To that end, tornado science’s biggest breakthrough would come with the arrival of Doppler radar in the 1970s. Since 1953, it had been known that tornadic storms create a unique image on radar called a hook echo—so named because the radar picks up the arc of rain or hail that wraps around a strong mesocyclone. But before the arrival of Doppler radar—an advance that allowed radar to track movement in real time—meteorologists were unable to learn much about the supercell’s evolution or internal structure.
The first Doppler scan of a tornado would leave little doubt about just how much the technology had to offer. Like most advances in the field, it came about only with a rare alignment of chance elements. Throughout the sixties and early seventies, a small number of Doppler weather radars had started dotting the states, but it took until May 24, 1973, for a recordable tornado to form and touch down within range of any. That day, researchers at the National Severe Storms Laboratory in Norman, Oklahoma, noticed the first signs of a broad vortex forming several miles off the ground, and just within range of their Doppler. They scrambled to record its every moment—as the circulation focused, touched down, and marched six miles through the open country into nearby Union City. Their scan, of an entire tornado from birth to death, was historic—but the fruits of their labor took several more months to reveal themselves.
In the aftermath, researchers Rodger Brown, Donald Burgess, and Les Lemmon at the NSSL pored over every angle, every azimuth, on the magnetic tapes that recorded the day’s radar feed. When they struck upon a place where wind shear and velocities spiked over a short distance, they assumed they had found an error in the data. At that moment in the storm’s evolution, there had been no tornado on the ground; the increase in wind speed should have been smooth and gradual. Yet even when they checked different elevations, the same steep velocity gradient persisted. The researchers consulted the photographs taken from the field at the same time and identified a small funnel protruding from the cloud base, not yet in contact with the ground. In fact, it would be another forty-one minutes until touchdown.
What they’d stumbled across was no error. Their device had identified the first embryonic stages of a tornadic storm. In the race to warn those in the path, this was monumental. It meant that Doppler could reveal areas of concentrated rotation a few kilometers off the ground, a pattern detectable tens of minutes before touchdown. In the coming years, Doppler and follow-on advances would spread like wildfire across the meteorological community. While not all tornadoes produced visible radar signatures in their infancy—and not all signatures resulted in tornadoes on the ground—forecasters finally had some way of peering into the storm and spying a brewing tornado prior to its arrival. In the long history of tornado science, researchers could finally boast a clear victory over the vortex.
Through the thunderstorm, to the supercell, to the first spiraling wisps of the nascent twister, scientists had chipped away, steadily earning their insights. Yet as the field tried to isolate and discern the features of the funnel itself—What made the vortex descend? What determined its strength, or how long it would churn?—the tide suddenly turned.
Around 1980, after decades of scientific progress, the tornado seemed to set in its heels. At the lowest elevations—what’s referred to as the boundary layer—the twister refused scientists any further victories.
Tornadoes were such an improbability in the first place, such a miraculous confluence of variables that the trigger might be as inscrutable as chaos theory, and the beating of a butterfly’s wing. Vastly more data would be needed to detect any pattern. Doppler had proven its ability to capture in unprecedented detail the life cycle of a storm—but it had serious shortcomings. It struggled to capture the near-surface parts of the storm that scientists most prized. Unless a radar installation was practically run over, trees, houses, even the curvature of the earth, would conceal the vortex—to the point where scientists weren’t able to discern whether the wind was stronger near the surface or farther up. The boundary layer proved too low for the beam to reach.
It did not help that the technology was wholly reliant on chance encounters. Any particular spot in the country will meet a tornado, on average, once every 4,000 years. The odds were disconcertingly long that one would land within close range of a set of stationary meteorological sensors. By stringing mesonet instrument stations across the Oklahoma prairie through the sixties and seventies, the National Severe Storms Laboratory was able to improve the long odds. And the right storm in 1973 did yield its groundbreaking Doppler data set. But the run-ins were painfully rare, and storms that hit could be even more agonizing than those that missed: In 1977, when a tornado finally touched down near a site in Fort Cobb, it got too close. The site’s sensors were not designed to withstand tornadic wind speeds, and no useful data could be gathered.
Finally, there were a vast number of questions that Doppler would never be able to answer. The technology could say nothing about the temperature, humidity, or pressure inside the tornado. These are the fundamental data points of meteorology—what scientists consider essential to knowing where the wind will go, where it comes from, and what exactly is driving it. No fine-grain predictions of tornadic behavior would be possible without them.
Yet even as Tim Samaras takes SKYWARN classes and trawls Last Chance, these same basic questions remain unanswered. Upon reaching the unknowns at the tornado’s lowest levels, the science stumbled and sputtered.
A new tool, the next breakthrough, was needed.
In 1979, Dr. Al Bedard, a scientist at NOAA’s Wave Propagation Lab in Boulder, Colorado, knew it was time to risk a more daring approach. A stationary weather instrument wasn’t ever going to
