by Stephen S. Hall
[Editor's note: This 1992 telling of the Landsat story appears courtesy of the author, Stephen S. Hall. This article has not been reviewed for conformity with the U.S. Geological Survey's editorial standards, and opinions and conclusions expressed herein do not necessarily represent those of the U.S. Geological Survey. Stephen S. Hall, 1992, Mapping the next millennium; the discovery of new geographies: New York, Random House, p. 52-70. This book is available in paperback from Vintage Books. Copyright 1992, Stephen S. Hall. Reprinted by permission. This file was created by optical text recognition, and minor edits were made for HTML presentation.]
--Montaigne, Essays
On a warm August night in 1972, a barge burdened with the vilest of industrial wastes made its way slowly from the mid-Atlantic coast of the United States toward a watery refuse dump located about twelve miles offshore and known as the New York Bight. With the anonymity conferred by dark of night on open sea, the barge stopped short of its designated dumping area and proceeded to flush thousands of gallons of acid iron wastes into the Atlantic Ocean. The dumping took a matter of minutes. Then, with a tight hairpin turn and with the complicity of that same darkness, the barge chugged back to shore, leaving the scene without a trace. Or so it seemed. At approximately nine o'clock the following morning, August 16, a satellite launched only a few weeks earlier glided silently over the East Coast of the United States at an altitude of about 570 miles. With the satellite traveling at more than 14,000 miles per hour, its instruments made thirteen sweeps a second over the still waters of the New York Bight, canvassing 13,000 square miles of ocean and land in a mere twenty seconds or so, forming a picture in the morning light. There was no trace of the barge by then, of course. But there remained a map of its bad intent. Instruments aboard the satellite revealed a smear of white acid on the dark and pristine sea. The satellite showed that the dumping had occurred too close to shore; indeed, it even revealed the composition of the wastes as 8.5 percent sulfuric acid and 10 percent iron sulfates. Tugged by currents, the trail of acid curled back and forth. From space, it took on the appearance of a sinuous question mark.
The question it posed, then and now, is how attentively we wish to monitor the ongoing abuse of our planet. One of the earliest replies came in the form of that satellite. Known at the time as the Earth Resources Technology Satellite (or by its technological oink of an acronym, ERTS), it would eventually adopt the name Landsat, and that photomosaic of the New York Bight was one of its very first pictures sent back to earth. The image of mankind's polluting autograph, scribbled recklessly on the continental shelf, captured in one blink of the technological eye all the power of mapping from space, formally called "remote sensing." Now, nearly twenty years after Landsat's spectacular debut, we can look back at that image of the acid dump in the New York Bight and see how it anticipates the crucial role remote-sensing devices are destined to play in the future as we struggle to map and monitor and manage resources on earth that seem, with each passing decade, ever more precious, ever more finite, and, alas, ever more identifiable to exploiters as well as conservators because of those same satellites.
Long before Mission to Planet Earth, there was Landsat. The first remote-sensing satellite in this series was launched in July of 1972. Within days it had stunned geologists, rewritten the textbooks of forestry and land use, and revolutionized cartography. Its earliest transmissions revealed several dozen earthquake faults near Lake Tahoe and Monterey Bay in California, faults that had escaped the notice of geologists who had crawled over that same shifting seismic terrain for decades. Undiscovered lakes winked up at the camera. One early image Of Alaska showed a forest fire eating through a forest north of Fairbanks; nothing unusual about that, except people in Fairbanks didn't know about it. A ship stranded by chance in the Arctic Ocean received Landsat's satellite maps of the ice pack and picked its way to freedom. Previously unseen terrestrial patterns took on the beauty of modern art and modern metaphor: there were peeling fragments of landscape around Elephant Butte, Montana, that reminded one of Clyfford Still, Cairo as it might be imagined by Helen Frankenthaler, a network of Russian roads imitating astrocytes in the mammalian brain. From those first heady days, Landsat images have graduated to the level of cultural artifacts with coffee-table status: the posters hang in offices and homes, the maps have generated entire atlases, the images fill year after year of calendars. The satellite taught us, NASA administrator James C. Fletcher said at the time, "a new way to look."
The maps generated by Landsat, perhaps more than any other type of map discussed in this book, bridge the gap between old and new cartography. Landsat images cover common ground. They represent the ultimate refinement of fifty years of aerial photography and mapping, albeit in grander and more sophisticated terms; they are photographs and maps at the same time, and so they speak the age-old cartographic vernacular of hill and valley, basin and range, fault line and fracture zone, with perhaps crisper, more precise optical pronunciation. On the other hand, many Landsat images mark a qualitative breakthrough in mapmaking. The Landsat map showed not just the lay of the land, but-- through a happy convergence of technical tricks, computer enhancement, and interpretation-- the quality of the land: how it was used, what grew on it, whether the soil was dry or moist. Since the launch of Landsat 4 in 1982, there has even been an instrument circling the planet whose very name sums up this new qualitative cartography. It is called the Thematic Mapper.
Perhaps most important, Landsat has provided us with the most obvious and accessible benefits of remote sensing from space. Remote sensing refers to the use of robotic instrumentation, an eye in the sky in this case, to replace or enhance the human senses, and thus can be seen as a metaphor for how we explore and chart the physical world in the late twentieth century. Just as Voyager 2 is a robotic eye wandering through the solar system, just as radio telescopes are huge metallic ears listening for the dry rattle of radio waves in the cosmos, just as radioactive DNA probes poke like fingers through the maze of chromosomes; just so, the late-twentieth-century explorer need not stand on the bridge of a heaving ship or traipse dry, uncharted waste lands like explorers in earlier epochs. He (or she) more likely can be found at a computer terminal, manipulating data sent by a far-flung probe surfing the solar wind, by a quantum subatomic stylus riding the grooves of atoms like a phonograph needle on an LP. In the world of remote sensing, the vessels of discovery are instruments; our eyepiece is the computer. No enterprise of mapping has made that more apparent to more people than Landsat.
And no enterprise had to overcome more obstacles to prove the point. Landsat almost never happened.
The folklore surrounding the project and its origins offers up plenty of heroes and plenty of obstructionist villainy. Certainly one of the most visionary of the scientific pioneers was Robert N. Colwell, a native of Star, Idaho, and professor of forestry at the University of California at Berkeley. It was not simply that Colwell was one of the early experts in aerial photo interpretation, nor that he'd worked with Lockheed engineers on a secret Air Force satellite project called SAMOS in the mid-1950s. It was his premonitional realization that scientists could peer from a privileged, elevated perch-- a balloon, an airplane, ultimately a satellite-- and eavesdrop on all the electromagnetic chatter reflected or emitted by objects on the earth's surface, a chorus of voices ranging from X rays and infrared to radar microwaves. "Certainly, lots of people had the same kind of thoughts," Colwell said recently in his careful, erect manner of speaking, "though maybe not so gung-ho as me."
How gung-ho? Colwell, who would go on to head remote-sensing research at the Space Sciences Laboratory at UC-Berkeley, was willing to buzz the California state capitol building in Sacramento to make his point. Actually, he made low passes over both the capital and agricultural test plots in the 1950s in a Cessna 180 to prove the value of remote sensing. Peering down on the landscape from several thousand feet, he showed that by photographing a scene in several different wavelengths and later combining the images, you could distinguish grass from soil, even cement from asphalt, in black-and-white photographs. Colwell and others had shown that agricultural crops, trees, even different soils possess a telltale electromagnetic signature, as distinct as the handwriting of different individuals. This signature depended on how each type of plant reflected radiation in the near-infrared region of the spectrum, which is a frequency of electromagnetic radiation just beyond the gathering capacity of the human eye.
If we could see the world with infrared eyes, vegetation would appear to us not green but red (infrared detectors gained great importance during World War II precisely because of their ability to distinguish between real vegetation, which creates an infrared signal, and green plastic camouflage covering a tank, which does not). With instruments known as spectrometers, researchers could image a field of corn, for example, in a particular infrared wavelength; the image differed quite significantly from a regular black-and-white photograph. A Soviet scientist named E. L. Krinov had measured the spectral signatures of some 370 natural and man-made objects in the 1940s; he, too, had aerial mapping in mind. By the early 1960s, researchers had demonstrated that corn could be distinguished from wheat solely on the basis of its spectrographic signature, whether viewed from two thousand feet or five hundred miles above ground; soils heavy in clay could similarly be distinguished from sandy soil. Colwell even managed to show that diseased or stressed plants, like pale and infirm humans, had a different spectral tone than healthy plants, and that you could discern the difference more easily, and earlier, with an infrared sensor in space than if you inspected the plants at ground level and rolled the diseased leaves between your fingers. In the early literature of remote sensing, one detects a growing cockiness in the pioneers as they muse about distinguishing pigweed from soybeans, even a fat pig from a calf.
In a prescient article that appeared in American Scientist in 1961, Colwell described not only his own work along these lines; he intuitively grasped that emerging technologies in other fields would forever change humankind's ability to measure-- and therefore to map-- the physical universe. He cited the use of fluorescent and radiographic techniques to map biological objects microscopically small and nearby, the use of radio waves to detect celestial objects astronomically large and distant. Each different wavelength spoke to us in a different language, contained different information, if only we had the electronic ears to hear. "Just as our musical appreciation is increased greatly when more than one or two octaves are exploited," Colwell wrote, "so also is our appreciation of the physical universe, through multiband spectral reconnaissance, which already can exploit more than forty 'octaves.' " The first Landsat went up with just such a multiband spectrograph, thanks to a persuasive effort by Colwell and colleagues at the University of Michigan and Purdue, and its extra octaves have given scientific cartography the richness and tonal nuances that have emerged in the past thirty years.
But it was a Russian development that speeded up remote sensing from space. Even before October 4, 1957, that infamous date when the Soviets launched Sputnik 1, scientists-- especially agricultural scientists-- had been aware that aerial surveys revealed features of the earth that were either too hard to obtain on the ground or too costly and time-consuming. The space race merely stepped up the pace.
Satellites represented the other main stream of research at the time. Shortly after World War II, U.S. Army scientists obtained the first photographs of earth from space-shots of White Sands, New Mexico, and environs taken by captured German V-2 rockets, and so even before Sputnik, scientists had prepared wish-lists of possible applications for space satellites. What remained unclear in the initial days of remote sensing was the "platform," as they liked to call it-- the vehicle from which to peer down at the earth. Airplanes and balloons were the obvious candidates in the 1950s, but as early as 1955 the air force began planning what it called a Strategic Satellite System. Remote imaging of the atmosphere from space began in 1959 with the launch of the Explorer VI satellite, which sent back televised pictures of meteorological conditions; the first Tiros weather satellites, launched in 1960, used television cameras to return even better pictures of cloud movements.
A satellite's ability to take in very large, or "synoptic," scenes made it economically attractive. With sophisticated spectral-detecting technology, its power to convey information was truly remarkable. In principle, a mechanical device located five hundred miles above the surface of the earth could, with several spectral wave bands, reveal everything from nuclear missiles in Cuba to the average household income in any given neighborhood on earth. That is why people like Colwell wanted it; it is also why people in the military did not want civilians to have it. As early as 1783, when as ambassador to France Benjamin Franklin observed the first Montgolfier balloons in flight over Paris and predicted their use in "conveying Intelligence" about "an Enemy's Army," generals and spies have understood the power of aerial reconnaissance; the high-altitude surveillance possible from satellites promised both a revolution in surveillance and insurance against political embarrassments like the downing of U-2 spy pilot Gary Powers in 1960.
The two streams converged in February of 1962, when the Institute of Science and Technology at the University of Michigan, a research organization funded by the military, invited seventy agronomists, agriculturists, foresters, geologists, hydrologists, land-use experts, photogrammetrists, and of course cartographers to the First Symposium on Remote Sensing of Environment. From that date, civilian remote sensing as a discipline began to come of age. On that same date, the two converging streams began to back up behind a dam of security and secrecy. Scientists at that very first symposium did not bemoan the technical impediments that lay in their path; scientists rarely do. They complained about how the military scrambled to classify information about remote sensing.
"This came up over, and over, and over again," one symposium participant remarked, "and it was the one single wet blanket on the whole conference." All this was duly noted by William Fischer, a photogeologist from the United States Geological Survey, who at tended that first meeting. Fischer returned to Washington and suggested to his boss at the Department of the Interior, William T. Pecora, that the time might be right for a remote-sensing satellite.
The fact that Landsat almost never happened is a reminder that maps-- from Columbus's chart of the New World to that satellite view of the New York Bight-- depend to some degree on political resolve; Landsat's most persevering bureaucratic angel was Bill Pecora of the U.S. Geological Survey. He not only overcame military obstructionists but, it now appears, bluffed the National Aeronautics and Space Administration into reluctant cooperation, too.
NASA's participation was critical. If there were any lingering questions about the promise of space observation, they were dispelled-- to everyone's considerable surprise-- by John Glenn, Wally Schirra, and Gordon Cooper. Despite cramped conditions, these astronauts took large-format cameras aboard America's early space flights and came back with breathtaking photographs of the earth shot through the window of their Mercury and Gemini spacecraft. These were ordinary photos using visible light, not the information-rich spectral images pushed by the remote-sensing crowd, but the surprising amount of clarity and detail excited geologists. More important, the images excited the public.
Probably the most heralded episode occurred during the twenty two orbits of Mercury 9 in May of 1963, when astronaut L. Gordon Cooper, endowed with extraordinarily acute 20/12 vision, peered through a 70-mm Hasselblad camera and reported seeing individual houses in the plains of Tibet, even smoke curling out of chimneys. If that didn't strain credulity, there was his claim to have seen dust clouds kicked up by a vehicle traveling on a road in the southwestern United States. As John Noble Wilford writes in The Mapmakers, "Since the physiological effects of space flight were at that time largely unknown, a number of scientists suspected that Cooper, under the influence of prolonged weightlessness, had suffered hallucinations." But his claims were later substantiated. If the engineers could come up with a detector as sensitive as Gordon Cooper's eyes, remote sensing would have a great future.
But NASA administrators were not interested in a civilian satellite program, and the most generous interpretation for their reluctance is that it would have distracted time and resources from their mandate to put a man on the moon. A less generous interpretation would be that NASA cocked too willing and deferential an ear in the direction of the intelligence community. And so from the very beginning, Landsat was a political stepchild, sought by scientists but shunned by the military, claimed by the Interior Department but coveted by Agriculture, a nuisance to NASA and undermined at almost every step of the way by the Bureau of Budget (later the Office of Management and Budget). "There was a lot of pressure not to get it off," says Alden Colvocoresses, a senior cartographic expert at the U.S. Geological Survey and, in the 1960s, a military officer on loan to NASA. "The military and the National Security Council didn't like the idea of the civilian community looking at the earth at all. That was their prerogative. They wanted to keep this classified." Landsat historian Pamela E. Mack goes so far as to suggest that "NASA may have had a secret agreement with the Department of Defense or the National Security Council to go slowly on earth resources satellites."
Colvocoresses, and many others, credit William Pecora with forcing the issue. Born near Newark, educated at Princeton, and a member of the U.S. fencing team at the 1936 Olympics in Berlin, Pecora was a big-eared, balding, blunt-spoken man with a genial smile. A respected geologist, he joined the USGS in 1939, discovered a new mineral-- pecoraite-- in Brazil, intended to teach, but got waylaid by the federal bureaucracy; he ended up as director of the U.S. Geological Survey in 1965. From the moment Sputnik went up in 1957, Pecora wanted a satellite to monitor the earth and its resources; when his close aide William Fischer returned with reports from the first Michigan symposium, the desire became an obsession. And when higher-ups in NASA and the government didn't seem to share his enthusiasm, after several years of feasibility studies, Pecora decided to take matters into his own hands. He set up an impromptu and now-legendary press conference on September 20, 1966, at which his boss, Interior Secretary Stewart Udall, announced Interior's plans for the Earth Resources Observation Satellites. The EROS program, Udall said, would be run in cooperation with NASA. This came as a big surprise to NASA, which was pushing a plan for a manned observation platform and had little prior warning, apparently, about the press conference. The space agency was committed to manned missions and trying to put a man on the moon, and had no taste for receiving, storing, and analyzing data from an earth-monitoring satellite. Pecora announced that the first satellite could be launched by 1969 at a wildly optimistic cost of only $20 million. The story made the front page of the New York Times the following morning. Pecora was delighted. He was also convinced, he told his wife, that he would be fired the following day.
"The next day all hell broke loose," recalls his widow, Wynn Pecora. "The White House was ready to kill him, the Pentagon was ready to kill him, and the State Department was ready to kill him." "NASA and the military intelligence community were furious," confirms Colvocoresses. "The intelligence community felt it could keep a strong handle on NASA, but they were shocked when Interior got involved in this. Pecora nearly got fired, but President Johnson supported his action and so did Congress." Once the idea received an enthusiastic public response, NASA's policymakers decided they wanted a satellite, too, and geared up to coordinate the construction and operation of EROS in 1967 ("That the initials form the name of the Greek god of love and fertility," the Times article noted, "is a coincidence"). As Richard Mroczynski, who has worked on Landsat since the first launch, bluntly puts it: "NASA was the reluctant benefactor of a great thing."
Responding to continuing pressure from Interior, NASA began planning for a small unmanned satellite in early 1967. Then, just as the political waters began to calm, a technological dispute erupted. It would turn out to be another example of how a small piece of hardware-- a ball bearing, an 0-ring, a fan-jet part-- can determine the success or failure of a machine, a mission, an entire program.
The argument was all about the mirror. In 1968, out of the blue, NASA received an unsolicited bid from a group of scientists at the Hughes Aircraft Company in El Segundo, California. NASA and the Department of the Interior planned to use television cameras as the main sensors. The Hughes group, headed by a woman named Virginia T. Norwood, wanted to put a different type of mapping device on the proposed satellite. Known as a scanner, it featured that bete noire of aerospace engineers-- a moving part. The part that moved was the notorious "banging mirror." It measured nine by thirteen inches, and was designed to slam back and forth thirteen times a second. The Hughes team argued that it would perform better than television cameras.
"We were pushing a scanner because a scanner gives you some advantages," Virginia Norwood explained one day in her office. She is a handsome, robust woman with a hearty smile and a tan that takes longer than a vacation to get; pictures on her office wall established that she was a grandmother, just as technical papers in the literature have long established her crisp competence as an engineer, and you could hear both qualities, the sweetness and the toughness, in her voice. A self-described "Army brat," she was born in Fort Totten, an army base tucked away in a corner of Queens, New York. One of seventy women among five thousand undergraduate and graduate students at the Massachusetts Institute of Technology, she graduated in 1947 with a degree in mathematical physics and sought a job in private industry, only to get a rude introduction to the realities of the American workplace; her salary requirement stunned one prospective employer, who howled, "No woman in our plant has ever gotten that kind of money!" So she took a job with the U.S. Army Signal Corps, and worked her way up the electromagnetic ladder, as she puts it, specializing first in antennas, moving up to communications, and finally graduating to optics. Norwood joined Hughes Aircraft in 1954, where she worked for thirty-five years. She contributed to the Surveyor missions to the moon as well as on the microwave circuitry for Falcon missiles. By 1968 she had for many years worked in the space and communications division, where the design for the Landsat scanner was developed.
Norwood and Hughes entered the game late. By 1968 the Interior Department had formally requested congressional money for its Earth Resources Observation Satellites program, and NASA planned to adapt one of the Nimbus weather satellites for the job. The key mapping instrument aboard the satellite would be three RCA television-type cameras known as return beam vidicons, or RBVs. These detectors operated like a camera: they took whole-frame snapshots of the earth from airplanes, and each "saw" a different wavelength of light (green, red, and infrared). "People thought it would be straightforward to adapt that to remote sensing," Virginia Norwood recalls, "but it wasn't so darned straightforward."
In April of 1968, Norwood's group at Hughes first made its unsolicited bid to NASA. The company had developed a different type of meteorological sensor, called a scanner, for the early advanced technology satellite (ATS) weather satellites; this type of detector did not take snapshots of an entire scene. Rather, it scanned the landscape, line by line in rapid succession, as the spacecraft moved over it; computers back on the ground later assembled the complete scene out of the digitized data in each line. "We were an afterthought," Norwood admits. "NASA was planning to use the vidicons, but some of us just felt the scanner was the way to go, so we pounded. We visited people, showed them the idea, demonstrated it."
Since the satellite was destined to be rather small, the Hughes mapping device had to squeeze into what little room was left. The device was a compact, 120-pound package that looked like a telescope that never quite got removed from its packing crate. It possessed a nine-inch mirror much like a Cassegrain telescope; as the satellite flew over the earth and light reflected up from the surface, the infamous moving mirror scanned the landscape underneath it, sweeping back and forth in a motion likened to a person sweeping with a broom while walking across a room; its movement directed lines of light onto a focal point in rapid west-to-east scans that covered approximately 115 miles. They called it the multispectral scanner.
The scanner took big, quick swallows of data. In twenty seconds, the MSS could map a patch of land 100 nautical miles by 100 miles, and in so doing generate 7.6 million digitized picture elements, or pixels, for later visual display. It could swallow all of New Jersey in little more than thirty seconds; it could survey the coastline from Maine to Miami in about six minutes. The scanner resolved objects about 79 meters, a bit smaller than a football field.
But that kind of resolution was not the instrument's strong suit. The scanner, a four-eyed creature, embodied all the research done by the agricultural researchers-- Colwell, J. Ralph Shay at Purdue, and Archibald Park at the Department of Agriculture among others; it distinguished four different spectral wave bands of light. The scanner optics directed filtered light over four different detectors; these separated the incoming light waves just as a change separator distinguishes coins, but instead of nickels and dimes and pennies and quarters slipping into their respective slots, different wavelengths of reflected light fit into the separate spectral detectors. Two were for visible light; two were in the near infrared. Not only did each different wavelength create a different picture of the same landscape; analysts back on the ground could mix two, three, or four of those wavelengths together in one map, providing different information about the same landscape. Norwood's group knew how to build a scanner, but they didn't know what spectral bands would convey the best information about the earth, which is why, on one occasion, they carted bags of groceries (tomatoes, bell peppers, brussels sprouts, broccoli, and strawberries) from a market in Santa Barbara back to the lab to measure their spectral signatures.
Each spectral slot had been carefully chosen to tell something unique about the ground it covered. The wavelengths varied by as little .1 micron, or 1/10,000 of a millimeter, yet divined a rainbow of colors and a pot of gold in terms of information. Norwood quickly learned the immense versatility of the instrument. "The blue band, around .5 micron, is good for water quality," she explains. "The band between .5 and .6 micron gives an indication of the growth stage and vigor of crops. The band from .63 to .69 micron measures chlorophyll absorption. The band from .79 to .90 micron measures the water content of plants; we call that the 'blooming band.' That's where mature plants are brightest in the infrared spectrum, and it's a measure that can be used to assess biomass and stress. That is why false-color maps of vegetation are so red even though they're showing something green. And the band from 1.55 to 1.75 microns measures soil moisture. This is also the band that distinguishes clouds from snow and is used to measure the snowpack." The Hughes group crammed all this into an ambitious, six-band scanner. Then NASA, short of space on a satellite that was designed only to house the TV cameras, sent Norwood back to the drawing board. In 1971 the Hughes team came back with the more modest four-band multispectral scanner (a seven-band Thematic Mapper went up on Landsat 4 and Landsat 5).
The Hughes scanner triggered quite a debate. It was one of those behind-the-scenes, nitty-gritty, technical contretemps to which the general public is completely oblivious; yet, decades later, anyone who has marveled at the images from space can be grateful that the debate turned out the way it did. NASA scientists and mappers from the U.S. Geological Survey favored the vidicon cameras. Critics of the scanner expressed grave concern about that banging mirror and about the reliability of the mapping information it would provide. "Mapmakers like myself were very suspicious of the multispectral scanner, which we could not believe would have geometric integrity," admits Colvocoresses. "We were wrong on that one." "People were so emotionally opposed to a mechanical device," Norwood recalls. One wonders if the MSS might have gotten a slightly warmer reception if the engineer arguing its merits had been a man and not a woman.
In any event, the argument raged for well over a year and turned out to be unexpectedly crucial. "When Landsat was being conceived, the argument was: should we use RBVs or these newfangled, untested instruments on this satellite?" recalls David Landgrebe of Purdue University, an expert in the agricultural applications of remote sensing and an early advocate of the controversial scanner. "The whole community felt much stronger about a framing type of instrument than a line-by-line instrument, because their background was in aerial photography. Naturally, that was something they were more comfortable with. The government, in its usual fashion, decided to compromise and put on both."
So on July 23, 1972, a gloomy, overcast day in California, dignitaries and scientists (including Virginia Norwood) gathered at Vandenberg Air Force Base near Lompoc to watch as a Delta rocket lifted a refitted weather satellite, now bureaucratically burdened with both the vidicon cameras and the multispectral scanner, through the low cloud deck and into what science writer Boyce Rensberger called "a new era of earth exploration." This first Earth Resources Technology Satellite (later renamed Landsat 1) settled into an orbit over the poles approximately 570 miles above the earth. The orbit design itself was an unheralded work of art: as the satellite descended from North Pole to South over North America, for example, the earth would be spinning, so that the satellite's mapping instruments would cut diagonal swaths, and in such a manner the satellite could cover the entire planet every eighteen days. Moreover, every time the satellite crossed the equator, the local time was approximately 9:30 A.M., so each location was always photographed under identical lighting conditions. So great was the pressure to launch that there was an improvised quality to it-- the initial data, for example, would be beamed down to a computer originally designed to write checks.
The raging debate over the mirror resolved itself within hours of the launch. The TV cameras became afflicted with what officials later described as an "unexplained electrical problem"; within a month of launch, mission controllers quietly shut down all three vidicons. The scanner, on the other hand, worked perfectly; designed to last one year, it was operating ten years later. David Landgrebe, an advocate for the scanner, would later remark: "If they had decided to go just with the RBVs, I don't know that we'd have heard any more about this enterprise." Allen H. Watkins, who heads the EROS Data Center in Sioux Falls, South Dakota, where Landsat data and images are archived, says that would not have terminated the project. "But if ERTS had not functioned, if the multispectral scanner had not been on board," he admits, "it would have been very damaging."
The gloomiest thing about the launch was that it had been delayed for several months because of technical problems, and just four days before the successful lift-off, William Pecora died in a Washington hospital of surgical complications at the age of fifty-nine. He never saw his bird off, but one prediction he made would come true within a matter of days. "Sophisticated in program execution it is indeed," he promised, "but the products of its efforts are practical and wide-ranging-- and understandable."
One of the most powerful benefits of a map is its sheer "understandability." The human eye seems to have an aptitude, an understanding, of what a map conveys, and what makes it beautiful, and that seems especially true of Landsat images, which were immediately understood and appreciated by an immensely large audience. The first image from Landsat I came down two days after launch. It showed the Lake Texoma area on the Oklahoma-Texas border, about seventy-five miles north of Dallas. David Landgrebe compared the image with what mappers call "ground truth"-- knowledge about the types of trees and crops in a given region of land, for example. Landsat I passed muster. In its first three months of operation, the satellite showered fifty-three thousand photomaps down to earth.
No one could have predicted in 1972 just how versatile Landsat I and its successors would prove to be, or how many unexpected and sometimes profound insights they would fire down to us in their eighty-four-megabits-per-second cataract of data. The most straightforward Landsat images belong to the grand tradition of geographical and topographic maps; in the crisp folds and wrinkles of mountain ranges, the shimmer of shallow seas, the traditional iconography of mapmaking is preserved. Photomaps of this precision virtually merged the ancient cartographic skill of scaled illustration with the photograph: by calibrating with ground coordinates, the picture becomes the map. Paul D. Lowman, Jr., a research scientist at NASA's Goddard Space Flight Center, remarked, "Our first discovery from ERTS is that all our maps, topographic as well as geographic, are out of date." Even the defense establishment, once a bitter adversary, had to agree: the CIA and the Defense Mapping Agency began to use Landsat data because it was cheaper than data acquired by classified satellite, and the military has ultimately become Landsat's heaviest user.
Landsat stood traditional cartography on its ear. By seeing the world in electromagnetic increments beyond the normal range of human vision, Landsat revealed whole new worlds hidden within the folds of a familiar world we thought we knew so well. It was no great surprise that Landsat, from 570 miles up in space, could discern the health of corn plants on a half-acre plot of land in Iowa, could determine which were suffering from corn blight and where a fungal infestation appeared to be gaining a foothold. That is exactly what it was designed to do.
But even the most visionary and optimistic were surprised at how well it could detect a pink bollworm infestation in the cotton crops of California's Imperial Valley; that it would show the extent of defoliation by gypsy moths in a Pennsylvania forest; that it would outline the perimeter of grizzly bear habitats in Montana; that its infrared sensors would detect the ominous glow of the damaged reactor at the Chernobyl nuclear plant in 1986 and by similar means predict the imminent eruption of a Guatemalan volcano; that it could detect skid marks left by jets at the ends of airport runways and the morning shadow cast by the Washington Monument; that conservation groups would use its data to map the habitats of waterfowl in North America, and that states would use those photogenic stains of pollution, like the slick spotted in the New York Bight in 1972, to sue private industry for illegal chemical discharge; that entomologists could predict mosquito populations in inland Florida and epidemiologists could predict outbreaks of Fasciola hepatica, a liver fluke afflicting cattle, in the coastal zones of Louisiana; that a satellite could be used to gauge the affluence of urban and suburban neighborhoods (by measuring the amount of green space) and to chart the spread of that suburban fungus, the shopping mall; that it could look down at a stretch of coastline, in New York or Florida, and show us the stains of careless modern industrialism upon our waters. For a relatively modest cost, roughly $1.5 billion spread over twenty years, we have procured 2.5 million pictures, the most comprehensive and informative photo album of the planet ever assembled. And, military objections notwithstanding, the most democratic. Any citizen from any nation could obtain one of the original Landsat images of any spot on earth for $1.25. With the success of Landsat 1, NASA managed to find funding for additional satellites. Landsat 2 went up in January 1975, Landsat 3 in March 1978. The improved Thematic Mapper, again engineered and built by Virginia Norwood's group at Hughes, went up with Landsat 4 (in July 1982) and Landsat 5 (in March 1984), providing even more sparkling images. But the administration of Jimmy Carter instituted a phased plan to privatize Landsat, and that operation was accelerated, with perhaps ruthless dispatch, by the Reagan administration, which transferred operational responsibility for the satellites in 1985 to the Earth Observation Satellite Company, or EOSAT. Landsat was an orphan once again. John Pike, a space analyst for the Federation of American Scientists, says the "mania for commercialization in the early 1980s represented the triumph of ideology over common sense," and in the case of Landsat was "like throwing the thing into the deep end of a swimming pool and telling it to sink or swim, without even checking to see if there was any water in the pool."
An administrative orphan since inception, Landsat faces an even more hostile and indifferent world now. There is competition from abroad, in the form of the French SPOT remote-sensing satellite, which provides ten-meter resolution using an American-developed solid-state detector called a multilinear array), and there is indifference at home, in the form of a Congress that tried to shut down Landsat 4 and Landsat 5 in 1989 to save money, to say nothing of a blase public that has perhaps seen too many calendars and posters. Echoing its birthing pains, Landsat's very existence is again at the mercy of the political process. Only those people with long memories appreciated the irony in a short story about Landsat in the New York Times in November of 1989; the item reported that government officials were considering a proposal that would place the greatest civilian mapping enterprise in the history of civilization under the control and jurisdiction of the Department of Defense, whose interest has only increased since the Persian Gulf War. While Landsat's fate remains uncertain, the scientists-- as they always do-- are dreaming the next big cartographic dream: a stereo mapping satellite that can make three-dimensional maps of the earth.
Despite its many well-documented problems, Landsat was a spectacular vindication of remote sensing. One could argue that the greatest benefits of the space program have come not from hurling a few men and women into space for a few spectacular photo ops and long-distance sound bites, but from those satellites that, year in and year out, steadily and unspectacularly help us understand our planet better, improve lives, and, as in a recent program to identify breeding areas of the insect that causes sleeping sickness, even save lives. There is a place for manned flight, but over the long haul it can't compete with the satellites. Former astronaut Walter Schirra admitted as much back in 1972, when he said, "We should start looking down instead of up."
"We should send a man up when he can do something that an unmanned thing cannot do, or a robot cannot do," Virginia Norwood lamented one afternoon in her office. "Manned missions are so expensive, but they seem to stimulate greater interest in the public. The man on the street is not the least bit interested in Landsat. He doesn't even know what it is." That public indifference may be the most unfortunate thing of all, because as two decades' worth of remarkable images attest, Landsat has always had the right stuff, and we can now see that, compared with the space shuttle, it has the right stuff at the right price.