chad turned and gripped his father's hand to pull him into the house just as the twister hit. Winds yanked Paul out of Chad's grasp and into the night and blew Chad back down a hall, where he came to a stop near a bathroom. He opened the door, he recalls, hoping to find shelter in a tub, only to find nothing on the other side of the door.
Then a fireplace collapsed, covering him in debris.
In 10 seconds, he says, it was over.
Survivor awe and science
"It makes you wonder how people survive these things," says National Weather Service (NWS) meteorologist James LaDue, with a gentle shake of his head, as he stands the next afternoon amid the shattered wood, shards of glass, and tufts of fiberglass insulation that once made up the Lord family's high-ceilinged two-story home.
The Woodward tornado killed six people, including two children, and injured 29. Paul Lord survived – managing to curl up in a gutter and ride it out – with only minor injuries and the survivor's awe that Mr. LaDue and his colleagues in the forensics of tornadoes witness regularly.
LaDue operates at the intersection of science and humanity in making the initial assessment of the impact one of nature's most violent phenomena has on individuals and communities. Their work and that of weather researchers, engineers, and public-health officials around the country – before and after storms – has advanced the science of tornadoes and the potential for earlier warnings and safer structures.
Tornado-warning lead times are longer than ever – at 14 to 19 minutes on average – and may be about to increase dramatically because scientists are uncovering new features within thunderstorms that could serve as early-warning signs that a tornado is likely to form soon. Forecasters are getting new tools that can reveal in real time the inner workings of a thunderstorm, even as researchers are devising new ways to sort through the information to provide warnings with longer lead times and with more tightly focused warning areas. And engineers are undertaking new initiatives to identify construction techniques tailored to improve a home's ability to withstand the unique mix of winds a tornado delivers.
LaDue, from the NWS Warning Decision Training Branch at the US National Weather Center in Norman, Okla., and Rick Smith, warning coordination meteorologist with the NWS forecast office in Norman, arrive in Woodward the next day to collect information aimed at verifying and improving warnings. Their findings also determine the intensity of tornadoes on a scale from 1 to 6.
LaDue is a self-described "storm weenie" whose weather fixation began in junior high and was cemented later in rough-weather sailing in Long Island Sound. He often travels chasing tornadoes and their aftermath with his weathercentric family: his wife, Daphne, a University of Oklahoma weather researcher, and 4-year-old son Dylan, who loves to see a tornado spin into view.
Looking for damage patterns in structures and vegetation to determine a tornado's intensity – as well as width and length of track – is LaDue's just-the-facts-ma'am science job; but he is fully aware that his work site is often someone's personal disaster.
In approaching property owners, he says, "I don't want to treat them as victims. They're not to be pitied. They went through a horrible event, and if they come out reasonably OK, most wind up being thankful it wasn't worse."
He marvels at the hospitality and humor people show, despite their hardship. His serving as a patient listener helps. Many are just happy to have someone to talk to about their experience. During last spring's surveys in Alabama where 62 tornadoes killed 253 people, LaDue recalls: "We went to one house east of Tuscaloosa that was swept clean. The grandparents were in the hospital. Two [teens] were guarding what was left of the place. They said: 'We're so glad you came up and stopped to talk. Most people just gawk at us and drive on by.' "
Talking to residents is part of the boots-on-the-ground fact-finding that provides a reality check on the accuracy of forecasts for the storm that drops a tornado. And the investigations provide ground-truth tests for NWS tools for analyzing and predicting storm behavior.
The NWS is midway through a major upgrade to its network of weather radar that, by next spring, is expected to give forecasters a more detailed look at the inner workings of a storm and possible tornado triggers. It will also be able to spot tornado debris clouds, providing ready confirmation that a funnel cloud actually has reached the ground.
Researchers are working on another radar system that can scan the skies faster than current radar, with electronically steered beams that will allow forecasters to spot a twister radar signature several minutes earlier than mechanically rotating antenna now do. That means more lead time for people to reach shelter.
Also in the works is a warn-on-forecast system that will allow forecasters to routinely give at least 40 minutes' warning along a far more specific path than tornado warnings now cover.
US is the world capital of tornadoes
For all their tightly wound fury, tornadoes are very hard to produce and require specific geography. And nowhere else in the world hosts such optimal conditions as the Great Plains for building the towering thunderstorms that spawn twisters. So that's where scientists focus their inquiry on the forces that cause twisters.
Most significant tornadoes appear below so-called supercell thunderstorms – powerful, usually isolated storms. Only about 10 to 20 percent of these roiling behemoths form twisters, says Donald Burgess, a researcher with the Cooperative Institute for Mesoscale Meteorological Studies at the University of Oklahoma. And when they do form twisters, 80 percent of the tornadoes fall into the two weakest categories of the six-level tornado intensity scale.
Supercells rely on four basic ingredients to form: a source of warm, moisture-laden air near the ground and colder air at higher altitudes; a shift in wind speed or direction with altitude – known as wind shear – within a few thousand feet of the ground; something to trigger the rise of that low-level warm air; and a landmass that is closer to its hemisphere's pole than the source of the warm, moist air.
These ingredients are present elsewhere, such as South America, southeastern China, Bangladesh, and on the Tibetan Plateau, notes Paul Markowski, an associate professor of meteorology at Pennsylvania State University in State College. "But no place do these conditions occur on as vast a scale as the Great Plains of the United States," he says.
The Gulf of Mexico supplies warm, moisture-laden air that moves north near ground level. Winds flowing eastward over the Rocky Mountains provide an overlying layer of cold air. The initial lift can come from daytime heating, from wind deflected up as it blows across uneven terrain, or from a cold front moving east.
As the warm air rises and cools, water vapor in the air condenses, releasing back the heat that turned Gulf of Mexico seawater into a gas in the first place. This newly introduced heat source warms the surrounding air, giving the air parcel additional lift. If the surrounding air continues to get colder with altitude, the relentless flow of moisture-laden air into the updraft forms a sprawling, anvil-shaped thunderhead.
Meanwhile, wind shear at mid-levels sets up a rolling pin-like circulation of air drawn into the updraft, imparting rotation to the core of the storm.
All that happens well above the ground, Dr. Markowski says, but adds: "It can't ever explain why you would have spin right at the ground."
But maybe VORTEX 2, the largest tornado field study in history, can. During the 2009 and 2010 spring tornado seasons, more than 100 scientists crisscrossed the central US chasing supercells in hope of capturing the entire life cycle of tornadoes. Their tools: mobile radars, weather balloons, weather stations on wheels, and radio-controlled aircraft laden with instruments.
A picture of the final stages of tornado formation is still emerging from the data. But, says Markowski, spin at the ground appears to result from interplay between dense, rain-cooled air descending on the backside of a supercell's rainy core – a kind of exhaust system for the thunderhead – and a second source of shear near ground level that the storm itself can set up. In essence, it's a small-scale version of the process that forms the supercell in the first place. Typically, air in a thunderstorm's rear downdraft hits the ground, spreads, and dissipates. But under the right conditions, some of that air can get caught in small-scale updrafts in the rain-free base at the rear of the supercell.
If this cooler air has the right mix of temperature and moisture, the U-turn can result in a rotating "wall cloud" that descends from the rain-free base. If the storm's self-generated, low-level shear encounters the updraft, it is tugged upward, spins up, makes contact with the ground, and produces a tornado.
But are there early-warning signs of tornado formation? Enter the "blob." Researchers discovered that in some storms they intercepted, the downdraft seemed to flow in pulses that might be linked to discrete blobs of rain that the team's mobile radar picked up along the storm's rear flank. One of these blobs was present in a storm that generated a powerful tornado in Goshen County, Wyo., on June 5, 2009.
Some five to 10 minutes after the blob hit the ground, spin rapidly intensified and a tornado formed, says Markowski, who served on the project steering committee.
The radar units that picked up the blobs operate in the same way the NWS's weather radars will operate once the upgrades are complete. Known as dual-polarization Doppler radar, the upgraded radar will send out horizontal and vertical pulses that return detailed information on the size and distribution of particles in the clouds and clues to temperatures in the storm.
How to withstand vertical wind
When David Prevatt walked through residential areas in Joplin, Mo., and Tuscaloosa to survey damage after the tornadoes that hit each city last spring, he noticed an intriguing pattern: If a home wasn't obliterated, a house built before the 1940s tended to sustain less damage than homes built as recently as the mid-1990s.
Dr. Prevatt, an assistant professor of civil and coastal engineering at the University of Florida at Gainesville, noticed that the exterior walls, as well as the floors and roofs of the pre-1940s houses, typically were covered with 2-by-12-inch planks, set diagonally. More-heavily damaged postwar homes were sheathed in thinner, less-expensive plywood sheets.
It's "anecdotal evidence that may suggest that because we have changed the structural system" for homes, "we've changed the capacity and performance of these houses" to endure the stresses tornadic winds impose, he says.
In January, he received a National Science Foundation grant to explore ways to design new homes and retrofit old ones to resist tornadoes.
For all the attention paid to designing buildings to resist earthquakes and hurricanes, no generally accepted building standards yet exist in the US for tornado resistance, suggests a report by a team of researchers, including Prevatt, who analyzed damage from a tornado rated as a 4 on the Enhanced Fujita (EF) scale in Tuscaloosa and Birmingham, Ala., last spring.
One design, foam-and-concrete domes – built small as backyard shelters, mid-size as houses, or large for use as school gyms or churches – have proved robust in the face of even powerful tornadoes. Windows and doors may get blown in, but the structure remains intact.
LaDue testifies to the dome effect. "You're preaching to the person that surveyed one in Blanchard last May," he said when the subject came up during the drive to Woodward.
On May 24, 2011, an EF4 tornado plowed 32 miles across Oklahoma, injuring 15 people and killing one. During his survey of the damaged town of Blanchard, he came across an elderly couple's dome home that "performed remarkably well." Like water flowing around a rounded boulder in a stream, tornado winds easily flowed over and around the building, leaving the overall structure intact.
But most building departments have little or no experience with domed structures, so permitting can be as difficult as developing the dome's aesthetic appeal, advocates acknowledge.
Current codes don't address tornadoes well, Prevatt explains. Design standards account for horizontal winds, but tornadoes include vertical winds. Only recently have researchers developed laboratory tools that allow engineers to effectively estimate what those stresses may be on a full-scale building, he says. For now, engineers have no idea of how strong vertical stresses are at key points in a structure, he adds.
Moreover, the twister's winds hit like a boxer's punch, compared with the slow, steady increase in winds that come with an approaching tropical storm or hurricane.
Cost versus risk is an issue, too, says Prevatt. "It's not as much of an engineering problem as it is a societal one, accepting what its risk is and what it's willing to pay for its safety."
Last year 1,700 tornadoes may have touched down in the US, but the region at most risk covers 2 million square miles: The likelihood of a tornado hitting any single spot in a given year is tiny. But when that spot is shared by homes, schools, and businesses, the effect on individual lives and the community can be shattering.
Even in the absence of specific tornado standards, much more could be done to improve structures' resilience – at least with new construction, says Timothy Marshall, a meteorologist and principal engineer at Dallas-based Haag Engineering, a forensic engineering firm.
Today's codes are minimum standards. But the goal, Mr. Marshall argues, should be to exceed code. He notes that homes everywhere could benefit from the strapping and anchoring to secure roofs and walls that hurricane-prone states now require.
"It's going to cost more, but not a lot more – $500 to $1,000 more," he says.
He sounds a note of frustration at poor construction practices he found during surveys following severe tornadoes in Greensburg, Kan., in 2007 and the outbreak in the South last spring.
"It's like an offering to a tornado," he says.
He's heartened, though, by success stories in these events – people who'd installed safe rooms or shelters that allowed them to survive.
Race to investigate
Typically, a damage survey is a race against time. Residents aim to put their lives back together as quickly as possible.
"You have to work hard and fast to get the information before it's gone," observes Tanya Brown, who does post-storm surveys as a research engineer with the Insurance Institute for Business and Home Safety.
When LaDue pulled in to Woodward about noon the day after the tornado, police officers, firefighters, and other first responders had already been working nearly 12 hours.
LaDue's first step was to map the tornado's path. In Woodward, the damaged area was small enough to map by driving through neighborhoods until he ran out of visible damage.
Two areas drew particular attention: a trailer park, where the tornado pulverized several mobile homes; and a neighborhood a mile away, where the Lord home once stood.
His main task was to determine the tornado's intensity and wind speeds – estimated from the damage it inflicted. The tools: a tape measure, a camera, and a smart phone with software that contains a rogue's gallery of damage images that would help LaDue and Mr. Smith classify the twister on the EF scale.
During the May 24, 2011, outbreak over western and central Oklahoma, one twister picked up a Chevy Avalanche and carried it half a mile. "It was unrecognizable, a foot-wide piece of metal wrapped around a tree," says Smith, LaDue's NWS colleague. "Right in the core of what we believe was an EF5 tornado, an oil rig was pulled out of the ground – despite some 3 million pounds of force holding that heavy steel mechanism in the ground – and heavily damaged."
Damage indicators on the EF scale crib sheet don't include 1,500-ton oil rigs, Smith quips. Damaged and displaced vehicles are far more common, but aren't covered either. Despite hints from other damage indicators about a twister's intensity, its vehicular mayhem "really makes you scratch your head and wonder: Gosh, what kind of force, what kind of wind, and what kind of motion does it take to carry a giant pickup truck a half mile," shredding it in the process?
The neighborhood LaDue is surveying is a mix of homes built in the 1950s and – like the Lords' – homes built as recently as the '90s. As he works his way through the neighborhood, he stops to speak with residents for information on a damaged home's pedigree.
The need to work hard and fast doesn't trump compassion, however, or an ability to be amazed, even touched, at how residents spontaneously rally to help their neighbors or by remarkable stories of endurance and survival.
The tornado hit around 12:19 a.m. and by 1, more than 100 volunteers were combing the stricken area to look for survivors and begin to help them clean up.
"It's one of those small, Mayberry towns. Everybody knows everybody, and everybody lends a hand," says Chad Lord, above the buzz of chain saws and rumble of front-end loaders.
Early in the survey, LaDue works through the older part of the neighborhood, getting owners' permission to take photos and make measurements when he sees something that may have contributed to the severity of the damage. At one of the postwar-construction homes, he kneels to examine the rim of a foundation exposed when the tornado blew an exterior wall in. He hunts for signs of anchor bolts, and finds none. Exposed framing on interior walls reveals 1-by-3-inch wall studs, in contrast to the 2-by-4s used today.
Garages seem to be a weak link. Many homes have two-car garages with a single metal door with no braces on the back. On the periphery of the tornado's path, one house is intact, but its garage door is buckled inward. The garage doors of a house that was closer to the tornado blew in completely, which peeled the roof off.
"Notice the things not damaged," LaDue says. A room on the back of one blown-out garage is still standing. "This would have been a good place to take shelter," he says.
At this point, LaDue says he's inclined to give the twister a solid EF2 rating – indicating three-second wind gusts of between 111 and 135 miles an hour at the points where damage occurred – wind speeds comparable to a category 3 hurricane packed into a narrow funnel.
But then he walks up the block where volunteers are working through rubble that was the Lords' house, helping the family salvage what they can. As LaDue approaches and scans the debris and the foundation rim, he quietly observes: "This is an interesting candidate – all walls down, large house, anchor bolts."
Another two-story home next door is still standing, with low-end damage from the tornado itself, and collateral damage from the collapse of the Lords' home. At this point on its path, the tornado was one house wide. "A drill press," LaDue calls it.
Sue Lord, Chad's mother, approaches, and LaDue asks, "Were you here?"
"Yes, we rode it out," she says, explaining that her son Chad's warning sent her and other family members into an interior bathroom. Sue's daughter and grandson hunkered down in a bathtub; her other grandson huddled as far under the toilet tank as he could get, with Sue shielding him. Sue's son-in-law crawled under a built-in makeup table. The twister's suction pulled the toilet out of the floor, she explains, yet she and her grandson were unscathed.
Despite the harrowing experience of losing his dad's grasp in the twister, Chad marveled at the tornado's capriciousness. One moment, an ottoman was in front of a couch, complete with leg coasters. The next moment, the ottoman and couch had reversed positions – with the coasters still under each leg.
"Everything on top of you and you still made it," LaDue says with a tinge of awe. "You guys are surely blessed."
Over time, documenting experiences like this one, as well as those that end more tragically, can put one's own life and family into a new context, say LaDue and others involved in post-storm survey work.
For Smith, who acknowledges that on the open prairie, a supercell and tornado can still be things of beauty, "each time I go and walk through storm damage, it tempers that. It reminds me over and over that this is not a video game or a TV show. When we see those blobs on radar and there are tornadoes in those storms, this is a life-changing event" for someone.
LaDue says that he and his wife built their home in Norman before they were as keenly aware of the range of damage tornadoes can inflict – and they're now upgrading and are in the market for a storm shelter.
Haag Engineering's Marshall, who not only conducts surveys for Haag's clients but also works with the NWS on surveys and was one of the architects of the new tornado-intensity scale, says the work has changed him in profound ways. He doesn't put as much stock in personal possessions as he used to, before he started his work in forensic engineering. The resilience of survivors he's interviewed "has shown me that it really doesn't matter." The loss of a couch, a piece of china, or even a treasured photo is "not worth getting so upset about if you've got your life."
'It was kickin'!'
Sue and Chad move back into the flow of salvaging what they can from the debris when a voice booms out: "Hey, are you guys from The New York Times?"
Up strides Paul, his brow swathed in a thick layer of white gauze. "I don't know how my family came out of it," he says, voice cracking slightly. "It was a gift of God." Then he asks: "What are they rating this?"
LaDue replies that he's working on that.
"I can tell you how fast the wind was blowing," Paul deadpans. Then, he exclaims: "It was kickin'! It ... was ... kickin'!"