Chapter One The Frontier Overhead Nihil est in intellectu quod non antea fuerit in sensu. (All knowledge of the world must rest finally on one's sensory experience.) --John Locke I was a young child playing outside our house near Boston under a hazy yellow June sky in 1953 when my mother summoned me inside because a tornado had been reported in Worcester, about forty miles to the west. As further inducement to getting me in the house, she told me that tornadoes snatch children up into the air and abduct them. We didn't get many tornadoes in Massachusetts, but I knew what happened to Dorothy and her dog, Toto, in The Wizard of Oz . As it turned out, my mother needn't have worried. But in Worcester, the tornado killed ninety people, injured twelve hundred, and caused some $52 million (in 1953 dollars) in property damage. That year was a big one in the United States for tornadoes: A total of 516 people were killed and many hundreds more were injured. Most of the carnage took place in three days from tornadoes in Waco, Texas (May 11), Flint, Michigan (June 8), and Worcester (June 9). In those days, the U.S. government's Weather Bureau did not issue tornado forecasts very often, at least not publicly. My mother knew there was a tornado nearby because she heard about it on television. Tornado forecasting can be traced back at least to the mid-1880s, when the meteorologist J. P. Finley of the U.S. Army Signal Service (later called the Signal Corps) dared to suggest that tornadoes were predictable. In 1883, however, the U.S. government banned the word tornado From forecasts to avoid panicking the masses. As the chief Signal Service officer explained in his report for 1887: "It is believed that the harm done by such a prediction would be greater than that which results from the tornado itself." The ban was lifted in 1886, only to be reinstated the following year. The ban was lifted again in 1938, but only occasionally was the word tornado used and then only in forecasts issued to officials involved in disaster-relief efforts. On March 20, 1948, a tornado struck Tinker Air Force Base, near Oklahoma City, and inflicted some $10 million in property losses. Five days later, when similar atmospheric conditions again presented themselves, two meteorologists at Tinker, Major Ernest Fawbush and Captain Robert Miller, issued a tornado warning, although this was not made public. A tornado did indeed strike, causing about $6 million in damages. It was a lucky forecast. The odds of two tornadoes striking essentially the same place less than a week apart are infinitesimal. Perhaps it was this rare coincidence, which could not have escaped the meteorologists, as well as their incredible luck that sparked an unusual two-way exchange of government weather experts. During the following two years, the U.S. Weather Bureau in Kansas City invited Fawbush and Miller to discuss their forecasting procedures, and Tinker Air Force Base invited research meteorologists from Washington, D.C., to Missouri for talks. Of course, Fawbush and Miller had no special forecasting technique other than to examine weather maps and to determine that the maps looked the way they did five days before, when there was a tornado. The ultimate consensus of these meetings was that tornadoes really could not be forecast with much accuracy and therefore there was no good reason to change the existing policy of keeping tornado forecasts under wraps. It wasn't until March 1952 that the U.S. Weather Bureau (the old Signal Corps) began to issue tornado forecasts publicly. Rather than cause panic, the warnings drew a good deal of criticism for their inaccuracy and lack of precision. Our ability to predict tornadoes is still rudimentary. At best, we can say that in eight to twelve hours there might be some storms that might produce one or more tornadoes over a very broad area, sometimes covering several states. An hour or so in advance we can narrow down the area that might be affected, but we still cannot say whether a storm will bring a tornado. Much about tornadoes remains mysterious. Tornadoes are often called one of the last frontiers of atmospheric science. We know so little about them because they are so hard to observe and study. Tornadoes are elusive. Most people have never seen one. My experience--more than twenty years of studying and chasing tornadoes--attests to the fact that the chances of catching one are remarkably slim. Will Keller, a Kansas farmer, says he peered into the heart of a tornado's funnel cloud. According to his account, first published in the journal Monthly Weather Review in 1930, the tornado appeared in the late afternoon of June 22, 1928. Keller says: I was out in my field with my family looking over the ruins of our wheat crop which had just been completely destroyed by a hailstorm. I noticed an umbrella-shaped cloud in the west and southwest, and from its appearance, suspected that there was a tornado in it. The air had that peculiar oppressiveness which nearly always precedes the coming of a tornado. But my attention being on other matters, I did not watch the approach of the cloud. However, its nearness soon caused me to take another look at it. I saw at once that my suspicions were correct, for hanging from the greenish-black base of the cloud was not just one tornado but three. One of the tornadoes was already perilously near and apparently headed directly for our place. I lost no time therefore in hurrying with my family to our cyclone cellar. The family had entered the cellar and I was in the doorway just about to enter and close the door when I decided that I would take a last look at the approaching tornado. I have seen a number of these things and have never become panic-stricken when near them. So I did not lose my head now, though the approaching tornado was indeed an impressive sight. The surrounding country is level and there was nothing to obstruct the view. There was little or no rain falling from the cloud. Two of the tornadoes were some distance away and looked to me like great ropes dangling from the clouds, but the near one was shaped more like a funnel with ragged clouds surrounding it. It appeared to be much larger and more energetic than the others and it occupied the central position of the cloud, the great cumulus dome being directly over it. As I paused to look, I saw that the lower end, which had been sweeping the ground, was beginning to rise. I knew what that meant, so I kept my position. I knew that I was comparatively safe and I knew that if the tornado again dipped, I could drop down and close the door before any harm could be done. Steadily the tornado came on, the end gradually rising above the ground. I could have stood there only a few seconds but so impressed was I with what was going on that it seemed a long time. At last the great shaggy end of the funnel hung directly overhead. Everything was as still as death. There was a strong gassy odor and it seemed that I could not breathe. There was a screaming, hissing sound coming directly from the end of the funnel. I looked up and to my astonishment. I saw right up into the heart of the tornado. There was a circular opening in the center of the funnel, about 50 or 100 feet in diameter, and extending straight upward for a distance of at least one half mile, as best I could judge under the circumstances. The walls of this opening were of rotating clouds and the hole was made brilliantly visible by constant flashes of lightning which zigzagged from side to side.... Around the lower rim of the great vortex small tornadoes were constantly forming and breaking away. These looked like tails as they writhed their way around the end of the funnel. It was these that made the hissing noise. I noticed that the direction of rotation of the great whirl was anticlockwise, but the small twisters rotated both ways--some one way and some another. The opening was entirely hollow except for something which I could not exactly make out, but suppose that it was a detached wind cloud. This thing was in the center and was moving up and down. The tornado was not travelling at a great speed. I had plenty of time to get a good view of the whole thing, inside and out.... Its course was not in a straight line but zigzagged across the country in a general northeasterly direction. After it passed my place, it again dipped and struck and demolished the house and barn of a farmer by the name of Evans. The Evans family, like ourselves, had been out looking over their hailed-out wheat and saw the tornado coming. Not having time to reach their cellar, they took refuge under a small bluff that faced to the leeward of the approaching tornado. They lay down flat on the ground and caught hold of some plum bushes which fortunately grew within their reach. As it was, they felt themselves lifted from the ground. Mr. Evans said that he could see the wreckage of his house, among it being the cook stove going round and round over his head. The eldest child, a girl of 17, being the most exposed, had her clothing completely torn off, but none of the family were hurt. I am not the first one to lay claims to having seen the inside of a tornado. I remember that in 1915, a tornado passed near Mullinville and a hired man on a farm over which the tornado passed had taken refuge in the barn. As the tornado passed over the barn, the door was blown open and the man saw up into it, and this one, like the one l saw, was hollow and lit up by lightning. As the hired man was not well known no one paid much attention to what he said. Keller seems to be a credible observer and, at least to a meteorologist who studies tornadoes, a very lucky one. The only time I've ever seen inside a tornado funnel was on a radar display. Nor have I ever heard any "screaming" or "hissing" from the hundred or more tornadoes I've witnessed. I do know that you can't tell simply from looking at a thunderstorm whether it will spawn a tornado. And Keller was overly confident; another tornado a seemingly safe distance away may not have given him the chance to crawl into his cellar (or 'fraidy hole, as I've heard it called in the plains) and shut the door. Not all tornadoes are associated with "peculiar oppressiveness" or humidity, as Keller says. Nor is there necessarily any gassy odor associated with tornadoes. Perhaps what Keller smelled was gas from a line break. * * * The Fujita-Pearson Tornado Intensity Scale, formulated in 1971, rates tornadoes from F1 to F6 (Table 1.1). This scale estimates tornado intensity based solely on the degree of damage caused by tornadoes. It was designed to connect smoothly with the Beaufort scale, devised in 1806 by Sir Francis Beaufort, a British naval officer. The Beaufort scale gives wind speed from 0, meaning calm, to 12, for hurricane-force winds. The Fujita scale picks up where the Beaufort scale ends and eventually reaches Mach l, the speed of sound. An F6 tornado, according to Fujita, is "inconceivable." Tetsuya "Ted" Fujita is a meteorologist at the University of Chicago who would become one of the founding fathers of severe-storm studies. Fujita had begun his career in postwar Japan studying the pattern of damage from the atomic bomb attack on Hiroshima. The Fujita scale is not perfect. Estimating tornado intensity based on damage alone, not actual wind measurements, is risky. The structural integrity of the things hit by the tornado is often not very well known. A well-built structure can withstand very high winds, while a poorly built structure can suffer devastating damage even from less intense winds. Most tornado damage is a result of pressure induced by the wind. The pressure exerted on an object is proportional to the square of the wind speed and to a factor that depends on the shape of the object and the wind direction. Two identical objects that feel the wind from different angles may suffer different amounts of damage. Furthermore, many tornadoes occur over open country, where there is nothing to damage. Reports of tornado damage that imply incredibly high winds can be misleading. For example, people are amazed to learn that straw has been driven into wood posts and trees during a tornado. Experiments performed in the laboratory with a pneumatic gun have demonstrated that straw splinters can indeed be driven into soft wood by winds as weak as 60-70 mph. Much higher wind speeds are needed to drive the splinters into hard wood. There have also been reports of tornadoes stripping fowl of their feathers. In 1842, an American professor named Elias Loomis, presumably a scientist, conducted an experiment with the objective of determining how high the wind speed must be to blow all the feathers off a chicken. He did this because "the stripping of fowls [during tornadoes]," Loomis said, "attracted much attention." He continued: In order to determine the velocity needed to strip these feathers ... [a] six-pounder was loaded with five ounces of powder, and for a ball, a chicken just killed. The gun was pointed vertically upwards and fired. The feathers rose twenty or thirty feet, and were scattered by the wind. On examination, they were found to be pulled out clean, the skin seldom adhering to them. The body was torn into small fragments, only a part of which could be found. The velocity was 341 miles per hour. A fowl, then, forced through the air with this velocity is torn entirely to pieces; with a less [sic] velocity, it is probable most of the feathers might be pulled out without mutilating the body. More recent experiments indicate that the force needed to remove feathers varies widely, depending on the chicken's condition. There have been reports of tornadoes passing over ponds and subsequently "raining out" frogs. It is not difficult to imagine how a tornado can suck up the frogs along with some of the water and deposit them elsewhere, but I know of no experiments that have attempted to determine how strong a tornado must be to account for the frog-rain phenomenon. During the summer of 1997 two people and the cottage in which they were sheltered were lifted in a tornado and deposited in a nearby lake. The pair managed to swim back to shore and were hospitalized. One was reported to have remarked, "There was a hell of a bang, and I can remember saying `Where in the hell did all this water come from?'" I've been as close as a quarter of a mile to a tornado, but I've never seen airborne cows. I have seen dead cows on the ground after a tornado has passed through, and I have seen houses that have been knocked off their foundations. It may be that cows, houses, and people do fly through the violently rotating winds of tornadoes, but we usually can't see them because of the dark cloud of dirt and debris that tornadoes usually whip up. What would it take to lift a cow and toss it through the air? Consider that the speed at which air is rising in a tornado increases with height, from zero at the ground to some high value, say 100 mph, about three hundred feet above the ground. For a cow to become airborne, it would have to be lifted high enough so that the gravitational force acting downward on its body is more than counteracted by the upward drag force it would experience as it rises. But the upward wind speed is probably not great enough to lift the animal off the ground. More likely is that the horizontal wind speed at the top of the cow will be high enough to blow it laterally and flip it over; this might cause the cow to tumble and bounce high enough to reach the level at which it becomes truly airborne. The recent movie Twister , in which there is a flying cow, is a comic- book version of what meteorologists do. We do not put ourselves inside tornadoes to take measurements. This is simply not possible. Nor, as a rule, are we suicidal in our quest to observe a tornado firsthand. In terms of destructive power, tornadoes are the most violent storms on earth. Hurricanes and typhoons pack winds of 200 mph at most; the top speeds of tornadoes have been estimated at 300 mph. No instrument to measure wind speed directly has ever survived a strong tornado. Storms that spawn tornadoes may also bring hail plummeting from aloft at 70 mph, destroying crops, shattering windows, and pummeling anything and anyone in their path. The biggest hailstone I know of was found in Coffeyville, Kansas, on September 3, 1970; it weighed 1.67 pounds and was 17.5 inches around. Tornadoes occur throughout the world, over mountains, above plains and coastlines, in valleys, and over oceans, but they are by far most frequent in a hundred-mile-swath in the central United States that has come to be called "Tornado Alley." In this stretch, which extends from northern Texas and the Texas Panhandle through Oklahoma and Nebraska (Fig. 1.1), there are at least five tornadoes every year within a circle of radius of one degree of latitude-longitude (i.e., in a circular area approximately 50 miles across). Statistics on their frequency, however, must be viewed with a certain amount of skepticism. The data are probably biased toward urban areas, since it is more likely that tornadoes in remote parts of the world go unreported. Furthermore, counting tornadoes is not as straightforward as it may seem. Sometimes a series of tornadoes accounts for one long, continuous damage path. Sometimes a tornado will spin up for only a matter of seconds, disappear, and then reappear. If we included every such disappearing act logged in our records, we could produce a much higher frequency rate for tornadoes. There is no time of year and no time of day that tornadoes do not occur, but most are reported in the spring and early summer (Fig. 1.2), from midafternoon to early evening (Fig. 1.3), when the atmospheric temperatures at the ground are highest. However, tornadoes have occurred even over snow-covered ground. On Dec. 2, 1970, a tornado hit Timpanogos Divide, Utah, buried under 38 inches of snow. Most tornadoes rotate cyclonically, which is counterclockwise in the Northern Hemisphere and clockwise in the Southern Hemisphere. Exceptions to this have been documented. Although Northern Hemisphere tornadoes usually move from the southwest to the northeast, a tornado can move in any direction. The same tornado can change direction in midstream; the axes of some tornadoes appear to be stationary. A tornado typically lasts for less than half an hour, sometimes only several minutes or even seconds; its damage path may be as long as a hundred miles and as wide as two miles. Tornadoes affect very small portions of our planet and are short-lived, but they kill scores of people in the United States alone every year and cause hundreds of millions of dollars in property damage. (Statistics for other places are not as well known). Before 1950 the average annual death count from tornadoes in the United States often exceeded a hundred; in recent years, the figure has been well under that number, due probably to improved warnings and increased public awareness. But there are always exceptions. In what has come to be called the Palm Sunday outbreak of April 11, 1965, a family of tornadoes, at least thirty-seven and perhaps more than fifty, killed 256 people in Iowa, Wisconsin, Illinois, Indiana, Michigan, and Ohio. In another outbreak, on April 3-4, 1974, 148 tornadoes killed three hundred people in thirteen states stretching from Alabama to Michigan. On June 9, 1984, tornadoes killed four hundred in the former Soviet Union. In Bangladesh, a tornado killed more than five hundred people and injured some thirty-three thousand on May 13, 1996. Tornadoes come in so many shapes and sizes and guises that they seem to defy definition. What exactly is a tornado? Meteorologists are still trying to answer this question. The word tornado probably comes from the Spanish tornar , "to turn," and tornado , "thunderstorm." A tornado is commonly defined as a violently rotating column of air hanging from a tall, bubbly cloud from which rain is falling (a cumulonimbus). But I have seen tornadoes hanging from clouds that contain no rain. And "column" implies that the vortex of rotating air is vertically oriented; some tornadoes are stretched out horizontally on their sides (Fig. 1.4). Although most, but not all, tornadoes are associated with powerful thunderstorms, I've seen tornadoes come from weak thunderstorms. And not all tornadoes are spawned by thunderstorms. Hurricanes and typhoons, when they make landfall, can spawn tornadoes. To complicate the matter further, other meteorological phenomena--such as landspouts, waterspouts, and dust devils--are also violently rotating columns of air. The latter is not considered a tornado, but the first two are. So, what is a tornado? It is a violently rotating column of air, which may not be oriented vertically, that comes from beneath the base of a thunderstorm or a rapidly growing towering cumulus cloud. Tornado Research: Early History One of the earliest studies of tornadoes was a small pilot program conducted in 1950 near Washington, D.C. The goal was to test the hypothesis of Morris Tepper of the U.S. Weather Bureau, who suggested that the intersection of two pressure-jump lines--features on a weather map indicating a rapid rise in barometric pressure--is a preferred zone of tornado development. The experiment, dubbed the Tornado Project, was soon expanded to include parts of Kansas and Oklahoma. Apparently, tornadoes were frustratingly scarce, as the project's investigators noted. They wrote in their report: "Unfortunately for meteorological knowledge, the setting up of the Tornado Project system seems to have provided the people of Kansas with the best tornado insurance they ever had. For at the present writing (June), there have been no tornadoes in the `arc' area during 1952." It was finally determined that the majority of tornadoes are not related to intersecting pressure-jump lines. But the experiment marked an early use of mesonetworks, closely deployed surface instruments that measure such variables as temperature and barometric pressure. Surface measuring devices used operationally are spaced a hundred miles or more apart. In the mesonetwork, the instruments were one mile to ten miles apart, giving a more in-depth picture of atmospheric conditions. In the late 1950s meteorologists studied movies taken of tornadoes, usually by nonscientists. The first scientific analysis was made, in 1957, by Walter Hoecker of the U.S. Weather Bureau. Hoecker had acquired a film of a tornado ripping through Dallas, Texas, on April 2, 1957. Frame by frame, the film faithfully tracked swirling chunks of debris. Because he knew the distance from the camera to the tornado, he could compute how far the debris moved from frame to frame. Knowing these distances and the time interval between frames, he could estimate the wind speed. Using this kind of elementary geometry, Hoecker estimated wind speeds as high as 170 mph. Nobody had ever estimated the wind speed of a tornado before, although there had been talk that tornadoes could be supersonic, with wind speeds in excess of 700 mph. The same year, on June 20, a number of local citizens photographed a tornado in Fargo, North Dakota. One of the shots in this sequence of spectacular still photos showed an airborne automobile. The photographs made their way to Ted Fujita. From the Fargo pictures, he determined the relationship between a tornado and the architecture of cloud features associated with its parent storm. For his analysis, published in 1960, he coined the terms wall cloud, tail cloud, and others. But it would be years and many more photographs before the usefulness of his terminology would be appreciated. After all, the Fargo tornado was only one event; the scientific establishment was loath to apply Fujita's analysis to all tornadic storms, which are thunderstorms that spawn tornadoes. In the 1950s and 1960s aerial photographs of tornado damage were analyzed. A photograph of damage to a cornfield in Nebraska in 1955 shows a pattern of loops of flattened stalks. Meteorologist E. L. Van Tassel attributed the loops to scratch marks from large debris. Fujita later hypothesized, in 1970, that similar cycloidal (looped) or scalloped patterns of damage in grain fields from the infamous 1965 Palm Sunday outbreak were caused by "suction spots" that rotated around the tornado (Fig. 1.5). These spots supposedly sucked up debris from the ground as they rotated around the center of the tornado. What suction spots do is still controversial; it is not clear whether debris is sucked up or simply blown over during a tornado. On April 9, 1953, a weather radar in Champaign, Illinois, picked up a hook-shaped appendage on the southwest side of a tornadic storm (Fig. 1.6). Such a hook, which had been seen before but not captured by a camera, was conjectured to be associated with the circulation of a tornado. Meteorologists would later learn that not all hooks are associated with tornadoes, nor are all tornadoes associated with hooks. Meteorologists began using conventional (non-Doppler) radar in the late 1940s to study thunderstorm structure and behavior. It had been discovered during World War II that radar detects not only aircraft but also precipitation. Aircraft were small points of light on the radar screen and precipitation appeared as pulsating blotches. To the military, this was interference or "noise": It was not the enemy and did not shoot you, although precipitation in the form of large hail could indeed cause harm. To the meteorologist, radar illuminates the invisible by rendering it visible; it shows us the raindrops and cloud droplets up to several hundred miles away. At radar's centimeter wavelengths (which correspond to frequencies on the order of 3 to 10 GHz), raindrops (and cloud droplets) act as Rayleigh scatterers. Rayleigh scattering occurs when the wavelength is long compared to the width of the scatterers; the energy of the backscattered radiation decreases rapidly with increased wavelength. Precipitation backscatters to the radar, and this is the echo. (Electromagnetic radiation from radar is also scattered to the front and sides, but meteorologists usually aren't concerned with this because the radiation does not return to the radar directly, or at all.) The largest raindrops backscatter the largest fraction of the energy, while the tiniest raindrops and the cloud droplets backscatter the least. The actual power received back at the radar device is given by the weighted sum of the diameters of the drops or droplets raised to the sixth power. In other words, a raindrop 4 mm in diameter backscatters sixty-four times more powerfully than a raindrop 2 mm in diameter. Cloud droplets differ from raindrops in that the latter are large enough to have a fall velocity relative to the air. Cloud droplets, much smaller, do not--they just blow along with the wind. Radar bands in the 3-10 GHz frequency range are used because those with longer wavelengths are not as sensitive to raindrops and cloud droplets and require cumbersome antennas. Much more of the radiation from shorter wavelengths is absorbed by raindrops, so more of the beam is attenuated, and as a result, the sensitivity to rain at great distances is markedly diminished. Conventional radar can detect only the intensity of precipitation in a volume of air. Doppler radar can detect this and the rate of motion--that is, the speed of the scatterers along the line of sight of the radar. Doppler radar makes use of the Doppler effect, as expounded by Christian Doppler, a nineteenth-century physicist. The frequency of waves is changed by their motion relative to that of an observer. For example, consider the sound waves of a train's whistle: The pitch of the whistle rises as the train approaches and becomes lower as the train recedes. Light waves are another example. The visible spectrum of light from approaching stars shifts into the blue region of the spectrum, while the visible light from receding stars shifts into the red region of the spectrum. This red shift has been associated with an expanding universe in which virtually all galaxies more than three million light-years away are receding. Consider, finally, electromagnetic waves when they backscatter from raindrops and cloud droplets. The frequency of backscattered radiation shifts up if the raindrops or cloud droplets move toward the radar but shifts down if the water moves away from the radar. Doppler radar can sense whether the raindrops are moving toward or away from the radar, and this tells us how fast the wind is blowing because it is the wind that propels the raindrops. Suppose that raindrops, dust, and other debris are carried along by the wind in a tornado. The antenna of the Doppler radar concentrates electromagnetic energy into a narrow beam. If the beam is concentrated along an arc of 1 degree, then at a range of thirty miles, the beam will be spread out over a distance of about half a mile. Since all but the largest tornadoes are no wider than a mile or so, the radar beam will encompass most of the tornado. Since the left side of the vortex has scatterers (i.e., raindrops, dust, debris, etc.) that are approaching the radar, and the right side of the vortex has scatterers receding from the radar, the shifts in frequency of radiation backscattered to the radar are both up and down. In fact, if we simply compute the average Doppler shift, weighted by the intensity of the radiation at each frequency shift, we find that it is close to zero and the tornado would not be detected. But if we consider the spread of frequency shifts detected, the Doppler velocity spectrum , then we would see that some backscatterers are moving away and some are moving toward the radar. The maximum wind speeds in the direction of the line of sight from the radar can thus be determined (Fig. 1.7). Note that Doppler radar can sense only the motion along the line of sight of the radar, while the photogrammetric analysis technique can yield only the motion perpendicular to the line of sight. The first Doppler-radar measurements of a tornado were made in 1958 by U.S. Weather Bureau meteorologists R. Smith and D. Holmes. They obtained a Doppler radar from the U.S. Navy, set it up in Wichita, Kansas, and then waited. Just over a year later, on June 10, the radar detected a tornado in El Dorado twenty-five miles away. What luck that a tornado happened to touch down within range of the first Doppler radar used for meteorological purposes! The likelihood of a tornado striking within thirty miles of a given location during a given spring is extremely small. Although at the time it was not possible to discriminate between approaching and receding velocities, the wind spectra obtained by this radar detected maximum velocities (that is, wind speeds in the line-of-sight direction) of 200 mph. Smith and Holmes's radar was a continuous-wave (CW) radar at 3-cm wavelength. CW radars send electromagnetic energy continuously. An aural analogy might be a coyote that howls continuously in a canyon; the echo of the howl mixes with the howl going out. Thus, one cannot determine the delay between the time energy is radiated and the time backscattered energy reaches the radar. If the coyote were to howl in bursts, then, given the speed of sound, the time delay between each burst and its echo could be used to determine the distance to the canyon wall. A major drawback of these first Doppler-radar observations was that range information could not be obtained, owing to the way the signal was processed. Pulsed-Doppler radars can determine range. But in 1958 it was not yet possible to obtain both range and velocity from CW radar. The Doppler radar was not used after the first successful tornado measurements, and it was more than a decade before next major advances in Doppler-radar technology were put to meteorological use. Besides the remote measurements made by radar and the photographs and movies, measurements of wind and pressure were made serendipitously, when a tornado happened to pass near a weather station. Meteorologist Ed Brooks reported in 1949 that tornadoes often appear within low-pressure areas that are about ten miles across. He named these gyres "tornado cyclones." Only data from calibrated instruments should be trusted, and it is not always clear which of the early data are reliable. Anemometers tend to get blown away by tornadoes. One trace from an anemometer showed peak gusts of 145 mph when a tornado passed just north of a weather station during the Palm Sunday tornado outbreak. On May 27, 1896, a pressure deficit (deviation from ambient pressure) of 82 mb was measured very near the center of a tornado in St. Louis. A waterspout passing over a ship in 1958 dropped the pressure by 21 mb. On May 24, 1962, a pressure deficit of 34 mb was recorded near what was called a "tornado cyclone" in Newton, Kansas. An NSSL mesonet site also recorded a 10-mb pressure deficit some 1,300-2,600 feet from the center of a tornado on April 30, 1970. Although we don't know how good these measurements are, the largest credible pressure drop is probably about 100 mb, which is about 10 percent of the total atmospheric pressure near the ground. Pressure deficits in intense oceanic extratropical cyclones can be higher than 50 mb; pressure deficits in tropical cyclones can reach almost 100 mb. In the case of the extratropical and tropical cyclones, the large pressure drops cover much greater horizontal distances than those of tornadoes. Table 1.1 The Fujita F scale F number F-scale damage specification F0: 18-32 m/sec (40-72 mph): Light damage; some damage to chimneys; break branches off trees; push over shallow-rooted trees; damage signboards F1: 33-49 m/sec (73-112 mph): Moderate damage; the lower limit (73 mph) is the beginning of hurricane wind speed; peel surface off roofs; mobile homes pushed off foundations or overturned; moving autos pushed off the road F2: 50-69 m/sec (113-157 mph): Considerable damage; roofs torn off frame houses; mobile homes demolished; boxcars pushed over; large trees snapped or uprooted; light-object missiles generated F3: 70-92 m/sec (158-206 mph): Severe damage; roofs and some walls torn off well-constructed houses; trains overturned; most trees in forest uprooted; heavy cars lifted off ground and thrown F4: 93-116 m/sec (207-260 mph): Devastating damage; well-constructed houses leveled; structures with weak foundations blown off some distance; cars thrown and large missiles generated F5: 117-142 m/sec (261-318 mph): Incredible damage; strong frame houses lifted off foundations and carried considerable distance to disintegrate; automobile-sized missiles fly through the air in excess of 100 meters; trees debarked; incredible phenomena will occur F6-F12: 143 m/sec to Mach 1, the speed of sound (319-700 mph): The maximum wind speeds of tornadoes are not expected to reach the F6 wind speeds (Continues...)
